Adherence Conference 2014 Program

WELCOME

Welcome to the 9th International Conference on HIV Treatment and Prevention Adherence. We are honored to serve as the conference’s Scientific Co-Chairs and thrilled to celebrate with all of you numerous advances that have been made in biomedical and behavioral sciences, community alliances, and overall commitment towards reaching new levels of excellence in HIV prevention and treatment around the world, centering on the essential role of “adherence” in the broadest context. As you will learn through the conference plenaries and the oral and poster abstract presentations, these advances are attenuated by continued disparities affecting at-risk populations as well as “leaks” across the continuum of HIV care. Together, we are confident that these next few days will provide opportunities for discovery, new insights, and collaborations that will shape the future of research and practice centered on the complexities of HIV-prevention, linkage to quality HIV care, subsequent ongoing HIV care engagement, and antiretroviral therapy adherence.

Our ninth year of open discussion across disciplines, organizations, and individuals will continue to distinguish this conference as an exceptional example of collaboration between research, practice, policy, and community. We wish to sincerely thank the Planning Committee, Abstract Review Committee, faculty, delegates, oral abstract presenters, and the presenters of outstanding posters that will be on display at the Opening Reception and throughout the conference, for their contributions to this year’s program. We hope that this program will foster novel insights that will push the science and practice of HIV prevention and treatment adherence to the next level. In addition to the long-standing traditions of the conference, we note several new features to this year’s scientific program, including:

• Poster Session on the opening day in conjunction with the Opening Reception; • Recognition of the highest scored posters with a “palm tree” designation; and • Late-Breaker Oral Abstract sessions and late-breaker poster presentations.

We are thankful for the commercial support provided by AbbVie, Gilead Sciences, Merck & Co., and ViiV Healthcare. We are also thankful to José M. Zuniga, PhD, MPH, and Christopher M. Gordon, PhD, under whose leadership as Co-Chairs over the past four years this conference has become an exemplar of collaboration and excellence. We appreciate the traditions they have established and look for- ward to continuing to promote innovation, discourse, and discovery. We are equally indebted to the amazing team at the International Association of Providers of AIDS Care (IAPAC), including Dr. Zuniga and Angela Knudson, who have been instrumental in developing this year’s excellent program and managing countless logistics for us to all be able to be here.

With over 400 delegates from 28 countries, this year’s conference will offer a truly remarkable opportunity to learn from each other, find the synergies, and discover and innovate to advance the science and practice of HIV prevention and treatment adherence. As adherence drives efficacy, disseminating the innovative scientific discoveries and practical lessons learned and shared over the next three days is vital to achieving the individual and population health benefits afforded by contemporary HIV prevention and treatment modalities.

Welcome to Miami - we hope you enjoy the conference!

K. Rivet Amico, PhD1 Michael J. Mugavero, MD, MHSc2 Co-Chair Co-Chair

1 University of Connecticut, Storrs, CT, USA 2 University of Alabama at Birmingham, Birmingham, AL, USA 9th InternationalConference on Continuing Education General Information Acknowledgements Poster Abstracts Poster Map Oral Abstracts Oral AbstractsSchedule Conference Program-T Conference Program-Monday, June9,2014 Conference Program-Sunday, June8,2014 Pre-Conference Symposia Faculty Roster Hotel Maps Networking Opportunities ...... HIV ...... T ...... reatment andPrevention Adherence ...... uesday ...... , June10,2014 ...... 131 56 55 18 13 12 11 10 9 8 7 6 3 2 TABLE OF CONTENTS 1 2 GENERAL INFORMATION If youhaveanyquestionspost-conference, [email protected]. If youhaveanyquestionsduringthe conference,pleaselocateanIAPAC staffmemberattheconference’s RegistrationDesk. QUESTIONS 2014 eventforliveupdatesonthe conference.Jointheconference’s socialmediaconversation: #Adherence2014 If youhaven’t already, “like”ourFacebookpage(InternationalAssociationofProviders ofAIDSCare)andjointheAdherence SOCIAL MEDIA Tuesday, June10,2014,duringtheLunchPanel discussions. Lunch willonlybeprovidedonMonday, June9,2014,and MEALS and pages10-12,respectively). (see theHotelMapsandProgramScheduleonpage7 well asthePosterSessionwillbeheldonsecondlevel Hotel. Plenary, OralAbstract,andBreakoutSessions,as Adherence 2014isbeingheldattheLoewsMiamiBeach MEETING VENUE GENERAL INFORMATION cians, researchers,andpublichealthspecialists. requires multidisciplinarycooperationamongpatients,clini at individual,community ment andbiomedicalpreventionadherenceisacriticalgoal comes. Therefore,understandingandenhancingHIVtreat- success ofanymedicationadvanceanditshealthout- transmission ratesamongat-riskpopulations. other biomedicalinterventionsholdtodramaticallycurbHIV ing thepromisethatpre-exposureprophylaxis(PrEP)and ral-based preventioninterventionswillbecriticaltoachiev resistant strainsofHIV. Similarly, adherencetoantiretrovi- tion, poorersurvivalrates,andvirusmutationtotreatment- adherence canleadtotheresumptionofrapidviralreplica- and otherrequirementsarecloselyfollowed.Partialorpoor when adherencetoprecisedosingschedulesisrigorous clinical outcomes.However suppression ofHIVreplicationisassociatedwithimproved the bloodofinfectedindividuals.Significantandsustained viral loadtoapointwhereparticlesareundetectablein Antiretroviral therapycaninhibitviralreplicationandreduce STATEMENT OFNEED sionals workinginthefieldofHIVmedicine. organization (ASO)staff,andalliedhealthcarelayprofes- demiologists, socialworkers,casemanagers,AIDSservice psychologists, behavioralresearchers,socialscientists,epi includes physicians,nurses/nurse-practitioners,pharmacists, Treatment andPreventionAdherence(Adherence2014) The targetaudienceforthe9thInternationalConferenceonHIV TARGET AUDIENCE CONFERENCE INFORMATION Behavioral andclinicalinterventionsareintegraltothe , andpublichealthlevels,which , thesebenefitsareonlytenable 9th InternationalConference on - - - After completingthisactivity, participantswillbeableto: EDUCATIONAL OBJECTIVES ety ofdomesticandinternationalsettings. HIV treatmentandbiomedicalpreventioninterventionsinavari- tifically soundandpracticalstrategiestoenhanceadherence healthcare andhumanserviceprofessionalstoexaminescien- dence-based approaches.The2.5-dayprogramwillallow research willbepresented,discussed,andtranslatedintoevi- ence forHIVtreatmentandbiomedicalpreventionadherence Adherence 2014willprovideaforumwherethestate-of-the-sci- PROGRAM OVERVIEW available atwww.iapac.org post-conference. distributed atregistration,andelectronicversionswillbe post-conference. TheProgramandAbstractsbookwillbe Slide presentationswillbeavailableatwww.iapac.org SLIDE PRESENTATIONS/ABSTRACTS day ipants. WirelessInternetcanbepurchasedfor$14.95per Adherence 2014doesnotprovideInternetaccesstopartic- INTERNET ACCESS • • SummarizeHIVtreatmentandbiomedicalprevention • Integrateintoclinicalpracticeevidence-basedinter- • medical prevention tools toenhanceadherenceHIVtreatmentandbio- Utilize patientengagementandadherenceassessment ment incare ventions toimprovelinkageandlong-termengage achieve long-termvirologicsuppression adherence inordertomitigatedrugresistanceand Apply evidence-basedstrategiestoenhancetreatment community, andpublichealth adherence state-of-the-scienceinrelationtoindividual, , andiscomplimentaryinthehotellobby. HIV T reatment andPrevention Adherence - surate withtheextentoftheirparticipationinactivity. the conference’ out insertedinyourProgramandAbstractsbook,pleasevisit the content of this CME activity. hand- this find not do you If es/life partners,havewithcommercialinterestsrelatedto or devicesfaculty, planners,andmanagers,ortheirspous- reports offinancialrelationshipsortoproducts the ProgramandAbstractsbook.Thehandoutreflects business interestofacommercialinterest. ments orqualityinhealthcareandnotaspecificproprietary CME activitiesandrelatedmaterialsthatpromoteimprove- PIM iscommittedtoprovidingitslearnerswithhighquality are thoroughlyvettedandresolvedaccordingtoPIMpolicy. have asrelatedtothecontentofthisactivity. AllidentifiedCOI close anyrealorapparentconflictofinterest(COI)theymay are inapositiontocontrolthecontentofthisactivitydis instructors, planners,managers,andotherindividualswho The PostgraduateInstituteforMedicine(PIM)requires Disclosure ofConflictsInterest activity foramaximumof19.25 The PostgraduateInstituteforMedicinedesignatesthislive Credit Designation uing medicaleducationforphysicians. for MedicineisaccreditedbytheACCMEtoprovidecontin- Providers ofAIDSCare(IAPAC). ThePostgraduateInstitute Institute forMedicineandtheInternationalAssociationof (ACCME) throughthejointprovidershipofPostgraduate the AccreditationCouncilforContinuingMedicalEducation dance withtheaccreditationrequirementsandpoliciesof This activityhasbeenplannedandimplementedinaccor Accreditation Statement CONTINUING MEDICALEDUCATION PHYSICIANS &NURSE-PRACTITIONERS/NURSES Credit(s)™ 9th InternationalConference on hours. Certificateswillbeavailable atwww Accreditation. ANACdesignates this educationalactivityforamaximumof20CNEcontact nursing educationbytheAmericanNursesCredentialingCenter’s Commissionon The AssociationofNursesinAIDSCare(ANAC)isaccreditedas aproviderofcontinuing CONTINUING EDUCATION IN NURSING A DisclosureofConflictsInteresthandoutisinsertedin . Physicians shouldclaimonlythecreditcommen s RegistrationDesk. HIV T reatment andPrevention Adherence AMA PRACategory1 .cmeuniversity - - - .com, ProjectIDnumber9762. tions, contraindications,andwarnings. mation foreachproductdiscussionofapprovedindica of theplanners.Pleaserefertoofficialprescribinginfor- those ofthefacultyanddonotnecessarilyrepresentviews agent outsideofthelabeledindications. planners ofthisactivitydonotrecommendtheuseany cated bytheUSFoodandDrugAdministration(FDA).The lished and/orinvestigationalusesofagentsthatarenotindi- This educationalactivitymaycontaindiscussionofpub- Disclosure ofUnlabeledUse cate willbemadeavailabletoeachparticipant. Upon successfullycompletingtheevaluation,aCMEcertifi- “Find Post-Tests byCourse”andsearchProjectID9762. www.cmeuniversity.com. Onthenavigationmenu, clickon Participants maycompleteanonlineevaluationat Evaluation and comparisonwithrecommendationsofotherauthorities. review ofanyapplicablemanufacturer’ conditions andpossiblecontraindicationsondangersinuse, not beusedbyclinicianswithoutevaluationoftheirpatient’s sis ortreatmentdiscussedsuggestedinthisactivityshould ment. Anyprocedures,medications,orothercoursesofdiagno- activity isnotmeanttoserveasaguidelineforpatientmanage- own professionaldevelopment.Theinformationpresentedinthis acquired informationtoenhancepatientoutcomesandtheir Participants haveanimpliedresponsibilitytousethenewly Disclaimer The opinionsexpressedintheeducationalactivityare s productinformation, - CONTINUING EDUCATION 3 4 CONTINUING EDUCATION Learning Objectives: Opportunities Lunch Panel:EarlyARTInitiation:Caveats,Challenges,and 12:05 P.M. -1:05P.M. •Americana3 Learning Objectives: Continuum ofCareContext Panel 2:OptimizingOurResponse:PlacingAdherencewithinthe 9:30 A.M.-10:30•Americana3 Learning Objectives: Keynote Address:Treating Stigma:WhenaPillIsn’t Enough 9:00 A.M.-9:30•Americana3 MONDA Learning Objectives: The ChangingLandscapeofHIV/AIDSResearch:What’s Next? Gary S.Reiter, MD,andAndrewKaplan,MemorialLecture: 4:15 P.M. -4:45P.M. •Americana3 Learning Objectives: US: TheDevelopmentandImplementationofFiveResources Pre-Conference Symposium2:IncreasingPrEPAwareness inthe 1:00 P.M. -3:00P.M. •Poinciana1/2 SUNDAY, JUNE8,2014 tration deskpriortoleavingtheconference.Certificateswillbemailedtwoweekspost-conference. are sessionsthathavebeenapprovedforcredithours. approved forupto9.0contacthours(.9CEUs).Following ACPE# 0092-9999-14-(10-19)-L04-P. Thisprogramhasbeen Education asaproviderofcontinuingpharmacyeducation. accredited bytheAccreditationCouncilforPharmacy Nova SoutheasternUniversityCollegeofPharmacyis CONTINUING EDUCATION INPHARMACY PHARMACISTS • Explainadvantagesofearlyinitiation oftherapy • • Identifyadvantagestousingamulti-disciplinaryapproach • Identifystrategiesthatpharmacistscanusetosupportadher- • • • Describetheroleofstigmaasabarriertoentryintoandreten- • Identifyresearchprioritiestomeettheaimsofpreventingnew • Describekeyscientificadvances,whichprovideanopportunity • DescribetheacceptabilityofPrEPinvariouspopulations • IdentifybarrierstoPrEPuseamongat-riskpopulations • Applyevidence-basedinterventionstopromotetheuseofPrEP • Pharmacists claimingcreditsmustcollectandreturnacompletedCertificateofParticipationtoIAPAC staffattheregis- support adherence ence withinthecontextofcontinuumcare across thecontinuumofcare Describe theroleofpharmacistsinsupportingadherence antiretrovirals Describe thebarriersforeradicatingHIVepidemicusing tion inHIVcareandtreatment HIV infectionsandendingAIDS to endtheHIVpandemic for biomedicalHIVprevention Identify challengestoearlyinitiation ofART Define “early”ARTinitiation Y , JUNE9,2014 9th InternationalConference on Learning Objectives: Innovations Plenary 1:AdherenceMonitoring-StateoftheScienceandFuture 4:00 P.M. -4:30P.M. •Americana3 MONDAY, JUNE9,2014 Learning Objectives: Accountability Challenges Acceptability, Adherence,Affordability, Accessibility, and Closing Panel:SlayingtheHydra:ExaminingUniversalART’s 4:00 P.M. -5:00P.M. •Americana3 Learning Objectives: Relationship” istheIntervention Lunch Panel:Patient-ProviderTeams: SomeTimes the“Care 12:15 P Learning Objectives: Barriers toScaleUp Panel 3:PrEP-StateoftheScience,UptakeChallenges,andOther 9:00 A.M.-10:00•Americana3 TUESDAY, JUNE10,2014 Learning Objectives: Investments inIT-Based Technologies Plenary 2:DigitalAdherenceInterventions-AReviewofCDC 4:30 P.M. -5:00P.M. •Americana3 Describeadvantagesfornewermethodsofadherencemonitoring • Identifypotential rolesforpharmacistsinpromotinguniversalART • Listchallengesassociatedwithimplementation ofuniversalART • Listtwoinnovativemethodsformonitoringadherence • Explainthebenefitofwidespread rolloutofuniversalART • • Applyprinciplesofrelationshipbuildingtopharmacypractice • • ListconsiderationsforimplementingPrePinhealthcenters • • DescribetheacceptabilityofPrePinvariouspopulations • Identifycommonfeaturesofsuccessfuldigitaladherence • establishing atherapeuticalliance Identify pharmacypracticeswhichmaybedetrimentalto tionship” Identify successfulelementsofbuildingapositive“carerela tion projects List factorsassociatedwithadherencetoPrePindemonstra- interventions .M. -1:15P .M. •Americana3 (continued) HIV T reatment andPrevention Adherence - PSYCHOLOGISTS

CONTINUING EDUCATION IN PSYCHOLOGY Postgraduate Institute for Medicine (PIM) is approved by the American Psychological Association to sponsor continuing education for psychologists. PIM maintains responsibility for this program and its content. This program offers 18.50 continuing education credits for psychologists. The following sessions have been approved for continuing education credits for psychologists. Participants must verify their attendance by signing in during each session. Sign-in sheets will be placed in the back of every room.

Participants may complete an online evaluation at www.cmeuniversity.com. On the navigation menu, click on “Find Post- TION Tests by Course” and search by Project ID 9762. Upon successfully completing the evaluation, a CE certificate will be made available to each participant.

SUNDAY, JUNE 8, 2014 MONDAY, JUNE 9, 2014 (continued) 10:00 A.M. - 12:00 P.M. • Poinciana 1/2 5:00 P.M. - 6:00 P.M. • Americana 3 Pre-Conference Symposium 1: Networking: Adherence Update Plenary 3: Three Top-Rated Oral Abstracts from the International HIV Networks

1:00 P.M. - 3:00 P.M. • Poinciana 1/2 TUESDAY, JUNE 10, 2014 Pre-Conference Symposium 2: Increasing PrEP Awareness in the 9:00 A.M. - 10:00 A.M. • Americana 3 US: The Development and Implementation of Five Resources Panel 3: PrEP - State of the Science, Uptake Challenges, and Other Barriers to Scale Up 4:15 P.M. - 4:45 P.M. • Americana 3 Gary S. Reiter, MD, and Andrew Kaplan, MD, Memorial Lecture: 10:15 A.M. - 11:30 A.M. • Americana 3 The Changing Landscape of HIV/AIDS Research: What’s Next? Thematic Oral Abstract Session 9: Engagement in Care - Interventions and Correlates 4:45 P.M. - 5:45 P.M. • Americana 3 Panel 1: Patient Perspectives on Treatment as Prevention 10:15 A.M. - 11:30 A.M. • Salon 1 Thematic Oral Abstract Session 10: Adherence - Technology

MONDAY, JUNE 9, 2014 11:35 A.M. - 12:05 P.M. • Americana 3 9:00 A.M. - 9:30 A.M. • Americana 3 Invited Speaker Session 1: Patient-Provider Communication Keynote Address: Treating Stigma: When a Pill Isn’t Enough CONTINUING EDUCA 11:35 A.M. - 12:05 P.M. • Salon 1 9:30 A.M. - 10:30 A.M. • Americana 3 Invited Speaker Session 2: Stigma Panel 2: Optimizing Our Response: Placing Adherence within the Continuum of Care Context 11:35 A.M. - 12:05 P.M. • Salon 2 Invited Speaker Session 3: Health Literacy 10:45 A.M. - Noon • Salon 1 Thematic Oral Abstract Session 2: Adherence - Measurement 12:15 P.M. - 1:15 P.M. • Americana 3 and Epidemiology Lunch Panel: Patient-Provider Teams: Some Times the “Care Relationship” is the Intervention 1:30 P.M. - 2:45 P.M. • Salon 2 Thematic Oral Abstract Session 7: Adherence - Intervention 1:45 P.M. - 3:00 P.M. • Americana 3 Thematic Oral Abstract Session 13: Adherence 2:50 P.M. - 3:50 P.M. • Salon 1 Late-Breaker Oral Abstract Session 2: ART Adherence: Toolkits 1:45 P.M. - 3:00 P.M. • Salon 2 and Technology Thematic Oral Abstract Session 15: Linkage and Retention - Correlates of Linkage and Retention 2:50 P.M. - 3:50 P.M. • Salon 2 Late-Breaker Oral Abstract Session 3: ART Adherence: Correlates 1:45 P.M. - 3:00 P.M. • Poinciana 1/2 and Consequences Thematic Oral Abstract Session 16: Biobehavioral Prevention - Perspectives: Consumers, Prescribers, and Communities 4:00 P.M. - 4:30 P.M. • Americana 3 Plenary 1: Adherence Monitoring - State of the Science and Future 4:00 P.M. - 5:00 P.M. • Americana 3 Innovations Closing Panel: Slaying the Hydra: Examining Universal ART’s Acceptability, Adherence, Affordability, Accessibility, and 4:30 P.M. - 5:00 P.M. • Americana 3 Accountability Challenges Plenary 2: Digital Adherence Interventions - A Review of CDC Investments in IT-Based Technologies

9th International Conference on HIV Treatment and Prevention Adherence 5 6 NETWORKING OPPORTUNITIES visit theAdherence2014registration desktosign-uptoday! to Adherence2014attendeesasawayhelpyoumakethemostofyour conferenceexperienceandconnectwithyourcolleagues.Please Networking dinnerattendeeswillberesponsibletopayfortheirownmeal, drinks,andassociatedtip.Thenetworkingdinnersareoffered vation isfilledonafirst-come,first-servedbasistoofferautomaticdinner plans,networking,andgreatfoodallinone. Reservations willbeofferedfortablestoseatsix-10peopledinnerMonday, June9,2014,between6:30P.M. and8:30P.M. Thereser- tunity toparticipateinacasualdinnerwithfellowdelegatesandguests whowanttoenjoyanicemealandinterestingconversation. delegates foranetworkingdinner. Whetheryou’re travelingalone,afirst-timeattendee,orjustlookingforgreatmeal,thisisoppor- Maximize yournetworkingopportunitiesandenjoyamouthwateringmeal atoneofMiami’s toprestaurantsbyjoiningotherAdherence2014 Monday, June9,2014 NETWORKING DINNERS program. Participantsshouldstayonsidewalks,usecrosswalksandalerttotrafficpedestrians. caution whenperformingexerciseand/orengaginginstrenuousphysicalactivity in thisevent.ParticipantsalsofullyunderstandtherigorsofactivityandhavepreparedthemselvesphysicallyforPleaseuse sponsors, andanyotherparticipatinggroupswithrespecttopersonalloss,illness,bodilyinjuryordeathresultingfrom Disclaimer: Runnersacknowledgethattheyareparticipatingattheirownriskandwaiveallclaimsofeverynatureagainsttheorganizers, to getknoweachotherinarelaxed,informalenvironment.Seeyouthere! of allendurance,paces,distancesandexperiencelevelsareencouragedtocomeruntogetherinanon-competitiveatmosphereasway Calling allrunners!Comemeetuseverymorningforapre-conferencerunaroundMiamiBeach.Adherence2014participantsandrunners Meet UpPoint:HotelLobby RUNNING CLUB The IssueInterestCoffeeT Men whohaveSexwithMen,Women, Transgender Individuals,Children/Adolescents,MentalHealth Issue InterestCoffeeTalk, 8:00A.M.-8:50A.M.,Tuesday, June10,2014 The GroupInterestCoffeeTalk willfocusontheindividualswhoservepeoplelivingwithHIV/AIDS:serviceproviders. Pharmacists, Nurses&NursePractitioners,CivilSociety, Psychologists,Physicians Group InterestCoffeeTalk, 8:00A.M.-8:50A.M.,Monday, June9,2014 tunity willprovideaspaceforyoutodiscussissuesspecificHIV/AIDSserviceprovidersandspecialpopulations. IAPAC invitesyoutostopbyaCoffeeTalk tomixand mingleamongcolleaguesandnewfriends.Thisuniquecasualnetworkingoppor- Adherence 2014coversavastspectrumofissues.Comebreakitdownwithusovercoffeeandanearlybirdbreakfast. COFFEE TALKS NETWORKING OPPORTUNITIES alk willfocusonspecialpopulationsofindividualslivingwithoraffectedbyHIV/AIDS. , 7:00A.M.(Monday, June9,2014&Tuesday, June10,2014) 9th InternationalConference on . Pleaseconsultyourphysicianbeforebeginningexercise HIV T reatment andPrevention Adherence LOEWS MIAMI BEACH HOTEL

Poinciana 4 Salon 1 Americana e

Ballroom c y a r r e r l e l T a r s e a G w d r k n e e r n a a y a o n Poinciana 3 N c o S i Y a r F c i e r a e n m A a A

i Banquet m c A

n Office i

o Salon 2 Salon 3 Salon 4 P Poinciana 2

Es ca la to HOTEL MAPS rs Poinciana 1 E Women’s le va Restroom to rs

Americana Foyer

Rotunda Men’s Business Restroom Center

Second Level

9th International Conference on HIV Treatment and Prevention Adherence 7 K. Rivet Amico, PhD Linda J. Koenig, PhD, MS Kenly Sikwese University of Connecticut Centers for Disease Control and Prevention African Community Advisory Board Storrs, CT, USA Atlanta, GA, USA Lusaka, Zambia

Michele Peake Andrasik, PhD Deborah Konkle-Parker, PhD, FNP Jane M. Simoni, PhD Fred Hutchinson Cancer Research Center University of Mississippi Medical Center University of Washington Seattle, WA, USA Jackson, MS, USA Seattle, WA, USA

Mary Catherine Beach, MD, MPH Ann Kurth, PhD, CNM, FAAN Papa Salif Sow, MD, MS Johns Hopkins University NYU College of Nursing Bill & Melinda Gates Foundation Baltimore, MD, USA New York, NY, USA Seattle, WA, USA

Patrick Buzzell Kenneth Mayer, MD Christian Stanley Boston, MA, USA Harvard University Boston, MA, USA TY ROSTER Boston, MA, USA Megan Canon, MPH Sean Strub San Francisco AIDS Foundation Alan McCord The Sero Project San Francisco, CA, USA Project Inform Milford, PA, USA San Francisco, CA, USA Amanda Castel, MD, MPH Betsy Tolley, PhD, MA George Washington University Maria T. Mejía FHI 360 ACUL Washington, DC, USA Miami, FL, USA Durham, NC, USA F Jennifer Cocohoba, PharmD, MAS Julio S.G. Montaner, MD Ariane van der Straten, PhD, MPH University of California British Columbia Centre for Excellence in RTI International San Francisco, CA, USA HIV/AIDS Washington, DC, USA Vancouver, BC, Canada Dazon Dixon Diallo, MPH Mitchell Warren Sister Love Michael J. Mugavero, MD, MHSc AIDS Vaccine Advocacy Coalition Atlanta, GA, USA University of Alabama at Birmingham New York, NY, USA Birmingham, AL, USA Ruth DeRamus Sheri Weiser, MD, MA, MPH UAB 1917 Clinic Julie Myers, MD University of California Birmingham, AL, USA NYC Department of Health and Mental San Francisco, CA, USA Hygiene Thomas P. Giordano, MD, MPH New York, NY, USA Alan Whiteside, DEcon Baylor College of Medicine Wilfrid Laurier University Houston, TX, USA Rob Newells Waterloo, ON, Canada AVAC PxROAR Christopher M. Gordon, PhD Oakland, CA, USA Ira B. Wilson, MD National Institute of Mental Health Brown University Rockville, MD, USA Jim Pickett Providence, RI, USA AIDS Foundation of Chicago Robert M. Grant, MD, PhD Chicago, IL, USA Phill Wilson University of California Black AIDS Institute San Francisco, CA, USA Robert H. Remien, PhD Los Angeles, CA, USA Columbia University Kathleen Green, PhD New York, NY, USA Theresa Wolters, MA Centers for Disease Control and Prevention Elizabeth Glaser Pediatric AIDS Foundation Atlanta, GA, USA Steven A. Safren, PhD, ABPP Washington, DC, USA Harvard Medical School Jessica E. Haberer, MD, MS Boston, MA, USA Anna Zakowicz, MPH, MA Massachusetts General Hospital Global Network of People Living with HIV Boston, MA, USA Jeffrey T. Schouten, MD, JD Amsterdam, Netherlands Fred Hutchinson Cancer Research Center Sybil G. Hosek, PhD Seattle, WA, USA José M. Zuniga, PhD, MPH John H. Stroger, Jr. Hospital of Cook County International Association of Providers of Chicago, IL, USA James D. Scott, PharmD, MEd AIDS Care Western University Washington, DC, USA Pomona, CA, USA

8 9th International Conference on HIV Treatment and Prevention Adherence 9th InternationalConference on CME accredited Jim Pickett Alan McCord Sybil Hosek,PhD Dazon DixonDiallo,MPH Megan Canon,MPH Panelists: K. RivetAmico,PhD Moderator: Increasing PrEPAwareness intheUS:TheDevelopmentandImplementationof FiveResources 1:00 Symposium 2 Sponsored by: CME accredited and SybilHosek,PhD,StrogerHospitalofCookCounty, Chicago,IL,USA Jeffrey T. Schouten,MD,JD,presentingonbehalfofCarlosDelRio,EmoryUniversitySchool,Atlanta,GA,USA, Betsy Tolley, PhD,MA Ariane vanderStraten,PhD,MPH Steven A.Safren,PhD,ABPP Michele PeakeAndrasik,PhD Panelists: Jeffrey T. Schouten,MD,JD Moderator: Networking: AdherenceUpdatefromtheInternationalHIVNetworks 10:00 Symposium 1 SUNDAY, JUNE8,2014 PRE-CONFERENCE SYMPOSIA P . A M . M . -3:00 . -Noon/Poinciana1/2 , AIDSFoundationChicago, IL,USA , ProjectInform,SanFrancisco,CA, USA P . M . /Poinciana1/2 , StrogerHospitalofCookCounty, Chicago,IL,USA , UniversityofConnecticut,Storrs,CT, USA , SanFranciscoAIDSFoundation,Francisco,CA,USA , FHI360,Durham,NC,USA HIV , SisterLove,Atlanta,GA,USA T , FredHutchinsonCancerResearchCenter, Seattle,WA, USA reatment andPrevention Adherence , HarvardMedicalSchool,Boston,MA,USA , FredHutchinsonCancerResearchCenter, Seattle,WA, USA , RTIInternational,Washington, DC,USA

9 PRE-CONFERENCE SYMPOSIA 10 PROGRAM SCHEDULE IEACTIVITY TIME SUNDAY, JUNE8,2014 N 4:45 4:15 3:00 6:00 5:45 4:00 POINCIANA1/2 ACTIVITY 10:00 TIME 1:00 OON P P P P P P P –1:00 ...... A M M M M M M M . M .–8:00 .–6:00 .–5:45 .–4:45 .–4:15 .–4:00 .–3:00 .–Noon P . M P P P P P P P Break ...... M M M M M M M ae :PatientPerspectivesonTreatment asPrevention Panel1: . GaryS.Reiter, MD,andAndrewKaplan,MemorialLecture . ConferenceWelcome . Break . Pre-Conference . ¡BienvenidosaMiami! . Break . Symposium 1 Pre-Conference Symposium 2 Moderator: Christopher M.Gordon,PhD Michael J.Mugavero,MD,MHSc K. RivetAmico,PhD José M.Zuniga,PhD,MPH Panelists: Christian Stanley;AnnaZakowicz,MPH,MA OPENING RECEPTION&POSTERSESSION•AMERICANA4 Panelists: Moderator: Increasing PrEPAwareness intheUS:TheDevelopmentandImplementationofFiveResources Jeffrey T. Schouten,MD,JD,presentingonbehalfofCarlosDelRio,andSybilHosek,PhD Panelists: Moderator: Networking: AdherenceUpdatefromtheInternationalHIVNetworks Jim Pickett Alan McCord Sybil Hosek,PhD Dazon DixonDiallo,MPH Ariane vanderStraten,PhD,MPH Betsy Tolley, PhD,MA Steven A.Safren,PhD,ABPP Patrick Buzzell;MariaT. Mejía;RobNewells;KenlySikwese; Phill Wilson Megan Canon,MPH Michele PeakeAndrasik,PhD OPENING SESSION•AMERICANA3 K. RivetAmico,PhD Jeffrey T. Schouten,MD,JD 9th InternationalConference on HIV T reatment andPrevention Adherence 9th InternationalConference on MONDAY, JUNE9,2014 8:50 Noon–12:05 3:50 2:45 IEATVT MRCN AO SALON2 SALON1 POINCIANA1/2 5:00 AMERICANA3 SALON2 4:30 4:00 SALON1 ACTIVITY AMERICANA3 POINCIANA1/2 2:50 TIME ACTIVITY SALON2 1:30 TIME SALON1 1:05 AMERICANA3 12:05 ACTIVITY 10:45 TIME 10:30 9:30 9:00 ACTIVITY 8:00 TIME P P P P P P A A A A P P ...... A A P . . . . M M M M M M M M M M M M . . . M M M .–6:00 .–5:00 .–4:30 .–4:00 .–3:50 .–2:50 .–2:45 .–1:30 .–10:30 .–9:30 .–9:00 .–8:50 .–1:05 .–Noon .–10:45 P . M P P P P P P P P A A A ...... P M M M M M M M M Break . A . . . . M M M M A . Plenary3:ThreeTop-Rated OralAbstracts . Plenary2:DigitalAdherenceInterventions -AReviewofCDCInvestmentsinIT-Based Technologies . Plenary1:AdherenceMonitoring-State oftheScienceandFutureInnovations . Break . . Break . . Break . M Treating Stigma:WhenaPillIsn’t Enough . Break . . . . M Panel2:OptimizingOurResponse:PlacingAdherencewithintheContinuumofCareContext . Break . Early ARTInitiation:Caveats,Challenges,andOpportunities Sessions Oral Abstract Thematic Sessions Oral Abstract Late-Breaker Sessions Oral Abstract Thematic HIV Pharmacists Moderator: Kathleen Green,PhDandLindaJ.Koenig, PhD,MS Jessica E.Haberer, MD,MS Panelists: Discussants: Presenter: Sean Strub Moderator: Moderator: T reatment andPrevention Adherence GROUP INTERESTNETWORKINGBREAKFAST •SALON1/2 Amanda Castel,MD,MPH;KenlySikwese Julio S.G.Montaner, MD José M.Zuniga,PhD,MPH Mitchell Warren Papa SalifSow, MD,MS Ann Kurth,PhD,CNM,FAAN; JamesScott,PharmD,Med;JaneM.Simoni,PhD Epidemiology Measurement and Linkage andRetention- Session 1: and Methods Prevention -Correlates Biobehavioral Session 5: Nurse-Practitioners KEYNOTE ADDRESS•AMERICANA3 the CareContinuum Systems ApproachesAcross Structural andHealth Session 1: Nurses/ LUNCH PANEL •AMERICANA3 PLENARY •AMERICANA 3 PANEL •AMERICANA3 Epidemiology Measurement and Adherence - Session 2: Epidemiology Measurement and Linkage andRetention- Session 6: ii oit scooit Physicians Psychologists Civil Society and Technology ART Adherence:Toolkits Session 2: Epidemiology Measurement and Engagement inCare- Session 3: Intervention Adherence - Session 7: and Consequences ART Adherence:Correlates Session 3: Epidemiology Measurement and Prevention - Biobehavioral Session 4: and Correlates Continua ofCare Engagement inCare- Session 8:

11 PROGRAM SCHEDULE 12 PROGRAM SCHEDULE TUESDAY, JUNE10,2014 11:30 8:50 12:05 3:00 3:45 IEATVT MRCN AO SALON2 SALON1 5:00 AMERICANA3 4:00 ACTIVITY SALON2 3:15 TIME SALON1 POINCIANA1/2 1:45 TIME 1:15 AMERICANA3 SALON2 12:15 SALON1 ACTIVITY AMERICANA3 11:35 TIME ACTIVITY 10:15 TIME 10:00 9:00 ACTIVITY 8:00 TIME P P P P P A A A P P ...... P A A A A P . . . M M M M M M M M M M ...... M M M M M M .–4:00 .–3:45 .–3:15 .–3:00 .–1:45 Adjourn . .–5:00 .–10:00 .–9:00 .–8:50 .–1:15 .–12:15 . .–11:35 .–11:30 .–10:15 –12:05 P P P P P P A A ...... P M M M M M M A . . . M M P M P A A A ...... Break . . M . M Break . . . . . M Patient-Provider Teams: SomeTimes the“CareRelationship”is . M M M Panel3:PrEP-StateoftheScience,UptakeChallenges,andOtherBarrierstoScaleUp . Break . . Break . . Break . and AccountabilityChallenges Break Slaying theHydra:ExaminingUniversal ART’ Break ACTIVITY Sessions Oral Abstract Thematic Invited Speakers Sessions Oral Abstract Thematic Invited Speakers Men whohave Sex withMen Moderators: Deborah Konkle-Parker Panelists: Panelists: Moderator: Discussants: Presenter: GROUP INTERESTNETWORKINGBREAKFAST •SALON1/2 Jennifer Cocohoba,PharmD,MAS;RuthDeRamus;ThomasP Christopher M.Gordon,PhD;AlanWhiteside, DEcon Kenneth Mayer, MD Ira B.Wilson,MD K. RivetAmico,PhDandMichaelJ. Mugavero,MD,MHSc Patrick Buzzell;RobertM.Grant,MD,PhD;SybilG.Hosek,JimPickett Correlates Interventions and Engagement inCare- Session 9: Adherence Session 13: AMERICANA 3 , PhD,FNP;RobertH.Remien,PhD Julie Myers,MD Public HealthApproaches Session 4: Mary CatherineBeach,MD,MPH Patient-Provider Communication Session 1: CLOSING P Women LUNCH P PANEL •AMERICANA3 ANEL •AMERICANA3 ANEL •AMERICANA3 9th InternationalConference on s Acceptability Technology Adherence - Session 10: Interventions Engagement inCare- Session 14: SALON 1 Transgender Individuals Sheri W Food Insecurity Session 5: Sean Strub Stigma Session 2: , Adherence,Affordability the Intervention eiser , MD,MA,MPH Interventions Linkage andRetention- Session 11: and Retention Correlates ofLinkage Linkage andRetention- Session 15: SALON 2 . Giordano,MD,MPH; HIV hlrnAoecnsMentalHealth Children/Adolescents T reatment andPrevention Adherence , Accessibility Theresa W Health SystemsinTransition Session 6: Anna Zakowicz,MPH,MA Health Literacy Session 3: POINCIANA 1/2 Demonstration Projects - Interventionsand Biobehavioral Prevention Session 12: and Communities Consumers, - Perspectives: Biobehavioral Prevention Session 16: olters, MA , Prescribers,

MONDAY, JUNE 9, 2014

THEMATIC ORAL ABSTRACT SESSIONS

SESSION 1 SESSION 3 Linkage and Retention - Measurement and Engagement in Care - Measurement and Epidemiology Epidemiology 10:45 A.M. - Noon / Americana 3 10:45 A.M. - Noon / Salon 2 Moderator: Michael J. Mugavero, MD, MHSc Moderator: Alan Whiteside, DEcon

287 Patterns of HIV Service Use and HIV Viral Suppression 330 2014 Federal Recommendations about HIV Prevention among Patients Treated in an Academic Infectious Diseases Services for Persons with HIV: Promoting Synergies between Clinic in North Carolina Clinicians, Community-Based Organizations, and Health Anton Palma presenting Departments Amrita Tailor presenting 427 Antiretroviral Refill Adherence is an Early Predictor of Retention in HIV Care 436 Examining Clinic-Based and Public Health Approaches to Robert Gross presenting Ascertainment of HIV Care Status Katerina Christopoulos presenting 444 Public Health Campaigns as a Vehicle for Quality Improvement in HIV Services 441 in+care Campaign: Improving Retention in HIV Care Michael Hager presenting Michael Hager presenting

448 Comparison of Engagement in Care Measures Using Self- SESSION 2 Report and Clinical Records Data Adherence - Measurement and Epidemiology Amanda Castel presenting ABSTRACTS SCHEDULE 10:45 A.M. - Noon / Salon 1 Moderator: Jane M. Simoni, PhD SESSION 4 315 Changing Trend in Adherence to Highly Active Antiretroviral Biobehavioral Prevention - Measurement and Therapy in the Multicenter AIDS Cohort Study and the AIDS Epidemiology Linked to Intravenous Experience 10:45 A.M. - Noon / Poinciana 1/2 Shilpa Viswanathan presenting Moderator: Yvette Calderon, MD, MS ORAL 345 Development of an Evidence-Based Alert for Risk of ART 275 DNA and Protein Biomarkers Offer Increased Accuracy for Failure within an Electronic Medical Record in Haiti Assessing Vaginal Microbicide Gel Adherence Nancy Puttkammer presenting Andrea Thurman presenting

414 Comparing Adherence Items of Missed Doses with Different 297 What Do We Really Mean By ‘Adherence’ in Vaginal Timeframes and their Associations with Viral Load in Routine Microbicide Trials? A Comparative Study Clinical Care Lori Miller presenting Heidi Crane presenting 366 FEM-PrEP: Participant Explanations for Study Product 443 Using the Visual Analogue Scale to Measure Adherence to Adherence Antiretroviral Therapy: A Meta-Analysis Amy Corneli presenting David Finitsis presenting 382 Daily Short Message Service Surveys Detect Greater HIV Risk Behavior than Monthly Clinic Questionnaires in Kenya Kathryn Curran presenting

9th International Conference on HIV Treatment and Prevention Adherence 13 14 ORAL ABSTRACTS SCHEDULE 428 408 383 303 Moderator: ZazaT 1:30 Epidemiology Linkage andRetention-Measurement 6 SESSION 440 396 374 347 Moderator: KennethMayer, MD 1:30 Methods Biobehavioral Prevention-Correlatesand 5 SESSION MONDAY, JUNE9,2014 P P . . M M Women intheSouth States Enrolled intheiPrExOpenLabelExtension(OLE)United Mirjam-Colette Kempfpresenting Postpartum RetentioninHIVCareamong Lillian Lourencopresenting Canada, from2000-2012 Time toRetentionamong PersonsLinkedtoHIVCareinBC, HIV OutpatientClinics Factors AssociatedwithIntervalsbetweenWomen’s Visits to Florence Momplaisirpresenting Outcomes inPregnantW Time ofHIVDiagnosisandEngagementinPrenatalCareImpact Hailey Gilmorepresenting T Alexander Kintupresenting Adherence Study Adherence Gaps:DatafromthePartnersPrEPAncillary Abstinence areAssociatedwithLowerAdherenceand Sexual RelationshipsOutsidePrimaryPartnershipsand Albert Liupresenting MSM andTransgender Women intheGlobaliPrExStudy Patterns andCorrelatesofPrEPDrugDetectionamong Nima Machoufpresenting Study ofPost-SexualExpositionProphylaxis Why DoPeopleComeBackforFollow-up?AProspective Fiona Burnspresenting -2:45 . -2:45 . o T ake orNottoT P P . . M M /Salon1 . /Americana3 . sereteli, MD,MPH ake PrEP:PerspectivesfromParticipants omen withHIV (CONTINUED) -Infected 9th InternationalConference on 370 301 273 Moderator: JoséM.Zuniga,PhD,MPH 1:30 Adherence -Intervention SESSION 7 402 375 337 334 Moderator: EvelynP. Tomaszewski, MSW, ACSW 1:30 Correlates Engagement inCare-Continuaofand SESSION 397 P P . . M M and QualityofLife for OlderPeopleLivingwithHIVonAntiretroviralAdherence Impact ofaSelf-ManagementTelephone SupportProgram Sebastian Linnemayrpresenting Adherence InterventioninUganda First FindingsofaNovel,V Marvin Belzerpresenting Antiretroviral Therapy Intervention forYouth LivingwithHIVNonadherentto Acceptability andFeasibilityofaCellPhoneSupport Becky Genbergpresenting of HIVCareinWestern Kenya:FromHIVTesting toRetention Population-Based EstimatesofEngagementintheContinuum Rituparna Patipresenting Center ofNewYork City Factors AssociatedwithLosstoClinicinaLargeHIVCare Kaja-Triin Laisaarpresenting Engagement intheContinuumofHIVCareEstonia Christine Oramasionwupresenting Engagement inOutpatientCareforPatientsLivingwithHIV Allen Giffordpresenting Adherence: InitialResultsoftheMedCHECRandomizedT A Tablet-Computer ClinicalInterventiontoSupportAntiretroviral Sheryl Catzpresenting -2:45 . -2:45 . 8 P P . . M M /Salon2 . /Poinciana1/2 . HIV T reatment andPrevention Adherence ariable, Rewards-Based rial Across CareContinuum Structural andHealthSystemsApproaches 1 SESSION 486 472 471 Moderator: ThomasP. Giordano,MD,MPH 2:50 ART Adherence:CorrelatesandConsequences SESSION 3 493 490 473 Moderator: BenjaminYoung, MD,PhD 2:50 ART Adherence:T SESSION 483 482 Moderator: PapaSalifSow, MD,MS 2:50 MONDAY, JUNE9,2014 9th InternationalConference on LATE-BREAKER ORALABSTRACTSESSIONS P P P . . . M M M Adherence Motivations Rhondette Jonespresenting T Every DoseDay:APeekintoCDC’s Newe-Learning Kathleen Bradypresenting the HIVCareContinuum Care, andViral Suppression:ANewMethodforMonitoring Geospatial PatternsPredictLinkagetoCare,Retentionin Matt Hickeypresenting Mfangano Island,Kenya Intervention forPromotingEngagementinHIVCareon Pulling theNetworkTogether: ANovelSocialNetwork Marcia Holstadpresenting Health LiteracyisRelatedtoHIV Shilpa Viswanathan presenting Results fromtheVeterans AgingCohort Study Adherence andHIVRNASuppressionintheCurrentEra- Parya Saberipresenting Group StudyParticipants Adherence andHIVViral LoadamongAIDSClinicalTrials Association betweenBarrierstoAntiretroviralTherapy Lora Sabinpresenting Real-T The ChinaAdherencethroughTechnology Study:TheEffectof Anik Patelpresenting Antiretroviral AdherenceinKenya Proven Text-Messaging ProgramsforImproving The Impacton5-Year MortalityandtheCost-Effectivenessof . -3:50 . -3:50 . -3:50 raining T ime FeedbackonAdherencetoAntiretroviralTherapy 2 P P P . . . M M M oolkit toImproveAdherenceAntiretroviralTreatment . /Salon2 . /Salon1 . /Americana3 oolkits andTechnology HIV T -Positive PersonsMedication reatment andPrevention Adherence (CONTINUED) 395 377 283 Moderator: JoséM.Zuniga,PhD,MPH 5:00 THREE TOP-RATED ORALABSTRACTS P . M for RapidLinkagetoHIVCare Intervention toFacilitateEmotionalandCognitiveResponses Linkage toCareDemonstrationProject:APractice-Advised Albert Liupresenting in theDemoProject Clinics: HighUptakeandDrugConcentrationsamongMSM Implementation ofPrEPinSTDandCommunityHealth Linda Beerpresenting Monitoring Project,2009 Therapy UsebyRacialandEthnicGroup-Medical Gender DisparitiesinViral SuppressionandAntiretroviral Robert Bolanpresenting -6:00 . P . M /Americana3 .

ORAL ABSTRACTS SCHEDULE 15 16 ORAL ABSTRACTS SCHEDULE Correlates Engagement inCare-Interventionsand 9 SESSION 369 323 276 266 Moderator: BenjaminYoung, MD,PhD 10:15 Adherence -T 10 SESSION 431 430 372 363 Moderator: ChristopherM.Gordon,PhD 10:15 TUESDAY, JUNE10,2014 THEMATIC ORALABSTRACTSESSIONS A A . . M M Lora Sabinpresenting with ‘Wisepill’ Positive IndividualsinSouthernChina:EarlyExperiences Real-Time Antiretroviral Treatment Monitoring amongHIV- Robert Grosspresenting Assessment ofPresentationFormats Feedback ofHIVMedicationAdherenceDatatoPatients: Keith Petriepresenting Adherence toAntiretroviralTherapy Personalized Health-RelatedVisual ImageryinImproving The EffectivenessofaNewSmartphoneAppUsing Laura Whiteleypresenting An iPhoneApp/GametoImproveARTAdherence Janet Myerspresenting Engagement inHIVCareFollowingReleasefromJail Rhodri Dierst-Daviespresenting Program atSafetyNetHIVClinicsinLosAngelesCounty Six-Month OutcomesfromaMedicalCareCoordination Stephanie Kochpresenting Engagement inHIV-Related MedicalCare A QualitativeInvestigationofPatientsTransitional Statesof Stephanie Chamberlinpresenting Care CoordinationPrograminNewYork City Robust Short-Term EffectivenessofaComprehensiveHIV . 11:30 - . - 11:30 A A . . M M . /Americana3 . echnology / Salon1 9th InternationalConference on 387 248 Moderator: CelsoRamos,MD,MSc 10:15 Linkage andRetention-Interventions SESSION 11 409 350 339 317 Moderator: K.RivetAmico,PhD 10:15 Demonstration Projects Biobehavioral Prevention-Interventionsand SESSION 456 437 A A . . M M April Monroepresenting Randomized Trial inRakai,Uganda Antiretroviral Therapy:AQualitativeEvaluationofa and CareServicesamongHIV-Infected PersonsNotYet on Impact ofaPeerInterventiononEngagementinPrevention Alan Lifsonpresenting Client AttitudesandOutcomesafter1Year Community HIVSupportWorker PrograminRuralEthiopia: Sarit Golubpresenting Health Center Demonstration/Implementation HybridinaCommunity-Based Developing andImplementingaPrEP Ryan Kofronpresenting Pre-Exposure ProphylaxisDemonstrationProject Real Time PlasmaTFV LevelstoSupportAdherenceina Ronald Brookspresenting in LosAngeles:ImplicationsforPrEPDissemination of EffectivenessamongLowSESBlackGayandBisexualMen Acceptability ofHIVPre-ExposureProphylaxisatV Jessica Habererpresenting Demonstration Project:PreliminaryFindings Adherence toPre-ExposureProphylaxisinthePartners John Lizcanopresenting Stepped Wedge Study Kenya: FindingsfromtheFirstInterventionPeriodofa Seek, T Kimberly Koesterpresenting HRSA/SPNS SystemsLinkageandAccesstoCareInitiative and RetentionInterventionsforPeopleLivingwithHIV: The Evaluating theDevelopmentandImplementationofLinkage . - 11:30 - . . - 11:30 - 12 est, T A A . . M M /Salon2 . . / Poinciana1/2 reat andRetainforPeopleWhoInjectDrugsin HIV T reatment andPrevention Adherence arying Levels 413 400 322 274 Moderator: Tia Morton,RN,MS 1:45 Engagement inCare-Interventions SESSION 379 371 351 260 Moderator: MarkR.Nelson,MBBS 1:45 Adherence 13 SESSION TUESDAY, JUNE10,2014 9th InternationalConference on P P . . M M Controlled T Care: FindingsfromaSystematicReviewofRandomized Linkage to,Retention,andRe-EngagementinHIVMedical Identification ofEvidence-BasedInterventionsforPromoting Y among MACH14Patients Medication AssociatedwithLowerAntiretroviralAdherence Symptomatic HIVDiseaseorPerceivedSideEffectsof David Moorepresenting Antiretroviral MedicationNon-Adherence Active MethamphetamineUseisAssociatedwith Patricia Garviepresenting Functioning inAdherenceforY Roles ofMedicationResponsibility, ExecutiveandAdaptive Susan Grahampresenting who haveSexwithMeninCoastalKenya Facilitators andChallengestoARTAdherenceamongMen Rebecca Dillinghampresenting Linkage toHIVCareinRuralVirginia Positive Links:SmartPhonesandSupportforSuccessful Rhodri Dierst-Daviespresenting and Re-EngagingLostHIVClinicPatients The NavigationProgram:AnInnovativeMethodforFinding Jennifer Kellerpresenting Carolina and BridgeCounselingtoImproveRetentioninCareNorth Bridging theGaps:TheUseofHealthInformationTechnology Darrel Higapresenting -3:00 . -3:00 . an Wang presenting 14 P P . . M M /Salon1 . /Americana3 . rials, 1996-2012 HIV T outh withPerinatalHIV reatment andPrevention Adherence (CONTINUED) 381 264 255 Moderator: ChinkholalThangsing,MBBS 1:45 and Retention Linkage andRetention-Correlatesof SESSION 15 416 386 329 270 Moderator: JeffreyCrowley 1:45 Consumers, Prescribers,andCommunities Biobehavioral Prevention-Perspectives: SESSION 16 389 P P . . M M Uganda Linking KeyPopulationstoHIVCareServicesinKampala, Baligh Yehia presenting with LinkagetoCare Location Matters:SiteofHIVDiagnosisStronglyAssociated Lameck Machumipresenting Treatment inDaresSalaam Follow UpamongHIV-Positive PatientsonCareand Who areThey?IdentifyingRiskFactorsofLossto Charlene Flashpresenting Willingness toPrescribeHIVPre-ExposureProphylaxis Houston PrimaryCareProviders’Perceptionsofand K. RivetAmicopresenting Development andPilotofaPrEPEducationVideo PrEP AudioV Kathleen Morrowpresenting Microbicide Development Choosing Prevention:TheCaseforInvolvingUsersinEarly Douglas Krakowerpresenting Antiretrovirals forPrevention Knowledge, andPracticesRegardingtheUseof New EnglandHealthcareProviders’Perceptions, Oni Blackstockpresenting W Correlates ofNotReceivingHIVCareamongHIV-Infected Stephen Alegepresenting -3:00 . -3:00 . omen EnrollinginaHRSASPNSMulti-SiteInitiative P P . . M M /Salon2 . /Poinciana1/2 . isual Representation(PrEPREP)Project: , MPH

ORAL ABSTRACTS SCHEDULE 17 Community HIV Support Worker Who are They? Identifying Risk 248Program in Rural Ethiopia: Client 255 Factors of Loss to Follow Up among Attitudes and Outcomes after 1 Year HIV-Positive Patients on Care and Treatment in Dar es Salaam Alan Lifson1 (presenting), Sale Workneh2, Abera Hailemichael2, Workneh Demissie2, Lucy Slater3, Tibebe Shenie2 Lameck Machumi1 (presenting), David Sando1, Expeditho Mtisi2, Guerino Chalamilla1, Irene Andrew1, Humphrey Mkali1, Ester 1 University of Minnesota, Minneapolis, MN, USA Mungure1, Eric Aris1 2 National Alliance of State and Territorial AIDS Directors, Addis Ababa, Ethiopia 1 Management and Development for Health, Dar es Salaam, 3 National Alliance of State and Territorial AIDS Directors, Tanzania Washington, DC, USA 2 Harvard School of Public Health, Boston, MA, USA Background: To improve retention in treatment for HIV-positive Introduction: Loss to follow up (LTFU) is a challenge in care and treat- ABSTRACTS patients newly enrolled in care in rural Ethiopia, we implemented a ment programs in Sub Saharan Africa, Tanzania included. We ana- program using trained lay community support workers (CSWs) who lyzed risk factors LTFU among HIV+ patients receiving care and treat- were also HIV positive. CSWs provided education, counseling, social ment in Dar es Salaam, so as to inform strategies to retain patients in support and linkage by cell phone to the HIV clinic for antiretroviral care and treatment services. therapy (ART) or other medical questions. Methods: A cohort of patients enrolled in care and treatment facili- Methodology: 142 clients (75% ART, 25% pre-ART) from Arba Minch ties in Dar es Salaam from 2004-2011 were studied. LTFU was defined town and surrounding villages in southern Ethiopia were each ORAL as missing clinic visit for 90 and 180 consecutive days after the last assigned a CSW, based on residence. Baseline and follow-up sur- appointment date among patients on antiretroviral therapy (ART) and veys after 12 months were conducted. Composite scores were calcu- care respectively. For univariate and multivariate analysis, Cox pro- lated for HIV treatment knowledge; symptoms of chronic illness; portional hazard regression model was employed to identify the risk physical and mental quality of life (QOL); feelings of social support; factors. and internalized stigma. HIV clinic data included CD4+ count and body mass index (BMI). Results: Among 85,608 patients studied, 77% were on ART. Median age of participants was 34 (IQR: 29- 41 years) and median CD4+ count Results: CSWs visited clients 1-4 times/month, and typically spent an was 206 cells/µL (IQR: 84-378 cells/µL). For those on ART, it was found average of 1-2 hours/month with each client. During the first year, 7 that patients aged 50 years and those with CD4+ cell count <100 clients (median baseline CD4+ = 107) died and 2 transferred out; there ³ cells/µL had an independent significantly increased risk of loss to fol- were no losses to follow-up. Of 133 (94%) clients retained in the proj- low up (RR: 1.11, 95% CI 1.03 - 1.19, p < 0.0001 and RR: 1.22, 95% CI 1.10 ect through 12 months, significant changes were seen in client - 1.24, p = 0.01 respectively). Among patients on care and monitoring knowledge, attitudes, and clinical and mental status (see Table). male patients, patients with advanced disease and lower CD4 cell count had significantly increased risk of LTFU (RR: 1.06, 95% CI 1.01 - Conclusions: HIV-positive patients in this pilot CSW program had sig- 1.14, p < 0.04), (RR: 1.26, 95% CI 1.14 - 1.39, p < 0.0001) and (RR: 2.10, nificant improvement in knowledge about HIV, physical and mental 95% CI 2.07 - 2.22, p< 0.0001 respectively). Patients on care and mon- QOL, and perceived social support, with a significant reduction in itoring were more likely to be LTFU compared to patients on ART symptoms of chronic illness and internalized stigma. Client retention (LTFU rate 0.45 vs. 0.21 respectively). in this program was excellent. Conclusions: Determining risk of LTFU at enrolment and initiation of ART and active and focused tracking of patients at risk is important MEAN VALUES AT BASELINE AND 12-MONTH FOLLOW-UP to improve retention rates both for patients on ART and care and monitoring. Baseline 12-months HIV knowledge (# correct = 0-6) 4.7 5.5† Chronic illness (# symptoms = 0-7) 2.3 0.1† Physical QOL (summary score = 0-14*) 7.4 13.6† Mental QOL (summary score = 0-8*) 6.7 7.6† Social Support (summary score = 0-24*) 18.2 21.5† Internal stigma (summary score = 0-5**) 1.6 0.05† CD4+ counts (cells/mm3) 225 396† BMI (kg/m2) 20.5 21.7†* * higher score = better QOL/social support p-value: † = <0.001 ** higher score = greater stigma †* = 0.004

18 9th International Conference on HIV Treatment and Prevention Adherence 260 as apatientnavigator. through theuseofatrainedpeeroutreachworkerwhowouldserve encourage self-acceptanceandincreaseknowledgewouldbe an importantfacilitator. Mostmensuggestedthatthebestwayto Kiswahili) despiteahighlystigmatizingenvironmentalsoemergedas nomic empowerment.Thequestforself-actualization(‘Shikamana’in knowledge, disclosuretoselectindividuals,peersupport,andeco- ART adherenceandcareengagementincludedself-acceptance, poverty, suchashungerandhomelessness.Factorsthatfacilitated MSM-friendly services,depression,substanceabuse,andeffectsof and ofsexualorientation,pooraccesstohealthinformation include lackofpsychosocialsupport,non-disclosureHIVstatus been carriedoutandwereincludedforanalysis.Barriersidentified developing aninterventiontoimproveoutcomesinthisgroup. and challengesfacedbyHIV-positive KenyanMSM,withthegoalof ral therapy(ART)adherence.We aimedtoidentifykeyfacilitators group thatfacesmanybarrierstocareengagementandantiretrovi- among menwhohavesexwith(MSM),ahighlystigmatized Conclusions: Results: Methods: Background: 5 4 3 2 1 Susan Graham Murugi Micheni Sex withMeninCoastalKenya initial conceptualmodel,butemergedasimportantintheIDI. ment incaredespitestigma.Theseconceptswerenotincludedour actualization, alongwithincreasedknowledge,canfacilitateengage and engagementinHIVcareservices.Self-acceptanceself- nation, andsocialisolationthatcanimpedeoptimalARTadherence to identifycommonfactorsinfluencingARTadherence. English, detailedinterviewernotesandtranscriptionswerereviewed focus onaccesstocareandtrustinproviders.Aftertranslationinto motivation-behavioral skillsmodelofadherence,withanadditional open-ended topicguidesusedanapproachbasedontheinformation- discussions (FGD)withtheirhealthcareproviders.Semi-structured, positive MSMrecruitedthroughpurposivesampling,andfocusgroup Jane Simoni 9th InternationalConference on University ofWashington, Seattle,WA, USA Brown University, Providence,RI,USA University ofWashington, Seattle,WA, USA Kenya MedicalResearchInstitute,Kilifi, Kenya MedicalResearchInstitute,Mombasa, Twenty-eight of50plannedIDIand34FGDhave W e conductedindividualin-depthinterviews(IDI)withHIV 3 , DonOperario Adherence amongMenwhohave Facilitators andChallengestoART HIV-positive MSMarechallengedbystigma,discrimi- In coastalKenya,20%oftotalHIVinfectionsoccur 5 1 (presenting) , ElisevanderElst 4 , EduardSanders HIV 2 , BernadetteKombo T reatment andPrevention Adherence 2 , 2 , - - 264 We thereforeevaluatediflinkageratesvariedbyHIVdiagnosissite. known abouthowthesiteofHIVdiagnosisimpactslinkagetocare. improves survivalandisessentialtoHIVpreventionefforts.Little inpatient settings,CTCs,andcorrectionalfacilitiesareneeded. age tocare.Interventionsfocusedonlinkingpatientsdiagnosedin 1 Methods: Background: 3 2 Baligh Yehia with LinkagetoCare Conclusions: Results: Michael Eberhart and calendaryear linkage tocare,adjustingforage,sex,race/ethnicity, HIVriskfactor, regression examinedassociationsbetweensiteofHIVdiagnosisand treatment facilitywithin90daysofdiagnosis.Multivariablelogistic defined ashavingaCD4countorHIV testing center(CTC),orcorrectionalfacility clinic, inpatientsetting(includingemergencyroom),counselingand through 2012.SiteofHIVdiagnosiswascategorizedasoutpatient diagnosed withHIVinPhiladelphia2010and2011,follow-up link tocarecomparedthosediagnosedinanoutpatientclinic. correctional facility(0.06,0.03-0.12)weresignificantlylesslikelyto inpatient setting(AOR0.49,95%CI0.36-0.68),CTC(0.50,0.32-0.79),or al facilities.Inmultivariateanalyses,individualsdiagnosedinan ics, 53%forinpatientsettings,52%CTCs,and12%correction- 60% overall,andvariedbyHIVdiagnosissite:65%foroutpatientclin- respectively (p<0.05).Theproportionofpersonslinkedtocarewas women, andthosewithheterosexualorinjectiondruguserisk, likely tobediagnosedatoutpatientclinicsorCTCsthanolderadults, facilities. Y ics, 15%ininpatientsettings,7%CTCs,and6%correctional MSM HIVrisk.Mostpeople(72%)werediagnosedinoutpatientclin- 2011; 74%weremale,68%Black,60%18-39yearsold,and41%with Philadelphia DepartmentofPublicHealth,Philadelphia,PA, USA Temple University, Philadelphia,PA, USA University ofPennsylvania,Philadelphia,PA, USA Overall, 1,362newHIVdiagnoseswereidentifiedin2010- Retrospective analysisofadults( ounger adults,men,andthosewithMSMriskweremore 1 Site ofHIVdiagnosisisstronglyassociatedwithlink- Diagnosis StronglyAssociated Linking andretainingHIV-infected personsincare (presenting) Location Matters:SiteofHIV 3 . , KathleenBrady , ElizabethKetner 3 -1 RNAtestatanoutpatient 1 , FlorenceMomplaisir . Linkagetocarewas ³ 18-years-old) newly 2 , ORAL ABSTRACTS 19 An iPhone App/Game to Improve New England Healthcare 266 ART Adherence 270Providers’ Perceptions, Knowledge, and Practices Regarding the Use of Laura Whiteley1 (presenting), Larry Brown1, Michelle Lally1, Jacob van den Berg1, Virginia Curtis1, Amethys Nieves1 Antiretrovirals for Prevention

1 Rhode Island Hospital, Providence, RI, USA Douglas Krakower1 (presenting), Jennifer Mitty1, Ira Wilson2, Ann Kurth3, Kevin Maloney4, Donna Gallagher5, Kenneth Mayer6 Introduction: Despite the need for consistent adherence, youth and young adults living with HIV (YLWH) have suboptimal rates of adher- 1 Beth Israel Deaconess Medical Center, Boston, MA, USA ence to antiretroviral treatment (ART). 2 Brown University, Providence, RI, USA 3 New York University, New York, NY, USA Description: We have developed the first iPhone game to promote 4 Beth Israel Deaconess Medical Center, Boston, MA, USA ART adherence among YLWH in the United States. A smart pill bottle 5 New England AIDS Education and Training Center, monitoring device is integrated into the app/game to both measure Boston, MA, USA ABSTRACTS adherence and enhance game play. In-depth interviews were con- 6 Harvard University, Boston, MA, USA ducted with 20 YLWH (ages 14-24) to develop an action-oriented adventure game to increase health information, improve motivation, Background: Prescribing antiretroviral therapy (ART) to HIV-infected and build skills for adherence. Using the information-motivation- persons irrespective of CD4+ count (“early ART”) and to high-risk, behavior model (IMB) as an organizing framework, the interviews HIV-uninfected persons (pre-exposure prophylaxis/”PrEP”) reduces elicited feedback on game characteristics, suitability, and interac- HIV transmission. However, many healthcare providers have not yet tions between the game, the smart pill bottle, and phone. adopted these strategies. ORAL Lessons Learned: Consistent themes emerged from the interviews, Methods: In September-December 2013, 197 providers affiliated with which were integrated into the app/game and will be tested in a ran- New England AIDS Education and Training Center answered online domized controlled trial: (1) gaming characteristics important to surveys assessing perceptions, knowledge, and practices regarding YLWH included directly killing HIV with weapons, improving health by early ART and PrEP. Analyses regarding provision of early ART were taking pills, increasingly harder levels of play to increase self-effica- restricted to ART-prescribing clinicians (n = 112). Multivariable cy, and HIV-relevant images and facts. Participants remarked that regression identified factors associated with intentions to prescribe seeing facts about the deadly consequences of HIV were important early ART. for motivation. (2) YLWH found it acceptable to have HIV pictured in the scenes, but did not want HIV in the game title due to concerns Results: Respondent median age was 44, 57% were female, and 78% about stigma. Several participants noted that many of their older were white. Practitioners included physicians (20% infectious dis- friends with HIV (up to age 29) played iPhone games and would also eases [ID], 20% primary care provider [PCP], 26% other), associate benefit from this game. (3) YLWH desired multiple interactions clinicians (25%), and other (9%). On average, participants had provid- between the technologies (smart pill bottle cap and app/game) and ed HIV care for 12 years and had 53 HIV-infected patients in their requested adherence feedback to their phone and “bonus game care. Most respondents (96%) agreed that early ART decreases HIV points” for adherence. transmission; 56% were aware that treatment guidelines recommend early ART. Sixty-six percent would initiate ART in all patients irre- Recommendations: With a gaming company, Mission Critical spective of CD4+ count, and 69% often/always recommended ART for Studios, we have developed an engaging, IMB-consistent, HIV-spe- CD4+ count ³500 cells/mL. However, providers tended to defer ART if cific app/game to improve ART adherence. It can be played by con- patients were not ready to initiate (78%) or had uncontrolled depres- ference participants on an iPhone. sion (28%) or alcohol/substance abuse (23%). In multivariate analysis, ID physicians were at 4 times greater odds of intending to prescribe early ART than PCPs. Seventy-five percent were aware of nor- mative guidance regarding PrEP provision. Few (9%) were concerned about PrEP efficacy, though 51% indicated safety concerns. Thirteen percent had prescribed PrEP, though 57% indicated they would be likely/very likely to prescribe PrEP to high-risk patients.

Conclusions: Providers believe early ART and PrEP are efficacious. However, only 2/3 report prescribing intentions and practices consistent with implementing early ART, and few have prescribed PrEP despite positive intentions. Interventions to optimize practices regarding ART/treatment-as-prevention and PrEP are needed.

20 9th International Conference on HIV Treatment and Prevention Adherence Acceptability and Feasibility of a Identification of Evidence-Based 273Cell Phone Support Intervention for 274 Interventions for Promoting Youth Living with HIV Nonadherent to Linkage to, Retention, and Re-Engagement in Antiretroviral Therapy HIV Medical Care: Findings from a Systematic Review of Randomized Marvin Belzer1 (presenting), Karen Kolmodin MacDonell2, Sylvie Naar-King2, Shoshana Kahana3, Moussa Sarr4, Controlled Trials, 1996-2012 Sarah Thornton4, Jennifer Haung4, Johanna Olson5, Leslie Clark1 Darrel Higa1 (presenting), Nicole Crepaz1, Cynthia Lyles1 1 Children’s Hospital Los Angeles/University of Southern California, Los Angeles, CA, USA 1 Centers for Disease Control and Prevention, Atlanta, GA, USA 2 Wayne State University, Detroit, MI, USA 3 National Institute on Drug Abuse, Bethesda, MD, USA Background: Engaging and retaining persons living with HIV (PLWH) 4 WESTAT, Rockville, MD, USA in medical care are associated with individual- and population-level ABSTRACTS health benefits, but only 37% of PLWH in the United States are esti- Background: A recent pilot randomized controlled trial of youth ages mated to be in routine HIV medical care. Identifying evidence-based 15-24 nonadherent to antiretroviral therapy (ART) utilizing daily cell interventions (EBIs) that promote linkage to, retention, and re- phone support was found to have significant improvement in self- engagement in care (LRC) can help to increase the proportion of reported adherence and HIV RNA both during the 24 week interven- PLWH receiving regular medical care. tion and 24 weeks post intervention. Understanding acceptability and feasibility is critical to future implementation in clinic settings. Methods: A comprehensive search strategy was developed, includ-

ing searching four electronic databases (MEDLINE, EMBASE, ORAL Methods: Exit interviews were obtained from intervention participants PsychINFO, CINAHL) and manually searching 16 peer-reviewed jour- and adherence facilitators (AF). Acceptability was assessed from nals, reference lists, and listservs. Interventions were eligible if they content analysis of exit interviews. Feasibility was assessed via inter- were published between 1996 and 2012, randomized controlled trials, vention retention and study retention rates. AFs also reported aver- and reported LRC outcome data not solely based on self-report. age length of calls and sites reported on the cost of cell phone plans. Outcomes included linkage to care, retention in care (e.g., multiple visits within a time period), and re-engagement in care. We evaluat- Results: Thirty-seven eligible subjects were enrolled with 19 ed each eligible intervention against established criteria on the qual- assigned to the intervention. Seven subjects (37%) were discontin- ity of study design, implementation, analysis, and strength of findings. ued from the intervention either due to missing over 20% of calls for 2 consecutive months (N = 5) or missing 10 consecutive calls (N = 2). Results: Eight United States and 3 non-United States interventions Calls averaged 3-6 minutes. Scheduling and making calls required were eligible. Four EBIs were identified, including three from the less than one hour per week per subject and phone plans costs var- United States. and one from Uganda. Of the United States-based ied from $25-75 per line. Sixteen of the 19 intervention subjects com- interventions, the strengths-based case management intervention pleted exit interviews, 15 reported the call length was just right, 13 found effects for linkage to and retention in care, while the other two reported they would have liked to continue calls after the 24 week interventions, clinic-based treatment for opioid addiction and inter- intervention and all would recommend the intervention to a friend. active provider alerts, found effects for retention in care. The Uganda Twelve of 13 AFs felt youth made use of the problem solving discus- study, using specialized posttest counseling and home visits by com- sions, 4 reported calls were intrusive, and 9 reported the most diffi- munity workers, found an effect for linkage to care. Of the 7 remain- cult part of being an AF was coordinating times for the calls. ing eligible interventions, 6 reported null findings. No EBIs were identified for re-engaging PLWH who fell out of care. Conclusions: Providing cell phone support to youth non-adherent to ART was acceptable and feasible. Seven youth did not complete the Conclusions: Despite the emerging body of LRC research, only a few intervention and future studies will need to determine if lowering call evidence-based interventions were identified. More studies are need- adherence requirements improves intervention completion without ed, especially those focusing on re-engagement. Integrating EBIs into adversely impacting adherence outcomes. Future studies will need practice, particularly targeting PLWH less likely to access medical to explore if providing for the cost of cell phone plans is essential. care, is needed to improve health outcomes and reduce disparities.

9th International Conference on HIV Treatment and Prevention Adherence 21 DNA and Protein Biomarkers Offer The Effectiveness of a New 275Increased Accuracy for Assessing 276Smartphone App Using Vaginal Microbicide Gel Adherence Personalized Health-Related Visual Imagery in Improving Adherence to Antiretroviral Andrea Thurman1 (presenting), Terry Jacot2, Christine Mauck3, Jill Schwartz3, Gustavo Doncel4 Therapy

1 CONRAD, Norfolk, VA, USA Keith Petrie1 (presenting), Anna Perera1, Mark Thomas1, 2 CONRAD/Eastern Virginia Medical School, Norfolk, VA, USA John Moore2 3 CONRAD, Arlington, VA, USA 4 CONRAD EVMS, Norfolk, VA, USA 1 University of Auckland, Auckland, New Zealand 2 Massachusetts Institute of Technology, Boston, MA, USA Background: Current measures of product adherence and protocol compliance in human immunodeficiency virus (HIV) prevention trials, Background: Mobile technologies have potential to deliver adher-

ABSTRACTS which include participant report of product use, visual inspection of ence promoting interventions which are cost-effective and scalable. returned applicators (VIRA), dye stain assays and ultraviolet light However, existing mobile adherence interventions generally employ (UVL) assessment are suboptimal. DNA and protein based biomark- automated SMS reminders and have limited long-term efficacy as ers, obtained from returned, used vaginal gel applicators, were devel- they focus on the forgetfulness aspect of medication non-adherence, oped to determine vaginal insertion more accurately. The sensitivity and do not address the cognitive determinants of non-adherent and specificity of DNA/protein biomarkers in detecting vaginal inser- behavior. This study aimed to investigate the effectiveness of a tion were determined and compared to VIRA and UVL assessment. smartphone app that links personal adherence to real-time simulat- ed visual displays of viral activity on adherence to antiretroviral ther- ORAL Methods: Forty non-pregnant, HIV-negative women were given 12 apy (ART), illness perceptions, and medication beliefs. These fea- applicators containing hydroxyethyl cellulose (HEC) universal place- tures were intended to modify participants’ conceptualizations of bo: four applicators were handled but not vaginally inserted (“sham”) their HIV infection and ART regimen to be more conducive to optimal and eight applicators were vaginally inserted under direct observa- adherence. tion. Half of the applicators (n = 240) were assessed within 7 days of the clinic visit, and half (n = 240) were assessed after storage for 30 Methods: 28 people on ART were randomized to use either a standard days. VIRA and UVL inspection of gel applicators was performed by or augmented version of the smartphone app. The augmented version three independent, blinded readers. Blinded laboratory staff ana- contained additional components which illustrated to the participant, lyzed vaginal markers (bacterial DNA and Cytokeratin 4) obtained by in real-time, the estimated concentrations of antiretroviral agents in swabbing the returned applicators. blood plasma as well as personalized immune activity. Adherence to ART was assessed at baseline and three month follow-up using HIV Results: The sensitivity of VIRA (52 - 54%) and UVL assessment (74 - viral load, pharmacy dispensing and self-report. Information was also 92%) was significantly lower at both the 7-and 30-day timeframes, collected on illness perceptions and medication beliefs. compared to DNA and protein biomarkers (95 - 100%). The specificity of VIRA (49 - 78%) and UVL assessment (66 - 73%) was significant- Results: Participants who received the augmented application dis- ly lower at both the 7-and 30-day assessments, compared to DNA played a significantly higher level of self-reported adherence to ART and protein biomarkers (100%). at follow-up, compared to individuals who received the standard version (p = .02). Participants who received the augmented version of Conclusions: DNA and protein biomarkers were superior to VIRA and the application also had a significantly lower HIV viral load at three UVL assessment in detecting the vaginal insertion of HEC placebo gel month follow-up than those who received the standard version (p = applicators when assessed within 7 days of use and after 30 days of .027). Further, there was a significant decrease from baseline to three storage at room temperature. These highly sensitive and specific bio- month follow-up in the proportion of participants classified as non- markers will be further investigated and optimized to accurately adherent in the augmented group p = .03). Greater usage of the app characterize product adherence and protocol compliance in large was associated with improvements understanding of HIV infection women’s health research trials. and perceived necessity for ART.

Conclusion: This research suggests that a smartphone app, using personalized health-related imagery and real-time feedback, is effective in facilitating adherence to ART. Study findings indicate that such an intervention may have considerable utility in the clinical care of people with HIV infection.

22 9th International Conference on HIV Treatment and Prevention Adherence Gender Disparities in Viral Patterns of HIV Service Use and 283Suppression and Antiretroviral 287HIV Viral Suppression among Therapy Use by Racial and Ethnic Group - Patients Treated in an Academic Infectious Medical Monitoring Project, 2009 Diseases Clinic in North Carolina

Linda Beer1 (presenting), Christine Mattson1, William Short2, Anton Palma1 (presenting), Lynne Messer2, David Lounsbury3, Jacek Skarbinski1 Evelyn Quinlivan4

1 Centers for Disease Control and Prevention, Atlanta, GA, USA 1 Columbia University, New York, NY, USA 2 Jefferson Medical College, West Deptford, NJ, USA 2 Portland State University, Portland, OR, USA 3 Albert Einstein College of Medicine, Bronx, NY, USA Background: Women comprise a quarter of persons living with HIV in 4 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA the United States and the majority of these women are black or Hispanic/Latina. Because female gender and non-white race/ethnic- Background: Many people living with HIV (PLWH) have difficulties ABSTRACTS ity have often been found to be associated with poor clinical out- participating in care, including keeping regularly-scheduled clinic comes, stratified analysis is needed to identify areas for targeted visits, taking medications and having regular laboratory assess- intervention relevant to specific groups of women, but few prior stud- ments, thereby preventing effective management of HIV disease. ies have had sufficient sample size to do so. Irregular medical care hinders HIV RNA suppression and impacts HIV-related health outcomes. However, the optimal frequency of Methods: We analyzed weighted data from the Medical Monitoring medical visits needed to sustain viral suppression and good health is Project (MMP), a national probability sample of HIV-infected adults a matter of ongoing deliberation.

receiving medical care in the United States between January-April ORAL 2009. We estimated the prevalence of self-reported antiretroviral Methods: To assess how medical care usage influences patient out- therapy (ART) use and medical record documentation of recent viral comes, we conducted cluster analysis of medical visit data from suppression (undetectable or ²200 copies/mL) among men and 1,748 PLWH attending a large academic medical center, using two women by racial and ethnic group. measures of HIV service use: annualized visits per year and the HIV/AIDS Bureau (HAB) performance measure (2 visits/year at least Results: Among 4,152 persons estimated to represent 414,335 per- 90 days apart). Results yielded three clusters of HIV service usage sons in care, 74% of men and 66% of women were virally suppressed patterns, which we validated on a randomly selected subset of the (“suppressed”) (p<0.001). Stratified by racial/ethnic group, 81% of data. We then identified demographic and clinic predictors of these white men vs. 69% of white women (p=0.001), 64% of black men vs. patterns and used the clusters to predict future health outcomes. 64% of black women (p=0.824), and 76% of Hispanic/Latino men vs. 72% of Hispanic/Latina women (p=0.430) were suppressed. Stratified Results: Patients in the “engaged in care” cluster exhibited most by racial/ethnic group, women were significantly less likely than men consistent retention in care (average 80% of follow-up years meeting to take ART (range: p<0.001- p<0.028), but among those prescribed HAB), optimal annualized visits (2.9 mean visits/year) and sustained ART, women were equally likely to be suppressed compared to men. viral suppression (>73% HIV RNA tests <400 copies/mL). Patients in “sporadic care” demonstrated lower retention (46-52% years meet- Discussion: Among persons in care, we saw disparities in suppres- ing HAB), visit use (1.7 visits/year) and viral suppression (56% tests sion between white men and women and between racial/ethnic <400 copies/mL). Patients with “frequent use” demonstrated higher groups. Among those taking ART, men and women of the same than average visit use (5.2 visits/year) and had more hospitalizations race/ethnicity did equally well. Further exploration is warranted, but (RR = 2.41, 95% CI [1.75, 3.07]) and emergency room visits (RR = 2.60 understanding barriers to ART use among women and decreasing [1.80, 3.41]) than “engaged in care” patients. Female gender, out of racial/ethnic disparities in suppression may help decrease gender state residency, low visit attendance during the first 12 months of disparities in viral suppression. observation and detectable first HIV RNA test were early predictors of subsequent service usage.

Conclusions: Results confirm earlier findings that under-utilization of services predicts poorer viral suppression and health outcomes and support current recommendations for visit frequency of 2-3 visits/year.

9th International Conference on HIV Treatment and Prevention Adherence 23 What Do We Really Mean By First Findings of a Novel, 297‘Adherence’ in Vaginal Microbicide 301Variable, Rewards-Based Trials? A Comparative Study Adherence Intervention in Uganda

Lori Miller1 (presenting), Richard Hayes1 Sebastian Linnemayr1 (presenting), Glenn Wagner1, Barbara Mukasa2 1 London School of Hygiene and Tropical Medicine, London, United Kingdom 1 RAND Corporation, Santa Monica, CA, USA 2 Mildmay Uganda, Kampala, Uganda Background: Understanding participant adherence is critical to interpret results from HIV prevention microbicide trials. Individual Background: Antiretroviral therapy (ART) requires life-long, consis- trial teams determine how to assess adherence in their trials, yet this tent medication adherence that many patients struggle with. can result in challenges for comparing adherence across trials. This Successful interventions need to address two fundamental issues: study critically examined how 6 completed trials measured adher- first, they need to bolster adherence and retention over a person’s

ABSTRACTS ence, analyzed adherence data, and reported gel adherence in the lifetime. Second, they need to be adapted to the resource-con- primary trial manuscripts. straints of developing countries to be sustainable. The Rewarding Adherence Program (RAP) is designed to improve motivation ART Methodology: Trials included were Cellulose Sulfate CONRAD, adherence and addresses the needs of a growing number of treat- Cellulose Sulfate FHI, HPTN 035, Carraguard, MDP 301 and CAPRISA ment-mature clients in sub-Saharan Africa at low cost. 004. Protocols, case report forms, and statistical analysis plans were obtained from study teams to identify variables used to collect adher- Description: RAP is a randomized controlled trial offering small ence data and understand how intermediate measures of adherence rewards allocated through a drawing conditional on timely clinic vis- ORAL were calculated. Primary trial manuscripts were reviewed to deter- its and ART adherence of 95% or higher to 157 HIV-positive clients in mine how overall estimates of adherence were reported. A survey Uganda’s capital Kampala. Primary outcome measures are timely was conducted with trial teams to clarify methods and understand clinic visits and ART adherence objectively measured by the lessons learned. Medication Event Monitoring System (MEMS). We evaluate this novel intervention with regard to acceptability/feasibility and also Results: Six trials used self-reported data to measure adherence; present preliminary evidence for short-term impact. three trials collected data on applicator use. Five trials collected data on sex acts in the past 7 days, 4 trials collected data on last sex act, Lessons Learned: Focus group discussions revealed broad accept- and 2 trials collected data on last 30 days. Two trials using self- ability and enthusiasm for the project that is also reflected by a low reported and applicator data used a mixed-methods measure to refusal rate (4.6%). Clients in the intervention group after 3 months report overall adherence. How adherence was calculated was not showed 6.8 percentage points higher mean ART adherence (83.9% always clear in statistical analysis plans and primary manuscripts, vs. 77.2%, p-value .05), and a higher fraction of 95% adherence or thus clarification from trial teams was required in some cases. higher (30.4% vs. 20%, p = .20). Similarly, clients in the intervention group show higher levels of kept timely clinic appointments, although Conclusions: Primary trial manuscripts did not always provide clear the result is not statistically significant. information about adherence estimates for the trial. Most trial teams agreed last sex act was a good recall period for general population Recommendations: At month 3 of the intervention, we find that small participants. Asking about reasons for non-use of gel can be a way rewards given out through a prize drawing conditional on two adher- to obtain data about sex acts not covered by gel, potentially reducing ence-related target behaviors are acceptable, and find preliminary social desirability bias created by standard questions about product evidence of their effectiveness on adherence and timely clinic visits. use. There was great variability in methods used for estimating and We will test the intervention for a total of 20 months to test longer- reporting adherence in these trials. The field would benefit from term effectiveness and arrive at more conclusive results. more standardization.

24 9th International Conference on HIV Treatment and Prevention Adherence Time of HIV Diagnosis and Changing Trend in Adherence to 303 Engagement in Prenatal Care 315 Highly Active Antiretroviral Impact Outcomes in Pregnant Women Therapy in the Multicenter AIDS Cohort with HIV Study and the AIDS Linked to Intravenous Experience Florence Momplaisir1 (presenting), Kathleen Brady2, Dana Thompson3, Baligh Yehia4 Shilpa Viswanathan1 (presenting), Roger Detels2, Bernard Macatangay3, Shruti Mehta4, Gregory Kirk1, 1 Temple University, Philadelphia, PA, USA Lisa Jacobson4 2 Philadelphia Department of Public Health, Philadelphia, PA, USA 3 Drexel University Hospital, Philadelphia, PA, USA 1 Johns Hopkins Bloomberg School of Public Health, Baltimore, 4 University of Pennsylvania, Philadelphia, PA, USA MD, USA 2 UCLA Fielding School of Public Health, Los Angeles, CA, USA Background: Little is known about the care continuum for HIV-infect- 3 University of Pittsburgh School of Medicine, Pittsburgh, PA, USA ABSTRACTS ed pregnant women. We examined the impact of time of HIV diagno- 4 Johns Hopkins Bloomberg School of Public Health, Baltimore, sis and engagement in prenatal care on use of antiretroviral therapy MD, USA (ART) and viral suppression during pregnancy. Background: The goal of the study was to determine whether the Methodology: Retrospective analysis of 836 pregnancies, involving effect of adherence on HIV RNA suppression changed over time with 656 HIV-infected women, in Philadelphia between 2005 and 2013. simpler and improved formulations, and to estimate the minimum cut- Time of HIV diagnosis was classified as: (1) before or (2) during preg- off of adherence to newer highly active antiretroviral therapy nancy. Engagement in prenatal care (adequate, intermediate, and ORAL (HAART) needed for population HIV RNA suppression. inadequate) was assessed using the Kessner index which accounts for time of entry into prenatal care, number of prenatal visits, and Methods: Longitudinal data from the Multicenter AIDS Cohort Study infants’ gestational age at delivery., and was grouped as (1) ade- (MACS) and the AIDS Linked to Intravenous Experience (ALIVE) col- quate, (2) intermediate, and (3) inadequate. Outcomes were: (1) use lected prospectively between March 2001 and December 2011 were of ART during pregnancy and (2) HIV viral load <200 copies/mL at the used. Adherence was calculated from reported use compared to time closest to delivery. Multivariable regression examined associa- prescribed use over the 4 days prior to the 6-month study visit, using tions between time of HIV diagnosis, engagement in prenatal care, an established algorithm. Linear mixed models were used to study and outcomes, adjusting for age, race/ethnicity, use of illicit drugs, the effect of time on adherence. The minimum needed adherence and calendar year. cutoff was defined as the level at which 78% of the population was suppressed, using an established metric observed in prior studies Results: Overall, 636 pregnancies (76%) involved women diagnosed with 95% adherence. with HIV before and 200 (24%) during pregnancy. Women were ade- ³ quately engaged in care in 39% of pregnancies, intermediately Results: Our study population consisted of 1,006 HAART users with engaged in 38%, and inadequately engaged in 23%. Compared to 10,971 person-visits in the MACS, and 197 HAART users with 1,745 pregnancies with HIV diagnosed during pregnancy, those with HIV person-visits in the ALIVE. The proportion reporting 100% adherence, diagnosed before pregnancy had greater (p <0.05) use of ART (89% and the proportion with suppressed viral load among those with vs. 73%) and viral suppression (73% vs. 46%). Similarly, 92% of preg- <95% adherence increased over time in both cohorts such that by nancies with women adequately engaged in prenatal care had ART 2011, 76% and 55% overall were suppressed in the MACS and ALIVE prescribed and 43% had viral suppression compared to 55% and 26% respectively. With current therapy, 78% suppressed HIV RNA of pregnancies where women were inadequately engaged, respec- between 75% and 79% adherence in the MACS, and did not signifi- tively (p <0.05). Multivariate models confirmed that time of HIV diag- cantly differ than the population suppressed at adherence ³95%. In nosis and engagement in prenatal care were strongly associated (p the ALIVE, we did not observe a minimum adherence cutoff below <0.05) with the outcomes. 95% because <78% suppression was observed even among those with 95% adherence (73.5%). Conclusions: Targeted interventions to diagnose women early and to ³ improve engagement in prenatal care are needed, and have the Conclusions: Although all HIV-infected persons should be instructed potential to improve outcomes for mothers and infants. to be 100% adherent, concerns of adherence should not hinder prescribing new HAART regimens early in HIV infection. Comprehensive adherence counseling sessions for current and former IDUs are still warranted.

9th International Conference on HIV Treatment and Prevention Adherence 25 Adherence to Pre-Exposure Bridging the Gaps: The Use of 317 Prophylaxis in the Partners 322 Health Information Technology and Demonstration Project: Preliminary Findings Bridge Counseling to Improve Retention in Care in North Carolina Jessica Haberer1 (presenting), Jared Baeten2, Renee Heffron2, Deborah Donnell3, Nelly Mugo4, Elly Katabira5, Elizabeth Bukusi6, Jennifer Keller1 (presenting), Kristen Sullivan2, Katherine Schafer1, Stephen Asiimwe7, Katherine Thomas2, Lara Kidoguchi2, Mary Beth Cox3, Amy Heine4, Jacquelyn Clymore3, Arlene Seña4, Connie Celum2, David Bangsberg1 Aimee Wilkin1 1 Massachusetts General Hospital, Boston, MA, USA 1 Wake Forest University School of Medicine, Winston Salem, NC, USA 2 University of Washington, Seattle, WA, USA 2 Duke University, Durham, NC, USA 3 Fred Hutchinson Cancer Research Center, Seattle, WA, USA 3 North Carolina Department of Health and Human Services, 4 KEMRI, Nairobi, Kenya Raleigh, NC, USA 5 Infectious Disease Institute, Kampala, Uganda 4 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA ABSTRACTS 6 Kenya Medical Research Institute, Kisumu, Kenya 7 Kabwohe Clinical Research Centre, Kabwohe, Uganda Introduction: Only 37% of people living with HIV (PLWH) engage in regular medical care. As part of NC-LINK, a statewide initiative to Background: Effectiveness of PrEP is highly dependent on adher- increase the number of PLWH engaged in regular medical care in ence. Data from the Partners PrEP Study showed high adherence NC, a retention intervention utilizing CAREWare was implemented in within HIV serodiscordant couples in East Africa. It is unknown if a 2,000 patient HIV clinic in Western North Carolina. similar adherence levels will be seen with open-label use outside clinical trials. ORAL Description: The Plan-Do-Study-Act model was used to develop a “Bridge Counseling” protocol to identify out of care (OOC) patients (no Methods: The Partners Demonstration Project involves high risk HIV medical visit 9 months) using CAREWare and return them to care serodiscordant couples in Kenya and Uganda. Couples are offered ³ (RTC). Contact was attempted using phone calls, letters, internet comprehensive HIV prevention services, including daily oral PrEP searches, and other providers/pharmacies. If contact was success- and antiretroviral therapy per national guidelines. Adherence is ful, staff addressed specific barriers to care and scheduled medical measured through the medication event monitoring system (MEMS) appointments. All efforts were tracked in CAREWare. If no contact and self-report. This analysis reflects up to the first three months of was made or no other definitive outcome (DO) identified such as relo- data per participant. cation, death, or incarceration, patients were referred via CAREWare to a state bridge counselor (SBC) for further follow-up. Results: Data are available on 437 HIV-uninfected participants; 422 (97%) initiated PrEP at enrollment. Of 200 participants with a month 3 Lessons Learned: Over 10 months, 510 OOC patients were identified visit, 2 (1.5%) had not yet initiated PrEP. and 351 (69%) had a DO of which 219 (43%) were RTC. One hundred fifty-six (31%) were referred to the SBC, who has linked 43 to care HIV-uninfected participant characteristics N (%) or Median (IQR) and identified other DO for 31 patients. The intervention improved Female 131 (30%) retention in HIV medical care from baseline. Medical visit frequency Age (years) 29 (26-36) (1 visit every 6 months) increased from 57% to 69% (p <0.0001) and Education (years) 8 (7-12) Partnership duration (years) 2.3 (0.8-6.0) new patient retention (1 visit every 4 months in their first year of care) Aware of HIV serodiscordancy prior to enrollment 122 (28%) increased from 51% to 63% (p = 0.32). CAREWare was an effective Unprotected sex with study partner in prior month 285 (65%) tool in finding, tracking, and referring OOC patients to the SBC. Belief of moderate/high risk of Collaboration between the clinic and the SBC program decreased HIV acquisition from partner 132 (31%) the numbers of OOC PLWH requiring SBC intervention. No concerns about daily use of PrEP 376 (87%) Recommendations: Bridge counseling at the clinic and statewide HIV-infected partner characteristics level can be effective in re-engaging OOC PLWH and CAREWare can CD4 count (cells/mm3) 423 (262-598) facilitate this process. Future directions include standardizing HIV RNA (log10 copies/mL) 4.6 (4.0-5.0) enhanced outreach for PLWH with missed appointments. Using MEMS, median adherence to PrEP was 98% (IQR 89-100%) and mean adherence was 90% (SD 20%); 18% of participants had adherence <80%. At month 3, 11 participants (6%) self-reported <80% adherence by missed doses, 84 (42%) reported less than “very good”/”excellent” adherence, and 32 (16%) reported taking less than “most”/”all doses.”

Conclusions: Preliminary data suggest high uptake and high adherence to PrEP among HIV serodiscordant couples in East Africa with a minority exhibiting suboptimal adherence in the first 3 months. Longitudinal data will be critical for understanding adherence patterns, factors influencing adherence, and decisions for continuation/discontinuation of PrEP.

26 9th International Conference on HIV Treatment and Prevention Adherence Feedback of HIV Medication Choosing Prevention: The Case for 323 Adherence Data to Patients: 329 Involving Users in Early Assessment of Presentation Formats Microbicide Development

Anita Lyons1, Warren Bilker2, Janet Hines3, Kathleen Morrow1 (presenting), Shira Dunsiger2, Joseph Fava3, Robert Gross3 (presenting) Rochelle Rosen2, Julia Shaw3, Sara Vargas1, Project LINK Study Team3

1 UPMC Internal Medicine Residency, Pittsburgh, PA, USA 1 Miriam Hospital/Brown Medical School, Providence, RI, USA 2 University of Pennsylvania Perelman School of Medicine, 2 Miriam Hospital/Brown University, Providence, RI, USA Philadelphia, PA, USA 3 Miriam Hospital, Providence, RI, USA 3 University of Pennsylvania, Philadelphia, PA, USA Background: User experience is critical to whether or not individuals Background: Medication adherence feedback is an important com- choose to use microbicides. Understanding patterns of experiences ponent of behavioral interventions. It is unknown which presentation elicited by product properties and their use is critical to development

format patients most easily understand. and impact of prevention technologies. The goal was to identify sen- ABSTRACTS sory perceptions women experience and discriminate during vaginal Methods: HIV-infected adults on antiretrovirals for ³6 months were gel use, and identify experience patterns that help predict use. presented examples of one month’s refill data in 5 formats: (1) medication possession ratio (MPR), (2) number of days late to refill, (3) Methods: 204 participants evaluated 4 vaginal gels (with differing calendar plot of days with/without medications, (4) pie chart of pro- properties; randomized) and chose a preferred product. Validated portion of doses taken/not taken, and (5) a color-coded letter grade. User Sensory Perception & Experience (USPE) scale scores differen- Five scenarios (>95%, 90-95%, 80-90%, 70-80% and <70%) were pre- tiated gels across experiences during coital activity. Latent class sented in each format. Participants were queried regarding how modeling identified patterns of USPEs based on participant respons- ORAL much adherence improvement was needed for each with “good es to their choice product. understanding” defined as answering >4 of 5 scenarios correctly and which format they preferred. We calculated relative risks (RR) for Results: Data support 4 classes of USPEs in preferred experiences, “good understanding” using logistic regression with “days late” as with frequencies as follows: 0.14, 0.28, 0.25, 0.33 (posterior probabili- the base case format. ties of belonging to each class). Class 1 is characterized by lower than average scores on initial penetration, lubrication and perceived Results: 124 participants were median age 48.5, 65% white, 34% wetness, but higher scores on intravaginal awareness and messi- African American, 71% male. Only 43% had good understanding of ness (compared to 2/3 of remaining classes). Class 2 had the highest MPR. The association between each format and understanding dif- overall averaged mean scale scores, with the exception of initial fered by education (test for interaction, p = 0.001). For ²12 years edu- penetration, on each USPE scale. The Class 2 experience overall cation, the relative risks were: calendar 0.9 (95% CI: 0.6-1.3), pie would appear to be the most noticeable experience from a user per- chart 0.6 (95% CI: 0.4-0.9), letter grade 0.5 (95% CI: 0.4-0.8), and MPR spective: higher lubrication, stimulation and wetness sensations, and 0.3 (95% CI: 0.2-0.4). For >12 years education, the RRs were calendar a higher degree of messiness and leakage. Qualitative interpreta- 1.8 (95% CI: 1.3-2.6), pie chart 2.2 (95% CI: 1.3-3.6), letter grade 1.3 tions can also be made for Classes 3 and 4. There were no between- (95% CI: 0.9-2.0), and MPR 0.4 (95% CI: 0.2-0.7). 33 (27%) did not class differences in sociodemographics, prior STD infection, or hor- achieve good understanding on their preferred format. monal contraceptive use. Conclusions: We found large differences in the ability to understand Conclusions: Microbicide effectiveness is predicated on optimal adherence data presented in different ways. Those with higher edu- drug delivery and use adherence. Both drug delivery and use are cation had a broader range of formats they understood. Interventions impacted by biophysical properties of gel formulations. Clinical trials should use calendar plots or number of days late for feedback. have been challenged by low adherence, obviating proof of concept. Patients who do not understand adherence data even in their By understanding the correspondence between product properties favorite format may need other approaches to feedback. and the user experience elicited by those properties, the likelihood of improved adherence and impact on HIV incidence could be realized.

9th International Conference on HIV Treatment and Prevention Adherence 27 2014 Federal Recommendations Engagement in Outpatient Care for 330 about HIV Prevention Services for 334 Patients Living with HIV Persons with HIV: Promoting Synergies Christine Oramasionwu1 (presenting), Stacy Cooper Bailey1, between Clinicians, Community-Based Terence Johnson1, Lu Mao2 Organizations, and Health Departments 1 University of North Carolina at Chapel Hill, UNC Eshelman School of Pharmacy, Chapel Hill, NC, USA Amrita Tailor1 (presenting), Kathleen Irwin1, Gema Dumitru1, 2 University of North Carolina at Chapel Hill, UNC Gillings School of Priya Jakhmola1, Anna Huang2 Global Public Health, Chapel Hill, NC, USA 1 Centers for Disease Control and Prevention, Atlanta, GA, USA Background: The HIV cascade illustrates the varying level of engage- 2 Health Resources and Services Administration, Rockville, MD, USA ment within the stages of HIV care by people living with HIV (PLWH). Prior studies that have applied the treatment cascade to various pop- Background: To advance national goals to reduce onward transmis- ulations were restricted to patients receiving care in HIV medical ABSTRACTS sion from persons with HIV (“prevention with positives”), Centers for care facilities. These data are not representative of care received by Disease Control and Prevention, Health Resources Services PLWH throughout the United States, as not all PLWH receive care in Administration, and six non-governmental organizations updated the HIV clinics. This study sought to estimate engagement in outpatient 2003 recommendations to incorporate HIV prevention into HIV med- care for PLWH, beyond facilities that specialize in HIV. ical care. The update was directed to clinicians, community-based organizations (CBO), and health departments (HD) to expand the Methods: Cross-sectional data from the 2009-2010 National Hospital workforce and resources focused on “prevention with positives.” Ambulatory Medical Care Survey (NHAMCS) were used to conduct

ORAL this study. ICD-9-CM codes were used to denote HIV infection (042, Methods: The workgroup developed a comprehensive set of recom- V08, 079.53). Levels of care included: receiving any care ( 1 clinic mendations based on other recent federal guidelines, systematic lit- ³ visit for a PLWH), receiving HIV care (clinic visit with a primary ICD- erature reviews, program evaluations, and opinions of >80 experts in 9-CM code for HIV), established in care (patient previously seen HIV prevention, care, and policy. The recommendations addressed within the clinic), engaged in care ( 2 clinic visits in the past year), linkage and retention in care, HIV treatment and adherence, behav- ³ and prescribed ARV (documentation of 1 ARV medication). Factors ioral risk reduction, partner services, STD services, reproductive and ³ associated with ARV prescription were determined by logistic pregnancy care, and other medical and social services that influence regression. HIV transmission. They included new and longstanding, but underuti- lized, interventions that apply social, structural, ethical, legal, behav- Results: ~2.6 million outpatient clinic visits for PLWH were analyzed. ioral, and biomedical approaches. Of these, 90% were receiving HIV-related care, 86% were established in care, 75% were engaged in care, and 56% were prescribed Results: Many recommended interventions promote synergies ARV. In stratified analysis, engagement in the various levels of care between clinicians, CBOs, and HDs. For example, CBO testing pro- did not vary by gender or by race/ethnicity. Routine engagement in grams can link newly diagnosed persons to HIV medical care by care and ARV coverage was less common in younger patients than helping with health insurance enrollment, appointments, and trans- in older patients (p <0.05). Type of provider seen was associated with portation. HD surveillance programs can identify testing sites that ARV prescription (OR = 0.27, 95% CI = 0.15-0.51) whereas routine warrant linkage facilitators by tracking reported cases of HIV not fol- engagement in care was not associated with ARV prescription lowed by reported CD4 cell counts. To complement adherence (OR=0.99, 95% CI = 0.96-1.03). advice from clinicians, CBOs can offer community-based ART support and surveillance programs can identify persons with persistent- Conclusions: PLWH with established outpatient care are still not ly high viral loads that warrant adherence support. Like clinicians, being prescribed ARV. With the change in ARV prescription guide- CBO and HD partner services specialists can inform HIV-uninfected lines, further research will be needed to understand factors deciding partners about options for post-exposure and pre-exposure prophy- when a PLWH initiates ARV in the outpatient setting. laxis to reduce risk of acquiring HIV.

Conclusions: By stressing synergies between different health sec- tors, these new recommendations promise to accelerate “prevention with positives” and reduce onward HIV transmission in the United States. Some communities have used these synergistic strategies and developed best practices to guide HIV prevention providers.

28 9th International Conference on HIV Treatment and Prevention Adherence Engagement in the Continuum of Acceptability of HIV Pre-Exposure 337 HIV Care in Estonia 339 Prophylaxis at Varying Levels of Effectiveness among Low SES Black Gay Kaja-Triin Laisaar1 (presenting), Mait Raag1, Heli Rajasaar2, Anneli Uusküla1 and Bisexual Men in Los Angeles: Implications for PrEP Dissemination 1 Department of Public Health, University of Tartu, Tartu, Estonia 2 University of Tartu, Tartu, Estonia Ronald Brooks1 (presenting), Vincent Allen2, Stanley Johnson1 Background: According to recent international reports, Estonia had 1 UCLA Department of Family Medicine, Los Angeles, CA, USA the highest rate of new HIV cases (23.5 per 100,000) in the European 2 UCLA Department of Psychology, Los Angeles, CA, USA Union, and the estimated HIV prevalence in the adult population was 1.3%. HIV care, including antiretroviral therapy (ART) is free of Background: In the United States, HIV incidence is highest among charge for persons living with HIV/AIDS (PLWHA), regardless of their Black men who have sex with men (BMSM). The goals of this study medical insurance status. Yet to maximize health benefits of care, were to assess perceptions of and adoption intentions for Pre- ABSTRACTS health system has to ensure effectiveness and quality of a cascade Exposure Prophylaxis (PrEP) among high-risk BMSM. of services, provided to PLWHA. Methods: Interviews were conducted with 224 HIV-uninfected Methods: To describe and quantify the spectrum of PLWHA engage- BMSM in Los Angeles. Survey domains included: PrEP awareness, ment in HIV care in Estonia, we reviewed published reports, retrieved intention to adopt PrEP, perceptions of PrEP, sexual risk behavior, data from surveillance and administrative databases. We defined the alcohol and substance use, and socio-demographic characteristics. cascade indicators based on Centers for Disease Control and Chi-square tests and regression analyses were conducted to deter- Prevention and the Institute of Medicine metrics: the number of indi- ORAL mine predictors of future PrEP adoption. viduals (1) infected with HIV, (2) diagnosed with HIV, (3) linked to HIV care, (4) retained in HIV care, (5) on ART, and (6) with suppressed Results: Participants ranged in age from 18 to 69 (M = 33.5). Only a viral load (HIV-RNA < 200 copies/mL). third (33%) of participants were aware of PrEP. More than half (60%) of the participants indicated a likelihood of adopting a PrEP medica- Results: In 2012 and 2013 the estimated number of PLWHA in Estonia tion that was 90% effective in preventing HIV infection. Acceptability was between 7,200 and 11,000, averaging 9,100. About 87% of them decreased for PrEP at lower levels of effectiveness: 7%, 16% and 30% were diagnosed, 59% were linked to and 33% retained in HIV care. acceptability for PrEP at 44%, 50% and 73% effectiveness, respective- 29% were receiving ART, yet only 19% were taking full advantage of ly. Participants were less likely to indicate an intention to adopt PrEP it (i.e. had suppressed viral load). Of PLWHA linked to care 92% were if they agreed with the following statements: “I would be very uncom- linked within 3 months of diagnosis, yet at linkage 61% had CD4 count fortable taking HIV medicines when I don’t have HIV” (AOR: Adjusted 350 cells/mm3. ² Odds Ratio = 0.39; CI: 95% Confidence Interval = 0.16-0.91)” and “Not knowing if there are long-term side effects of taking a daily HIV med- Conclusions: Identifying the gaps in connecting PLWHA to sustained icine makes me very uncomfortable” (AOR = 0.36; CI = 0.14-0.88). and quality care enables policymakers and service providers imple- Younger participants (18-28 years) were more likely than older partic- ment system improvements and service enhancements that better ipants to report intentions to adopt PrEP (AOR = 2.3; CI = 1.1-4.9). support individuals as they move from one step in the continuum to the next. In Estonia, PLWHA reaching the healthcare system with low Conclusions: BMSM may adopt a highly effective PrEP medication CD4 counts i.e. too late can be explained by delayed testing after HIV for HIV prevention but will have diminishing interest for PrEP medica- acquisition, the issue requiring most urgent attention in fighting tions with mid- to low-levels of effectiveness. Negative perceptions HIV/AIDS. of PrEP may also limit uptake of PrEP. Programs to address concerns and highlight the benefits of PrEP may help facilitate both greater awareness and uptake of PrEP among high-risk BMSM.

9th International Conference on HIV Treatment and Prevention Adherence 29 Development of an Evidence-Based Why Do People Come Back for 345 Alert for Risk of ART Failure within 347 Follow-up? A Prospective Study of an Electronic Medical Record in Haiti Post-Sexual Exposition Prophylaxis

Nancy Puttkammer1 (presenting), Steven Zeliadt2, Janet Baseman2, Nima Machouf1 (presenting), Réjean Thomas1, Sylvie Vézina1, Rodney Destine3, Nathaelf Hyppolite3, Nernst Raphael4, Daniele Longpré1, Danielle Legault1, Morency Duchastel1, Kenneth Sherr2, Krista Yuhas2, Scott Barnhart2 Michelle Milne1, Jason Friedman1, Amélie Mc Fadyen1, Benoit Trottier1

1 International Training and Education Center for Health (I-TECH), 1 Clinique Médicale L’Actuel, Montréal, Québec, Canada University of Washington, Seattle, WA, USA 2 University of Washington, Seattle, WA, USA Background: Patients who seek care for post sexual-exposure pro- 3 International Training and Education Center for Health - Haiti phylaxis (sPEP) are considered at high risk of contracting HIV. (I-TECH), Port-au-Prince, Haiti Counselling and HIV testing during sPEP is an important component 4 International Training and Education Center for Health - Haiti of the care protocol. At l’Actuel we recommend patients come back

ABSTRACTS (I-TECH), Kenscoff, Haiti for counselling and HIV/STI screening at 4 weeks and 16 weeks after a consultation for sPEP. In the context of increasing the use of and Background: In Haiti, the Ministry of Health’s iSanté electronic med- scaling up the sPEP strategy, we aimed at assessing the patient’s ical record system can help clinicians monitor patients’ adherence to determinants of a good PEP follow-up (FU). HIV antiretroviral therapy (ART) via self-reported or pharmacy-based measures. This analysis identifies which ART adherence measure is Methods: From 2000 to 2012, we prospectively enrolled patients con- most useful in predicting ART failure, identifies other patient-level sulting for sPEP in a single site cohort study. Our outcome was predictors of ART failure, and uses these results to design a simple adherence to week 16 FU visit. Factors associated with adherence to ORAL risk classification algorithm. FU-w16 were identified using backward stepwise logistic regression analyses by SPSS 17.0. Methods: The study analyzed data on 2,510 adult patients enrolled on ART from 2005-2013 at two large, public-sector hospitals in Haiti. ART Results: A total of 2,287 sPEP consultations were included. Patients failure was assessed based on immunologic and clinical criteria. consulting for sPEP were mostly MSM (80%), with a median age of 32 Performance of candidate prediction models was compared using years, with consultation after receptive anal intercourse in 38% of area under the receiver operating curve, and results were validated cases. Source person was known HIV+ in 26% of cases and risk of HIV using a randomly-split data sample. The selected prediction model acquisition as moderate or high in 81% of cases. In 1750 (78%) of them was used to generate a weighted risk score and risk classification, treatment was advised, mainly by CBV/LPV (N = 243), TVD/LPV (N = and its ability to differentiate ART failure risk over a 36-month treat- 1,248) or TVD/RAL (N = 51). Adherence to treatment was 1,206/1,750 ment period was tested using stratified Kaplan-Meier survival curves. (69%, ITT) varying based on the treatment selected. Globally, 70% of patients came back for their FU visit at w4 and 50% for the w16 visit. Results: Among 923 patients with data available on ART failure, 196 (21.2%) met failure criteria. The pharmacy-based proportion of days MULTIPLE LOGISTICAL REGRESSION ANALYSES FOR THE covered (PDC) measure performed best among five ART adherence DETERMINANTS OF FU-W16 VISIT. measures. Average PDC during the first 6 months on ART was 79.0% Crude OR Adjusted OR among ART failures and 88.6% among non-failures (p <0.01). Self- Independent variables (95% CI) (95% CI) reported adherence measures were not helpful in predicting ART Gender (Women vs. Men) 0.72 (0.47 - 1.09) -- failure and were frequently unavailable within iSanté. Beyond PDC, Age 1.04 (1.02 - 1.05) 1.02 (1.01 - 1.04) the variables sex, baseline CD4 count, and duration of enrollment in Risk Evaluation by Physician: Moderate vs. Low 1.47 (0.57 - 3.79) -- HIV care prior to ART initiation enriched the prediction model. The High vs. Low 1.36 (0.52 - 3.51) -- predicted risk groups strongly differentiated observed risk of ART Nb of episode failure within the first 12 months as well as over 36 months. (1st vs. subsequent PEP) 1.22 (0.92 - 1.61) -- Received ARV as Prophylaxis 2.22 (0.89 - 5.51) 1.81 (1.32 - 2.47) Conclusions: Pharmacy data are most useful for new ART adherence Regimen: alerts within iSanté. These alerts have potential to help clinicians CBV/LPV vs. TVD/LPV 1.52 (1.15 - 2.01) -- identify patients at high risk of ART failure so that they can be target- TVD/RAL vs. TVD/LPV 0.98 (0.56 - 1.71) -- ed with adherence support interventions, before ART failure occurs. Other vs. TVD/LPV 0.27 (0.95 - 1.71) -- Came to 4 weeks Follow up 6.96 (5.06 - 9.56) 3.74 (2.82 - 4.96) Was adherent to 4 weeks Treatment 4.07 (3.28 - 5.07) 1.87 (1.42 - 2.45)

Conclusions: Counselling and testing are an integral part of the PEP protocol. It is reassuring to see that this study demonstrates that patients who had high-risk behavior and needed treatment are also those who come back for their HIV testing and counseling at follow-up visits. However, additional effort is needed to enhance adherence to follow-up visits in all patients consulting for post-sexual exposition prophylaxis for HIV.

30 9th International Conference on HIV Treatment and Prevention Adherence Real Time Plasma TFV Levels to Roles of Medication Responsibility, 350 Support Adherence in a Pre- 351Executive and Adaptive Functioning Exposure Prophylaxis Demonstration Project in Adherence for Youth with Perinatal HIV

K Rivet Amico1, Christina Psaros2, Steve Safren3, Patricia Garvie1 (presenting), Susannah Allison2, Sean Brummel3, Ryan Kofron4 (presenting), Risa Flynn5, Robert Bolan5, Kathleen Malee4, Megan Wilkins5, Lynnette Harris6, Wilbert Jordan6, Kenneth Mayer7, Peter Anderson8, Miriam Chernoff3, Sharon Nichols7 Lane Bushman9, Rhodri Dierst-Davies10, Amy Rock-Wohl10, Mario Perez11, Keith Rawlings11, Raphael Landovitz4 1 Children’s Diagnostic & Treatment Center, Fort Lauderdale, FL, USA 1 University of Connecticut, Brighton, MI, USA 2 National Institute of Mental Health, Bethesda, MD, USA 2 Harvard Medical School, Boston, MA, USA 3 Harvard School of Public Health, Boston, MA, USA 3 Massachusetts General Hospital, Boston, MA, USA 4 Northwest University Feinberg School of Medicine, 4 UCLA CARE Center, Los Angeles, CA, USA Chicago, IL, USA

5 Los Angeles Gay and Lesbian Center, Los Angeles, CA, USA 5 St. Jude Children’s Research Hospital, Memphis, TN, USA ABSTRACTS 6 OASIS Clinic, Los Angeles, CA, USA 6 Baylor College of Medicine, Houston, TX, USA 7 Harvard University, Boston, MA, USA 7 University of California San Diego, La Jolla, CA, USA 8 University of Colorado Denver, Aurora, CO, USA 9 University of Colorado, Denver, CO, USA Background: Assuming responsibility for medication adherence is a 10 Los Angeles County Department of Public Health, Los Angeles, critical but challenging developmental task for youth with perinatally CA, USA acquired HIV (PHIV). To help prepare youth for transition to adult- 11 Gilead Sciences, Foster City, CA, USA hood, it is essential to understand how medication responsibility, executive (EF) and adaptive functioning (AF) contribute to adherence ORAL Introduction: Pre-exposure prophylaxis (PrEP) has great potential to and readiness for successful medication management among PHIV reduce HIV incidence with adequate adherence. Optimal interven- youth prior to adulthood. EF has predicted adherence in adults and tions for promoting daily PrEP adherence have yet to be defined. may, along with AF, inform adherence interventions for PHIV youth. Randomized placebo-controlled efficacy data from the Partners PrEP study suggest that adherence may be tri-modal, with relative distri- Methods: PHIV youth aged 7-16 years enrolled in the Pediatric butions varying by population: some individuals adhere well, some HIV/AIDS Cohort Study Adolescent Master Protocol, who were on adhere poorly, and some adhere episodically. This distribution sug- antiretroviral medication, with completed measures of responsibility gests that identifying and targeting poor and episodic adherers is for adherence, 7-day caregiver- and youth-reported adherence, likely to prove more efficient than a “one-size-fits-all” approach. Adaptive Behavior Assessment System, 2nd Edition, General Adaptive Composite (GAC), Behavior Rating Inventory of Executive Description: We designed a PrEP demonstration study that uses Function, General Executive Composite (GEC), in addition to demo- real-time PrEP drug (plasma tenofovir; TFV) concentrations to deploy graphic and health characteristics, were evaluated. a stepped-care adherence intervention. In step 1, all receive integrated “Next Step Counseling (iNSC)” at study visits (baseline and Results: 256 PHIV youth (mean age 12 years), primarily Black (76%) weeks 4, 8, 12, 24, 36, and 48). Plasma TFV concentrations collected and/or Hispanic (21%), and 49% male were included in analyses. Per at each visit are used to identify below the limit of quantitation (BLQ, 7-day recall, 72% were adherent (no missed doses). Per youth self- <10ng/mL) and trigger step 2, “Targeted iNSC,” a more detailed dis- report, 22% had sole medication responsibility with most reporting cussion that aims to develop strategies a) to support adherence and shared caregiver-youth medication responsibility (55%). Adjusted b) to maintain sexual health in the face of non-protective TFV levels. logistic models showed significantly higher odds of adherence for A repeat BLQ result prompts escalation to step 3, “PrEP-STEPS,” a caregiver vs. child solely responsible for medication (odds ratio (OR) six-session intervention based on principles of cognitive behavioral = 4.23, confidence interval (CI)[1.5,12.3], p = 0.008), and nadir CD4% therapy and the ART-supporting LIFE-STEPS intervention for HIV- <15% vs. >15% (OR = 2.24, CI[1.1,4.4], p = 0.019), adjusting for demo- positive individuals. graphic variables (age, race, caregiver education). No significant association of GEC or GAC was found with medication responsibility Lessons Learned: As of 1/15/14, 52 participants were enrolled and or adherence. only 1 BLQ assay was observed, which occurred at a week 4 study visit. While no participant has yet reached week 36, assays collected Conclusions: Study results suggest that among PHIV youth, contin- at week 8 (n = 28), 12 (n = 20) and 24 (n = 3) were all above limits of ued caregiver involvement in medication management during ado- quantification. Dried blood spots (DBS) will also be used to evaluate lescence is essential, despite youths’ growing autonomy. Global rat- gradients of adherence using intraerythrocytic TFV-DP levels. ings of EF and AF were not significantly associated with medication adherence. Given that EF and AF continue to develop throughout Recommendations: Real-time drug-level monitoring is feasible to adolescence, relationships with adherence should be evaluated lon- support adherence in PrEP demonstration projects and potentially in gitudinally, especially as youth transition to adulthood and shared clinical practice. Point-of-care drug level assays would facilitate caregiver responsibility diminishes. monitoring adherence when deploying enhanced adherence support.

9th International Conference on HIV Treatment and Prevention Adherence 31 Robust Short-Term Effectiveness of FEM-PrEP: Participant 363 a Comprehensive HIV Care 366 Explanations for Study Product Coordination Program in New York City Adherence

Mary Irvine1, Stephanie Chamberlin2 (presenting), Amy Corneli1 (presenting), Brian Perry1, Khatija Ahmed2, Rebekkah Robbins1, Julie Myers1, Graham Harriman1, Kawango Agot3, Kevin McKenna1, Fulufhelo Malamatsho2, Sarah Braunstein1, Beau Mitts1, Sarah Gorrell Kulkarni3, Joseph Skhosana2, Jacob Odhiambo3, Lut Van Damme4 Denis Nash3 1 FHI 360, Durham, NC, USA 1 New York City Department of Health & Mental Hygiene, New 2 Setshaba Research Centre, Pretoria, South Africa York, NY, USA 3 Impact Research and Development Organization, Kisumu, Kenya 2 New York City Department of Health & Mental Hygiene, Queens, 4 Bill & Melinda Gates Foundation, Seattle, WA, USA NY, USA 3 CUNY School of Public Health, New York, NY, USA Background: FEM-PrEP was a clinical trial of daily emtricitabine/

ABSTRACTS tenofovir disoproxil fumarate (FTC/TDF) for HIV prevention among Background: Evidence is needed regarding multi-component HIV women in sub-Saharan Africa. Study product adherence was too low interventions that improve engagement in care (EiC) and viral load to determine the effectiveness of FTC/TDF. However, a minority of suppression (VLS) across affected subpopulations. We assessed participants had evidence of recent pill use in drug concentration subgroup differences in EiC and VLS change following enrollment analyses. We conducted a follow-up study to identify facilitators of into a comprehensive medical case management intervention, the adherence among these participants in two sites. NYC Ryan White HIV Care Coordination Program (CCP). Methods: We conducted qualitative, semi-structured interviews with ORAL Methods: Using the local program reporting system and CD4 and 88 purposefully-sampled participants assigned FTC/TDF. To explore viral load (VL) records from a longitudinal, population-based HIV facilitators, we categorized a sub-set of participants into two adher- Surveillance Registry, we examined pre- and post-enrollment out- ence groups, “moderate” (n = 31) and “good” (n = 25), based on drug comes for 3,176 CCP clients enrolled by March 31, 2011 and diag- concentrations from samples collected at multiple times during FEM- nosed >1 year before enrollment. For the 12-month periods before PrEP. Participants viewed a graph displaying their adherence over and after enrollment, we estimated EiC (defined as ³2 tests ³90 days time and were asked about their adherence patterns, including fac- apart, with ³1 in each half of the 12-month period) and VLS (defined tors making adherence easy (“good” group) and reasons for adher- as VL ²200 copies/mL on latest VL test in the second half of the 12- ing some of the time (“moderate” group). We also asked participants month period). Relative risks (RRs) and confidence intervals (CIs) for about their perceptions of other participants’ adherence. We used the outcomes were estimated using generalized estimating equa- thematic analysis to analyze the data. tions. Results: Preliminary results suggest that many participants in the Results: The proportions with EiC and VLS increased from 74% to “good” group were motivated to adhere because of perceived HIV 91% (RREiC = 1.24, 95% CI: 1.21-1.27) and from 32% to 51% (RRVLS = risk and interest in learning if FTC/TDF can prevent HIV; several also 1.58, 95% CI: 1.50-1.66), respectively. Significant improvements held said partners reminded them. In the “moderate” group, many were across subgroups, except clients with baseline CD4 ³500 (VLS only, motivated to adhere after adherence counseling or enrollment, RRVLS = 1.03, 95% CI: 0.97-1.10) or “other/unknown” race (EiC only, although interest after enrollment waned over time; many also men- RREiC = 1.11, 95% CI: 0.99-1.23). By client characteristics at enroll- tioned perceived risk. When describing factors that influenced other ment, the greatest improvements were among those under age 45, participants’ adherence, numerous participants said coitally- diagnosed after 2004, not on antiretrovirals, born male (EiC only), dependent adherence was common. making <$9,000/year (EiC only), uninsured (EiC only), homeless (EiC only), unsuppressed (EiC only), and having CD4 <200 (VLS only). Conclusions: Some trial participants may need additional support Significant improvements were observed for EiC at 25 (89%) and VLS beyond adherence counseling to sustain their initial interest in using at 21 (75%) of 28 CCP agencies. a daily investigational product throughout trial implementation. Although perceived risk appeared to facilitate adherence, the con- Conclusions: Short-term EiC and VLS improvements were robust cept cannot be ethically used to promote adherence within trials across most subgroups examined. Given the cost/complexity of comparing an investigational drug with a placebo. Future research comprehensive care coordination, differences found suggest the could explore the role of risk perceptions and partner support in value of targeting recruitment to those with greater need of support FTC/TDF demonstration projects. to succeed in treatment.

32 9th International Conference on HIV Treatment and Prevention Adherence Real-Time Antiretroviral Treatment Impact of a Self-Management 369 Monitoring among HIV-Positive 370 Telephone Support Program for Individuals in Southern China: Early Older People Living with HIV on Experiences with ‘Wisepill’ Antiretroviral Adherence and Quality of Life

Lora Sabin1 (presenting), Mary Bachman DeSilva2, Sheryl Catz1 (presenting), Benjamin Balderson1 Christopher Gill1, Li Zhong3, Taryn Vian1, Wubin Xie4, Feng Cheng5, Jessica Haberer6, David Bangsberg6, Allen Gifford7 1 Group Health Research Institute, Seattle, WA, USA

1 Boston University School of Public Health, Boston, MA, USA Background: People living with HIV (PLWH) aged 50 and older have 2 Center for Global Health and Development, Boston University, increased rates of chronic health conditions and polypharmacy, and Boston, MA, USA unique social and health related quality of life (QOL) concerns. 3 FHI360, Beijing, China However, HIV support programs have not traditionally targeted this 4 FHI360, Nanning, Guangxi, China rapidly growing age group. ABSTRACTS 5 Global Health Strategies, Beijing, China 6 Massachusetts General Hospital, Boston, MA, USA Methods: The PRIME randomized controlled trial evaluated the effi- 7 Veterans Health Administration, Bedford, MA, USA cacy of a telephone-delivered individual self-management intervention for older PLWH. 452 PLWH aged 50 and older, currently pre- Background: The ‘China Adherence through Technology Study’ scribed ART, and reporting adherence lapses in the past 30 days (CATS) is assessing use of ‘Wisepill,’ web-linked medication contain- were recruited from AIDS service organizations in 9 cities, and ran- ers that allow real-time adherence tracking, among patients on anti- domized to one of three interventions (Individual, Group, Information)

retroviral therapy (ART) in China. Although Wisepill has been used to after completing baseline telephone surveys. All participants ORAL monitor medication adherence previously, its potential as an adher- received a book on living well with HIV. Individual intervention partic- ence support tool remains unknown. CATS will evaluate use of ipants also received up to 10 30-minute telephone calls that integrat- Wisepill-generated real-time reminders and adherence-data sup- ed chronic disease and HIV self-management skills training with ported counseling. One study component examined technical feasi- motivational and problem-solving counseling. Group participants bility and patient acceptability of Wisepill. received access to10 time-matched self-management support group calls. Information participants received a book only. Methods: Patients attending an ART clinic in Nanning, China, were provided Wisepill for 1 ART medication. After monitoring subjects’ Results: Individual telephone counseling significantly improved com- adherence for three months, we randomized them to intervention posite ART adherence scores as compared to Group and Information (reminders and counseling) or control (usual care). Socio-demo- (book-only) comparison arms, and these differences were main- graphic and self-reported adherence data were collected at base- tained at 12-month follow-up in intent-to-treat (ITT) analyses. At 6 line. After three months, we collected quantitative and qualitative months, ITT analyses showed the Individual arm had significantly data on patients’ experiences with Wisepill. Signal lapses of 48 hours higher Social Functioning than Group or Information controls, signif- or more were investigated during this 3-month period. icantly higher Mental Health Functioning than Information controls, and no differences in Physical Functioning. Mental Health Results: 120 subjects were enrolled; all but one completed the pre- Functioning for Individual intervention participants remained higher intervention period. Two-thirds (65.3%) were male; mean age was than Information controls at 9 months, but there were no differences 37.5 years. 78.6% had completed middle school and 49.0% were mar- in SF-36 subscales by 12 months. ried. 13.3% reported experience with injection drug use. Mean adherence was 91.7% (SD 12.4) measured by Wisepill, but 95.6% (SD Conclusions: PRIME trial results suggest that individual telephone 17.2) by self-report. 93.2% of subjects reported positive or very posi- counseling that is wellness-focused and recognizes the unique tive overall experience with Wisepill; 57.6% found it “very easy” to needs and comorbidities of the aging HIV population represents a use. However, 56.0% said Wisepill was inconvenient or very incon- promising intervention model for supporting self-management and venient, supported by statements that it was large and conspicuous. maintaining treatment adherence among older PLWH. 55.9% were worried that using Wisepill might disclose their HIV status, though no disclosures occurred. Over these 3 months, minor technical difficulties were encountered and addressed.

Conclusions: ART patients in China are generally positive about using a real-time, web-linked adherence monitoring device; Wisepill is also feasible. Concerns about convenience and potential stigma need further exploration. Wisepill holds potential for interventions that provide rapid adherence feedback directly to patients.

9th International Conference on HIV Treatment and Prevention Adherence 33 Active Methamphetamine Use is A Qualitative Investigation of 371 Associated with Antiretroviral 372 Patients Transitional States of Medication Non-Adherence Engagement in HIV-Related Medical Care

David Moore1 (presenting), Alexandra Rooney1, Jessica Montoya1, Stephanie Koch1 (presenting), Mary Gerkovich1, Karen Williams1, Kaitlin Casaletto1, Ben Gouaux1, Colin Depp1, J. Hampton Atkinson2 Julie Banderas1

1 University of California, San Diego, CA, USA 1 UMKC School of Medicine, Kansas City, MO, USA 2 VA San Diego/University of California, San Diego, CA, USA Background: Engagement in care following an HIV positive diagnosis Background: Methamphetamine use is associated with worse carries essential clinical and public health implications, including adherence to antiretroviral therapy (ART) and poorer HIV disease better management of the disease, minimization of viral load, mainte- outcomes. The present study assessed the association between nance of a healthy CD4 count, and transmission prevention. This daily texting responses of methamphetamine use and electronically- qualitative study was designed to investigate factors related to

ABSTRACTS monitored ART adherence among HIV-infected (HIV+) persons. engagement in HIV-related medical care with emphasis on transitions in and out of care. Methods: We analyzed data from 50 HIV+ persons who self-reported using methamphetamine within the last 30 days at the time of study Methods: Data were collected through person-level interviews fol- enrollment. Participants were enrolled in an ongoing randomized text lowed by thematic coding and analysis. The sample consisted of 22 messaging intervention designed to improve ART adherence over a 6 participants recruited from safety-net providers in an urban setting. week period. All participants received daily text messages assessing whether they had used methamphetamine in the last 24 hours. A sin- Results: Six main themes emerged from the data and were identified ORAL gle “sentinel” antiretroviral medication was tracked daily using an as factors related to engagement: health, treatment, personal, sup- electronic monitoring system and the proportion of correctly taken port, clinic, and resources, with each of these containing subthemes doses over the six weeks was used as the measure of adherence. that describe the issues more specifically. Results were analyzed in context of how they relate to phases of engagement on the HIV con- Results: There was a trend-level association between frequency of tinuum of care, with specific examples from the interviewees that reported methamphetamine use and ART adherence, such that characterize the transitional states of interest (e.g., engaging vs. dis- greater proportions of methamphetamine-using days were associat- engaging in care, being in vs. out of care). ed with worse mean ART adherence (␳ = -0.27, p = 0.05). Participants who reported methamphetamine abstinence during the entire six- Conclusions: Factors that enable or hinder care engagement vary week study period (n=9) had significantly better mean ART adher- among HIV-positive persons; the data demonstrate that a deterrent ence (mean adherence = 89.9%, SD = 13.3) as compared to those who for one patient may act as a motivator for another. The relative impor- reported some methamphetamine use (n = 41, mean adherence = tance of the six factors varied between individuals and over time. To 63.6%, SD = 27.3; p <0.001). Interestingly, using a matched-pairs t- reflect this complexity, factors related to engagement should be indi- test, there was no significant association between day-specific vidualized, based on evaluation upon care entry and during each methamphetamine use (yes/no) and day-specific ART adherence subsequent contact. Findings support an existing framework and (yes/no) (p >0.05). indicate the need for evaluation of funding priorities, as well as changes at the health care system and policy levels. Our conceptual Conclusions: Any active methamphetamine use, regardless of fre- framework addresses engagement transitions and can help quency, appears to confer risk of ART nonadherence, although there providers, case managers, clinic staff, and administrators under- was a trend toward increased methamphetamine use and worse stand the complexity of factors to be accounted for when developing adherence. The relationship between adherence and methampheta- efforts to support better engagement in care. mine use appears to be more complex than a simple one-to-one relationship (i.e., methamphetamine use on any given day does not specifically relate to nonadherence on that same day). Interventions designed to both lessen substance use behaviors, and improve ART adherence in the context of active methamphetamine use, continue to be needed.

34 9th International Conference on HIV Treatment and Prevention Adherence Patterns and Correlates of PrEP Factors Associated with Loss to 374 Drug Detection among MSM and 375 Clinic in a Large HIV Care Center of Transgender Women in the Global iPrEx Study New York City Albert Liu1 (presenting), David Glidden2, Peter Anderson3, Rituparna Pati1 (presenting), Christopher Beattie2, Joyce Park3, K Rivet Amico4, Vanessa McMahan5, Megha Mehrotra5, Denis Nash4, Victoria Sharp3 Javier Lama6, John MacRae7, Juan Carlos8, Orlando Montoya9, Valdilea Veloso10, Mauro Schechter11, Esper Kallas12, 1 Mount Sinai St Luke’s and Roosevelt Hospital, New York, NY, USA Suwat Chariyalerstak13, Linda-Gail Bekker14, Kenneth Mayer15, 2 Spencer Cox Center for Health - St. Luke’s-Roosevelt Hospital, Susan Buchbinder16, Robert Grant5 New York, NY, USA 3 1 Bridge HIV, San Francisco, CA, USA St Luke’s Roosevelt Hospital, New York, NY, USA 4 2 University of California, San Francisco, CA, USA CUNY School of Public Health, New York, NY, USA 3 University of Colorado Denver, Aurora, CO, USA 4 University of Connecticut, Brighton, MI, USA Background: Clinic-based electronic medical record data are under- 5 Gladstone Institutes, University of California, San Francisco, CA, USA utilized but uniquely informative data sources for development of tar- ABSTRACTS 6 Asociación Civil Impacta Salud y Educación, Lima, Peru geted strategies to improve suboptimal patient retention in HIV pri- 7 Investigaciones Médicas en Salud, Lima, Peru mary care. 8 Asociación Civil Selva Amazónica, Iquitos, Peru 9 Fundación Ecuatoriana Equidad, Quito, Ecuador Methods: We examined data on HIV-infected patients seen in New 10 Instituto de Pesquisa Clinica Evandro Chagas - IPEC/FIOCRUZ, York City’s Spencer Cox Center for Health for at least one HIV primary Rio de Janeiro, Brazil care visit in 2010 or 2011. Retention was defined as returning for at 11 least one medical visit between January 1, 2012 and September 30,

Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil ORAL 12 Universidade de São Paulo, São Paulo, Brazil 2012; lost to clinic (LTC) was defined as no medical visit during that 13 Research Institute for Health Sciences, Chiang Mai University, period. Multivariate regression models were used to estimate adjust- Chiang Mai, Thailand ed odds ratios (AOR) and 95% confidence intervals (CI) of factors 14 The Desmond Tutu HIV Foundation, Cape Town, South Africa associated with LTC. 15 Fenway Community Health, Brown University, Boston, MA, USA 16 San Francisco Department of Public Health, San Francisco, CA, USA Results: 5,750 HIV-infected patients were seen between January 1, 2010 and December 31, 2011 (median visits = 8, IQR:4-13). 4415 (76.8%) Background: Adherence to HIV pre-exposure prophylaxis (PrEP) is were retained in care and 1335 (23.2%) were LTC. Among those LTC, critical for efficacy. Antiretroviral drug concentrations are an objective median CD4 count at last visit was 338 cells/µL(IQR:169-528) and viral measure of PrEP use and correlate with efficacy. Understanding pat- load suppression was 31%. Factors associated with LTC include terns and correlates of drug detection can identify populations at risk insurance coverage (AOR commercial vs. Medicaid:0.7, CI:0.5-0.9), for non-adherence and inform design of PrEP adherence interventions. housing status (AOR unstable vs. stable:1.6, CI:1.4-1.9), incarceration history (AOR some vs. none: 2.0, CI:1.6-2.4), and clinic site (AOR mid- Methodology: Blood antiretroviral concentrations were assessed town vs. downtown: 3.5, CI:2.8-4.4). Mental health diagnosis, higher among active-arm participants in iPrEx, a randomized, placebo-con- CD4 count, suppressed viral load, and antiretroviral therapy (ART) trolled efficacy trial of emtricitabine/tenofovir in men who have sex were independently associated with lower risks of LTC. with men (MSM) and transgender women in 6 countries. We evaluated rates and correlates of drug detection among a random sample Conclusions: Patients who are underinsured, unstably housed, previ- of 470 participants at week 8 and a longitudinal cohort of 303 partici- ously incarcerated, and not on ART appear to be at highest risk of pants through 72 weeks of follow-up using logistic regression. LTC. Those with lower CD4 counts and high viral loads are also more Results: Overall, 55% (95% CI 49-60%) of participants tested at week 8 likely to be LTC, suggesting that strategies to reduce LTC could lead had drug detected. Drug detection was associated with older age and to improvements in both individual and public health outcomes. higher education and varied significantly by study site, with detection Further evaluation of the impact of mental health care and housing rates lowest in samples from Lima, Peru (35%) and highest in San support integrated with outpatient HIV care is warranted. Future Francisco, USA (90%). In longitudinal analysis, 31% never had drug studies would benefit from the ability to ascertain care status and detected, 30% always had drug detected, and 39% had an inconsistent vital status of those LTC. detection pattern. Overall drug detection rates declined over time. Drug detection at some or all visits was associated with older age, indices of This abstract will also be presented as a poster. risk behavior, and responding “don’t know” to a question about belief of PrEP efficacy. Among participants who had evidence of early drug detection, approximately one-third showed a pattern of non-persistence (detection stopped and remained undetectable at subsequent visits).

Conclusions: Distinct patterns of study-product use were identified, with a significant proportion demonstrating no evidence of initiating study-drug in iPrEx. Research literacy may explain greater drug detection among populations having greater research experience, such as older MSM in the United States. Greater drug detection among those reporting highest risk sexual practices is expected to increase the impact and cost-effectiveness of PrEP.

9th International Conference on HIV Treatment and Prevention Adherence 35 Implementation of PrEP in STD and Symptomatic HIV Disease or 377 Community Health Clinics: High 379 Perceived Side Effects of Uptake and Drug Concentrations among Medication Associated with Lower MSM in the Demo Project Antiretroviral Adherence among MACH14 Patients Stephanie Cohen1, Eric Vittinghoff2, Peter Anderson3, Susanne Doblecki-Lewis4, Oliver Bacon2, Wairimu Chege5, Yan Wang1 (presenting), Ira Wilson2, Marc Rosen3, Loren Miller1, Richard Elion6, Susan Buchbinder1, Michael Kolber4, Carol Golin4, Judith Erlen5, Jie Shen1, Jane Simoni6, Albert Liu7 (presenting) Robert Remien7, Honghu Liu8 1 San Francisco Department of Public Health, 1 University of California, Los Angeles, CA, USA San Francisco, CA, USA 2 Brown University, Providence, RI, USA 2 University of California, San Francisco, CA, USA 3 Yale University School of Medicine, New Haven, CT, USA 3 University of Colorado Denver, Aurora, CO, USA ABSTRACTS 4 University of North Carolina, Chapel Hill, NC, USA 4 University of Miami, Miami, FL, USA 5 University of Pittsburgh, Pittsburgh, PA, USA 5 National Institutes of Health, Bethesda, MD, USA 6 University of Washington, Seattle, WA, USA 6 Whitman Walker Health, Washington, DC, USA 7 HIV Center for Clinical and Behavioral Studies, 7 Bridge HIV, San Francisco, CA, USA New York, NY, USA 8 UCLA School of Dentistry, Los Angeles, CA, USA Background: Pre-exposure prophylaxis (PrEP) has been shown to be safe and efficacious in clinical trials. Demand for PrEP and adher- Background: Symptomatic HIV disease and medication side effects ORAL ence in real-world settings are unknown. We evaluated PrEP uptake are implicated in lower antiretroviral (ARV) medication adherence. and drug concentrations among men who have sex with men (MSM) However, which of common HIV symptoms and medication side in the US PrEP Demonstration (Demo) Project. effects are associated with lower medication adherence is unclear. Methodology: HIV-uninfected MSM and transgender women attend- Methods: MACH14 is collaboration of 14 sites that assessed anti- ing STD clinics in San Francisco (SF) and Miami, and a community retroviral therapy (ART) adherence using medication event monitor- health center in Washington, DC, were offered screening for The ing system (MEMS) electronic bottle caps. Ten studies collected per- Demo Project. Clients were approached in clinic while receiving ceived HIV- or ARV-related physical symptomatology data from 1997 services or self-referred to the program. Enrolled participants were to 2009, seven collected data on symptom severity, and four studies offered 48 weeks of open-label emtricitabine/tenofovir. Predictors of collected data that patients differentiated side effects from sympto- PrEP uptake were assessed using multivariable logistic regression. matic HIV disease. The association of HIV related symptoms or side Tenofovir-diphosphate (TFV-DP) levels in dried blood spots (DBS) effects of ARVs on adherence was analyzed through random inter- were measured to estimate average adherence in a random sample cept and slope (RIAS) models with repeated measures of adherence, at week 4. viral load and regimens clustered within studies. Results: From 9/2012-1/2014, 1097 clients were assessed for partici- Results: The RIAS model, adjusting for age, gender, ethnicity, educa- pation, 408 declined, 142 were ineligible, and 547 enrolled. 34% of tion, viral load, and current regimens, indicated that lower adherence clients assessed were self-referred. PrEP uptake varied among sites was associated with higher number of symptoms (p = 0.041) and (52% in SF, 61% in DC and 67% in Miami). Mean age of enrolled par- more severe symptoms (p = 0.001). Among the 20 self-reported symp- ticipants was 35; 10% were Black, 32% Latino, and 48% white; 74% toms, lower adherence was significantly associated with fever (p = had previously heard of PrEP. In adjusted analyses, participants from 0.003), weight loss (p = 0.04) and nausea (p = 0.048). Among the Miami or DC, those who were self-referred, and those reporting high- patients who reported the symptoms are not because of the side er indices of sexual risk were more likely to enroll. DBS from 136 par- effects of the medication, fatigue (p = 0.001), insomnia (p = 0.002), ticipants were tested: almost all (98%) had TFV-DP detected. Median cough (p = 0.02), appetite (p = 0.02) and fever (p = 0.04) were still sig- TFV-DP levels were higher in SF than in Miami and DC (975, 658 and nificantly associated with lower adherence, adjust for current regi- 812 fmol/punch respectively, p <0.001). Most participants (77%) had a mens and the viral load. TFV-DP level consistent with taking at least 4 doses/week (³550 fmol/punch) (92% SF, 57% Miami, 78% DC, p <0.001). Conclusions: Although there was no objective determination for Conclusions: Demand for PrEP and drug concentrations at week 4 symptoms by physicians in MACH14 studies, we found that among are high among MSM in the Demo Project. Factors contributing to HIV-infected persons, several symptoms including fever, weight loss differences in PrEP uptake and adherence across sites warrant fur- and nausea are associated with lower adherence. Among those who ther investigation. reported their symptoms are not related to their regimens, fatigue, insomnia, cough, appetite and fever are associated with lower adherence. Further research is warranted to identify the source of symptoms so interventions can target patients with symptoms caused by medications to improve adherence.

36 9th International Conference on HIV Treatment and Prevention Adherence Linking Key Populations to HIV Daily Short Message Service 381Care Services in Kampala, Uganda 382 Surveys Detect Greater HIV Risk Behavior than Monthly Clinic Stephen Alege1 (presenting) Questionnaires in Kenya 1 Makerere University, Kampala, Uganda Kathryn Curran1 (presenting), Nelly Mugo2, Ann Kurth3, Introduction: In Uganda, there is high incidence of stigma against Kenneth Ngure4, Renee Heffron5, Deborah Donnell6, Connie Celum5, key populations such as sex workers and men who have sex with Jared Baeten5 men (MSM). This situation is exacerbated by their illegal status, which makes it difficult to design and implement HIV friendly servic- 1 World Health Organization, Geneva, Switzerland es that meet their needs. As such the community and self-stigma 2 KEMRI, Nairobi, Kenya inhibits them from seeking HIV/AIDS services. 3 New York University, New York, NY, USA 4 Jomo Kenyatta University of Agriculture and Technology, Description: Uganda Health Marketing Group (UHMG) is implement- Nairobi, Kenya ABSTRACTS ing a 3-year project to provide HIV counseling and testing (HCT) serv- 5 University of Washington, Seattle, WA, USA ices and link HIV-positive key populations to HIV Care services. The 6 Fred Hutchinson Cancer Research Center, Seattle, WA, USA project is implemented through 56 private health clinics and 2 mobile HCT vans in Kampala. Clinics were supported to contract fulltime Background: Technology-based approaches offer novel opportuni- dedicated linkage facilitators to link and follow up key population ties to record HIV risk behaviors in real-time, potentially reducing posttest clients who test HIV positive to HIV care facilities of their biases related to recall and social desirability that are inherent in choice. other methods of self-report. We conducted a pilot evaluation of daily

short message service (SMS, i.e., text message) surveys to measure ORAL Results: Partnerships have been established with 5 key population sexual behavior and pill taking in a study of pre-exposure prophylax- groups and associations. Over a period of 3 months, 520 sex workers is (PrEP) for HIV prevention among HIV-uninfected members of HIV and 92 MSM have been provided with HCT services. Of these, 122 serodiscordant couples in Kenya. In this analysis, we compared self- were referred and linked to HIV care services with a total of 103 reported sexual and pill-taking behaviors collected over daily SMS to receiving HIV care services. A total of 31 HIV-positive key population concurrent monthly questionnaires among HIV-uninfected Kenyan clients were followed up in the community to ensure that they are adults taking pre-exposure prophylaxis (PrEP). retained on the care services. Methods: Eighty-five participants contributed 145 observation- Lessons Learned: Key population peer groups and associations are months. SMS data were collected daily through an automated sur- critical in mobilization of their peers. Training of staff in private health vey; clinic-based questionnaires were completed monthly. We care facilities in providing appropriate services is important. assessed agreement between reporting of sexual activity, sex unpro- Dedicated linkage facilitators at clinics increases linkage success and tected by condoms, and missed PrEP doses. retention into care. Collaboration with HIV-related care services has led to immediate enrollment into care services at the HIV testing site. Results: The proportion of observation-months with reports of any sex (93.8% vs. 85.5%), any unprotected sex (35.2% vs. 15.9%), or any Recommendations: Involve key populations in delivering HCT servic- missed PrEP doses (54.5% vs. 13.1%) were significantly greater by es so as to increase uptake by clients. daily SMS data collection than monthly questionnaires, respectively (all p <0.0001). Participants reported a median of 9 sex acts per month through daily SMS surveys and a median of 5 sex acts through monthly interviewer-administered questionnaires. Participants reported a median of 2 more sex acts per month on the daily SMS survey than the monthly questionnaire. The mean number of missed PrEP doses reported was 0.2 for monthly questionnaires and 1.2 for daily SMS surveys.

Conclusion: Daily SMS surveys delivered to personal mobile phones elicited significantly greater reports of sex, unprotected sex, and missed PrEP doses compared to monthly questionnaires, likely providing more accurate self-reported HIV risk behavior through enhanced privacy and reduced recall period. SMS diaries offer a promising method to assess sensitive or repetitive health behaviors (such as medication-taking) in real time.

9th International Conference on HIV Treatment and Prevention Adherence 37 Factors Associated with Intervals PrEP Audio Visual Representation 383 between Women’s Visits to HIV 386 (PrEP REP) Project: Development Outpatient Clinics and Pilot of a PrEP Education Video

Alison Evans1, Caroline Sabin1, Vanessa Apea2, Sophie Jose1, K Rivet Amico1 (presenting), Christopher Balthazar2, Tom Coggia3, Teresa Hill1, Fiona Burns1 (presenting) Sybil Hosek3

1 University College London, London, England, United Kingdom 1 University of Connecticut, Brighton, MI, USA 2 Barts Health NHS Trust, London, England, United Kingdom 2 Stroger Hospital of Cook County, Chicago, IL, USA 3 Iconatomic, Los Angeles, CA, USA Background: Studies find that women with HIV are less likely than men to engage with care. As part of the REACH project, we exam- Background: Despite the availability of pre-exposure prophylaxis ined factors associated with intervals between women’s visits to out- (PrEP) in the United States for over a year, PrEP knowledge is report- patient HIV clinics to better understand their engagement in care. edly low among individuals in communities at elevated risk for HIV

ABSTRACTS infection. Relatedly, educational tools to improve PrEP knowledge Methods: We conducted a secondary analysis of UK Collaborative are lacking. The PrEP REP team developed the PrEP REP video to HIV Cohort (CHIC) data, including 12 years of data from adult patients address this gap, where actions of PrEP are animated in an engag- who made two or more visits to 15 UK HIV clinics (1 January 2000 to ing, brief media clip intended for a wide range of audiences. The 31 December 2011). As clinic visits were not always reliably cap- video was pilot tested to determine immediate impact on PrEP-relat- tured, CD4 counts, viral loads and/or haemoglobin measures were ed knowledge and beliefs. used as surrogate markers of attendance. We conducted qualitative interviews with 6 HIV clinicians about fac- Methods: The PrEP REP video was developed in collaboration with ORAL tors associated with time to next appointment. medical experts and community advisors. Young men who have sex with men (MSM) in Chicago who were not on PrEP were recruited to Results: The UK CHIC analysis included 31,784 adults; 28.1% were evaluate the video. Changes in PrEP knowledge, perceived knowl- women (n = 8,930). Women were more likely than men to not attend edge and beliefs were evaluated using pre- and immediate-post for a year or more (17.7% vs. 16.2%). Women who did not attend for a video surveys, using non-parametric McNemar test with binomial year or more were more likely to be white (20.4%) or black-other eth- distribution and single sample mean difference t-tests depending on nicity (20.5% vs. 17.2% for black African, 11.4% for Asian, 18.3% for item. other), exposed to HIV through injecting drug use (32.2% vs. 17.5% for heterosexual, 24.0% for blood recipient, 10.9% for mother-to-child) Results: 46 young MSM (average age 21) viewed the video between and younger (21.2% of under 30s vs. 10.6% of over 45s), all p <.001. November and December 2013. 48% reported having heard of PrEP The average time between women’s clinic visits was 86.6 days. prior to the study. All but one of the 16 knowledge items significantly Women who became pregnant during the study had shorter gaps improved post-viewing, with 12 of these answered correctly by 80% between visits during pregnancy (46.8 days) than when they were not or more of the sample (none of the items reached 80% correct in pre- pregnant (95.0 days). The qualitative analysis highlighted the role of test). Rated efficacy of daily PrEP rose from 62% to 87% effective (on pregnancy in women’s engagement in care and the importance of average), as did perceived ability to take PrEP daily. Perceived comorbidities and psychosocial issues as key determinants of time to chances of getting infected with HIV did not significantly change, nor next appointment. did perceptions that PrEP would change condom use or number of partners. Perceptions that taking PrEP would promote mindfulness Conclusions: A range of factors contributes to positive engagement around sexual decision making significantly increased, as did overall in HIV care services and must be considered in the development of positivity towards PrEP adherence. any successful intervention. Conclusions: The video significantly increased PrEP knowledge and supported positive feelings towards PrEP and daily PrEP adherence among PrEP non-experienced at-risk young MSM.

38 9th International Conference on HIV Treatment and Prevention Adherence 387 Jeremiah MulambaBazaale Conclusions: Results: sized bytheme. Transcripts wereuploadedintoanalyticsoftware,codedandsynthe- guides wereusedfordatacollectionwithopen-endedquestions. patients, peers,andstaff(totaln=75).Semi-structuredinterview included 41in-depthinterviewsand6focusgroupdiscussionswith Methods: Background: 4 3 2 Maria Wawer April Monroe Uganda Evaluation ofaRandomizedTrial inRakai, Yet onAntiretroviralTherapy:AQualitative Services amongHIV-Infected PersonsNot engagement incareamongPLHIVwhoarenotyetonART. highlight thevalueofs-IMBmodelsforunderstandingandimproving scale-up oftheinterventionaspartintegratedcareservicesand ceptions ofthepeerintervention.Thesefindingslendsupportto skills model(s-IMB)ofengagementincare. tion wasbasedonasituated-InformationMotivationBehavioral ing PLHIVwhowerenotyetonARTinRakai,Uganda.Theinterven- ized controlledtrial(RCT)oftheimpactapeerinterventiontarget (ART). W tion andretentionamongpeoplenotyetonantiretroviraltherapy needed toassesstheeffectivenessofpeersinimprovingcareinitia- improve treatmentadherenceamongPLHIV, howeverevidenceis people livingwithHIV(PLHIV).Peersupporthasbeenproposedto 1 health staffandunderstandingofthesystem. tions withintheircommunities,andimprovedrelationshipswith including improvedknowledgeofhealthinformation,positivereputa conferred topatients,participantsreportedbenefitspeersreceived medication andkeepclinicappointments.Inadditiontothebenefits support networks,andidentifystrategiesforrememberingtotake Peers alsohelpedpatientstonavigatethehealthsystem,develop helped patientstobetterunderstandcomplicatedhealthinformation. Participants describedhowpeersreinforcedhealthmessagesand leading toincreasedengagementincareamongpatients. that peersimprovedinformation,motivation,andbehavioralskills, 9th InternationalConference on Johns HopkinsSchoolofMedicine,Baltimore,MD,USA National InstitutesofHealth,Kampala,Uganda Rakai HealthSciencesProgram,Kalisizo,Uganda Baltimore, MD,USA Johns HopkinsBloombergSchoolofPublicHealth, Participant narrativesparalleledthes-IMBmodel,indicating e conductedacomplimentaryqualitativestudyinrandom- The qualitativeevaluation(September 1 1 Engagement incareandtreatmentservicesbenefits This qualitativeevaluationfoundlargelypositiveper- Engagement inPreventionandCare Impact ofaPeerInterventionon , RonaldGray (presenting) , GertrudeNakigozi, 1 2 , StevenReynolds , WilliamDdaaki HIV T reatment andPrevention Adherence 2 , DavidSerwadda 3 , LarryChang -November 2013) 4 2 , - - 389 significant atp<0.05inthemultivariatemodel. and currentHIVriskbehaviors.We includedonlyvariablesthatwere demographic characteristics(agemodeled:<30;30-50;>50yearsold) ing HIVmedicalcareinthepast6months(self-report)andsocio- ate logisticregressiontoassesstheassociationbetweennotreceiv level ofengagementinHIVcare.W baseline interviewswereconductedwith924womenwhorangedin HIV prevalenceareasthroughouttheUnitedStates.From2010-2013, took placeat9community-basedorganizationsandclinicsinhigh Enhancing AccessandQualityHIVCareforWomen ofColorInitiative Niko ver 1 Methods: engagement andretentioninHIVcare. HIV careamongwomenenrollinginamulti-siteinitiativetopromote We soughttoidentifybaselinefactorsassociatedwithnotreceiving not retainedinHIVcare,littleisknownabouttheircharacteristics. Background: Oni Blackstock Initiative Enrolling inaHRSASPNSMulti-Site Conclusions: Results: future interventionsthatseekengageandretainwomeninHIVcare. ated withnotreceivingcarethatrepresentpotentialtargetsfor ing careinthepast6months.W mote engagementandretentioninHIVcare,halfreportednotreceiv- high-risk sexualbehavior(aOR=1.51,95%CI:1.08-2.11). to reportfair/poorhealth(aOR=1.58,95%CI:1.15-2.18)andcurrent 95%CI: 1.51-2.99),tobeuninsured(aOR=2.41,1.70-3.41),and years old;aOR=1.68,95%CI:1.01-2.79),liveinruralareas(aOR2.12, those notreceivingcareweremorelikelytobeyounger(<30vs.>50 care inthepast6months.ComparedtowomenreceivingHIVcare, stance use.Atbaseline,48.2%reportednotreceivingHIVmedical reported recenthigh-risksexualbehaviorand13.8%sub- 61.5% weresingle;41.3%hadnotcompletedhighschool;41.4% American/Black and27.1%Latina;30.5%residedinruralareas; Bronx, NY, USA Montefiore MedicalCenter/AlbertEinsteinCollegeofMedicine, 2 Mean agewas41.3years(SD11.1);66.8%wereAfrican , ChinazoCunningham Health ResourcesandServicesAdministration’ Although 60%ofUnitedStatesHIV-infected womenare Among womenenrollinginamulti-siteinitiativetopro- Care amongHIV-Infected Women Correlates ofNotReceivingHIV 1 (presenting) , ArthurBlank 2 e identifiedspecificfactorsassoci e usedchi-squareandmultivari 1 , JasonFletcher 2 , s - - - ORAL ABSTRACTS 39 Linkage to Care Demonstration Sexual Relationships Outside 395 Project: A Practice-Advised 396 Primary Partnerships and Intervention to Facilitate Emotional and Abstinence are Associated with Lower Cognitive Responses for Rapid Linkage to Adherence and Adherence Gaps: Data from HIV Care the Partners PrEP Ancillary Adherence Study Robert Bolan1 (presenting), Risa Flynn1, Steve Leyva1, K Rivet Amico2, Rhodri Dierst-Davies3, Amy Rock-Wohl3, Alexander Kintu1 (presenting), Susan Hankinson2, Mario Perez4 Raji Balasubramanian2, Karen Ertel2, David Bangsberg3, Jessica Haberer3 1 Los Angeles Gay and Lesbian Center, Los Angeles, CA, USA 2 University of Connecticut, Brighton, MI, USA 1 Harvard School of Public Health, Boston, MA, USA 3 Los Angeles County Department of Public Health, Los Angeles, 2 Division of Biostatistics and Epidemiology, School of Public ABSTRACTS CA, USA Health and Health Sciences, University of Massachusetts, 4 Los Angeles County Department of Public Health, Division of HIV Amherst, MA, USA and STD Services, Los Angeles, CA, USA 3 Massachusetts General Hospital, Boston, MA, USA Introduction: The Los Angeles Gay and Lesbian Center is one of the Background: High adherence is critical for effective pre-exposure largest HIV testing and care services providers in Los Angeles prophylaxis (PrEP) against HIV. Initial studies indicate that sexual County, performing over 12,000 HIV tests annually with a positivity behavior influences adherence; however, further data are needed on rate of 3-4%. In response to the National HIV/AIDS Strategy linkage ORAL different types of sexual behavior and their associations with levels to care (LTC) goals, a Linkage to Care Specialist (LTCS) was hired in and patterns of PrEP adherence. 2011. A 20% improved LTC rate (to 90%) was seen in the subsequent 24 months. We sought to characterize strategies introduced by the Methods: We enrolled 1,147 HIV-negative individuals in long-term LTCS that may have influenced the improved rate, and identify gaps serodiscordant relationships at three sites in Uganda from the in approach. Soliciting guidance from experts in research and prac- Partners PrEP Study- a randomized placebo-controlled trial of daily tice we analyzed the LTCS strategies in relation to social-behavioral oral tenofovir and emtricitabine/tenofovir in East Africa. Sexual theory and models of linkage. Guided by intervention mapping we behavior was assessed via monthly interviews and PrEP adherence then described a reproducible practice-advised LTC intervention. was measured through electronic monitoring. We used generalized estimation equations to assess risk factors of low adherence (<80%) Description: The intervention 1) focuses on early emotional and cog- and gaps in adherence. nitive reactions to receiving an HIV-positive diagnosis; 2) is one-on- one; 3) emphasizes tailoring of intervention components based on Results: Fifty-three percent were male, 51% were 18-34 years and specific client needs; 4) allows for referral to other services; 5) 24% were polygamous. Participants who had sex with someone includes phone-based follow-up as needed; and 6) uses integrated other than their primary partner and <100% condom use (N=40) were monitoring systems to identify individuals not linking into care. A more than twice as likely to have low adherence (OR = 2.48, 95%CI = unique assessment tool administered shortly following disclosure of 1.70-3.62) compared to those who had sex with only their primary HIV diagnosis will be used to explore domains of fear, stigma, knowl- partners and 100% condom use. Those who abstained from sex in the edge, attitudes, and support. Linkage rates, viral load, CD4, adher- previous month (N = 124) had 30% higher odds of low adherence (OR ence, and retention data will be gathered for 12 months following = 1.30,95%CI = 1.05-1.62). Participants in non-polygamous relation- diagnosis to evaluate the intervention. ships who reported sex with both their primary and other partners with <100% condom use (N = 15) were almost twice as likely to be Lessons Learned: Implementation of a research protocol in the con- low adherers (OR = 1.76,95%CI = 1.01-3.08). At least one 72-hour gap text of clinical care in a manner that leverages successes while offer- in adherence was seen in 598 participants (54.7%); 23.2% had at least ing potential solutions for gaps is feasible when all team members one one-week gap. Participants who had sex with other partners and with practice and research-based interests are well represented. <100% condom use (N = 47) had 50% higher odds of a 72-hour gap in adherence (OR = 1.50,95%CI = 1.19-1.91). Recommendations: Involvement of clinical staff in protocol development for research on linkage to care is critical to balance the require- Conclusions: Risk of low overall adherence and gaps in adherence ments of research with patient centered and compassionate care. were higher in participants who reported sex outside their primary partnerships and those who abstained from sex. Polygamy was associated with lower odds of low adherence. Gaps in adherence were common, potentially creating risk for HIV acquisition.

40 9th International Conference on HIV Treatment and Prevention Adherence 397 retroviral therapy(ART)adherence. positive patientsassignedto at 3urbanUSclinicswererandomizedtoreceive physician adherencecareduringroutineclinicvisits.AdultsonART clinic visit,withresultsproducedreal-timetosupportpatientand puter Julie Devasia period the MedCHECRandomizedTrial Antiretroviral Adherence:InitialResultsof Conclusions: tributed follow-updata. 23% reportedIVDU,32%MSM.N=238(51%MED,49%EDU)con non-white (72%),aged>50years(57%),anddosedoncedaily(65%); Results: analysis effects. Methods: domized trialofatouch-screencomputer adherence isunknown.We conductedamulti-siteprospectiveran- ness oftablet-deliveredpatient-facinginterventionstosupport inexpensive andcommoninhealthcare.Buttheclinicaleffective- Background: 4 3 2 Allen Gifford Effects ondose-timingadherenceandotheroutcomesareunknown. tion, buttheeffectdidnotpersistduring6-12monthperiod. ing adherencethanpatientsassignedtostandardeduca adherence 9.6%higherinMEDduringtheperiod. greater intheMEDgroupvs.EDU(p<0.03),withadjustedmean 1 ence wasnotdifferentbetweenthegroups. assure atleastoneclinicvisittablet-useopportunity. We tested separate randomeffectsperpatient.A45-day“wash-in”wassetto line adherence,healthandsocio-demographiccharacteristics,with ear modelswerefitasafunctionofstudyarm(MEDvs.EDU),base sive 7-dayperiodsduringtheyearoffollow-up.Randomeffectslin- fraction oftotaldailyexpecteddosestaken,averagedoversucces- and throughout1yearoffollow-up.Adherencewasdefinedasthe drug monitor(MEMSCap™)adherencewasmeasuredatbaseline, adherence careeducatorswasavailableatstudyclinics.Electronic or standardcareadherencepatienteducation(EDU).Referralto 9th InternationalConference on Department ofVeteran Affairs,LosAngeles,CA,USA Department ofVeteran Affairs,Bedford,MA,USA MA, USA Boston UniversityandDepartmentofVeteran Affairs,Bedford, Veterans HealthAdministration,Bedford,MA,USA -assisted andadaptedself-interviewcompletedpriortoeach (days 46-180)and The N=255randomizedwerepredominantlymale(81%), MedCHEC 1 2 Tablet computersareincreasinglyconvenient,familiar, Intervention toSupport A Tablet-Computer Clinical During oneyearoffollow-up,ART-experienced HIV- , JoelReisman (presenting) is abrowser-based HIVadherenceaudiocom- Early period , SarahMcDannold late period 3 MedCHEC , MarkGlickman HIV T reatment andPrevention Adherence (days 181-360)intent-to-treat adherence wassignificantly initially hadbetterdose-tak -based interventionforanti 2 2 , KeithMcInnes , MatthewGoetz Late period MedCHEC (MED), adher 2 early , 4 ------400 Re-Engaging LostHIVClinicPatients Amy Rock-Wohl Introduction: Rhodri Dierst-Davies Recommendations: Lessons Learned: regular provider. linked tocarewithin3months;andd)recentincarcerationwithno recent VL>200copies/mL;c)newly-diagnosedwithHIVandnot data; b)noreportedVLorCD4measurementin7-12monthsandmost ed VLorCD4measuresinprevious12monthsperHIVsurveillance sons 18yearsofageorolderwithanythefollowing:a)noreport Description: or moremedicalvisitsat6months. enrolled intheinterventionforatleast6monthshadcompletedtwo enrolled participantswerelinkedtocare,howeveronly59%ofthose over 90dayswasneededtolinkpatientscare.Atotalof98% records (36%).Anaverageof7appointmentsor15hoursNA tact informationcamefromHIVsurveillance(39%)andclinicmedical sured (51%)andreportedstablehousing(87%).Themostusefulcon enrolled (n=47).ThemajoritywasLatino(68%),male(72%),unin- able orhadinvalidcontactinformationand12%werelocated (in-care elsewhere,deceased,notLACresidents);43%wereunavail- 1 to care.Preliminaryresultsarepresented. fied ARTAS strength-basedinterventionwasusedtopromotelinkage HIV/STD surveillance,clinicrecordsandpublicdatabases.Amodi- to locate/contactpatientsincludinguseofcontactinformationfrom clinics. Enhancedpublichealthinvestigativetechniqueswereused developed toidentify, locateandre-engagelostpatientsatfourLAC load [VL]atleast3monthsapart).TheNavigationProgramwas Angeles County(LAC)wereretainedincare2011(2ormoreviral longer interventionwouldlikelyimproveretentionincareat6months. intervention waseffectiveatlinkingpersonstocare,howevera vided themostusefulcontactinformationandmodifiedARTAS Los Angeles,CA,USA Los AngelesCountyDepartmentofPublicHealth, Eligible lostclinicpatientsincludedHIV-infected per- Innovative MethodforFindingand The NavigationProgram:An An estimated47%ofHIV-infected personsinLos 1 Among 499lostclinicpatients,43%wereineligible HIV surveillanceandclinicmedicalrecordspro 1 (presenting) , SalonikiJames 1 , V time - - - ORAL ABSTRACTS 41 Population-Based Estimates of Time to Retention among Persons 402 Engagement in the Continuum of 408 Linked to HIV Care in BC, Canada, HIV Care in Western Kenya: From HIV from 2000-2012 Testing to Retention Lillian Lourenco1 (presenting), Guillaume Colley1, Dmitry Shopin1, Keith Chan1, Bohdan Nosyk1, Kate Heath1, Julio Montaner1, Becky Genberg1 (presenting), Violet Naanyu2, Juddy Wachira3, Viviane Lima1 Paula Braitstein4 1 British Columbia Centre for Excellence in HIV/AIDS, Vancouver, 1 Brown University, Providence, RI, USA BC, Canada 2 Moi University School of Medicine, Eldoret, Kenya 3 AMPATH Consortium, Eldoret, Kenya Background: Previous analyses have demonstrated the greatest 4 Indiana University School of Medicine, Indianapolis, IN, USA attrition observed in the HIV cascade of care for the province of British Columbia (BC), Canada, occurs in the transition from ‘linked’ Background: Few estimates of engagement in the early phases of the ABSTRACTS to ‘retained’ in care. Thus, we performed a time to retention analysis continuum of HIV care from community-based samples exist in sub- among persons linked to HIV care in BC to understand why leakage Saharan Africa. The objective of this study was to report engage- occurs. ment within one district in Kenya served by AMPATH (Academic Model Providing Access to Healthcare). Focused on care delivery, Methods: We included individuals in the BC STOP HIV cohort ‘linked’ AMPATH is a partnership between Moi University, Moi Teaching and to care between 2000 and 2012 who had at least one calendar year of Referral Hospital, and a consortium of North American institutions. follow-up. Individuals were followed until death, last contact date or end of December 31, 2011. ‘Linked’ was defined as the first instance ORAL Methods: 3,485 HIV-positive individuals >13 years were identified of an HIV-related service following HIV diagnosis. ‘Retained’ in care during population-wide home-based counseling and testing (HBCT) was defined as having, within a calendar year, either: i) an HIV-relat- between December 2009-February 2011. AMPATH medical records ed physician visit or diagnostic test (CD4 or pVL) or ii) 2 ARV dispen- verified engagement in care (initial encounter with an HIV provider), ³ sations, 3 months apart. Weibull survival analysis was used to treatment eligibility (CD4 testing), antiretroviral therapy (ART) initia- ³ determine factors associated with time to retention among linked tion, retention (clinic visit within three months of scheduled return), persons. We adjusted for several fixed and unfixed clinical and and mortality as of April 2013. Individuals reporting care outside of demographic variables. AMPATH were excluded. Logistic regression analysis was used to examine socio-demographic factors associated with losses in the Results: A total of 5,231 persons were linked to care between 2000 continuum. Secondary analysis examined the subset newly diag- and 2012 (79% male, median age ([Q1-Q3]): 39 (32-46) years, median nosed during HBCT. time to retention: 8 (5-12) months), of which 540 (10%) were never ‘retained’ in care. The adjusted survival model estimated that Results: Of 3286 eligible individuals, 64% were female, with a mean females (adjusted Hazard Ration (aHR) = 0.81 (95%CI:0.75,0.87)), indi- age of 36. As of April 2013, 59% (n = 1,940) had engaged in care, 59% viduals with non-HIV related visits to a general practitioner (GP) dur- had received CD4 testing, and 52% had initiated ART. Among those ing ‘linked’ year (aHR = 0.75 [95%CI:0.70, 0.80]) or non HIV-related engaged in care, 10% were not retained in care and 3% died. Men hospital admissions, during linked” year (aHR = 0.60 (OR = 0.64, 95% CI: 0.55,0.77) and those <25 years (OR = 0.33, 95% CI: [95%CI:0.54,0.67]) were less likely to be retained in care. Individuals 0.27,0.39) were less likely to have engaged in care. Men and those 40-49 (aHR = 1.20 (1.10, 1.31) and ³50 years of age (aHR = 1.16 (1.05, <25 years were also less likely to have had CD4 testing, initiated ART, 1.28), and those with non-HIV related specialist visits, during ‘linked been retained in care, and more likely to have died. Among those year’ were more likely to be retained in care (aHR = 3.06 (95%CI: 2.71, newly diagnosed during HBCT (n = 1,349), 12% linked to care, 67% 3.45)). within 90 days. Conclusions: Individuals who saw non-HIV-related specialists during Conclusions: Engagement estimates were similar to recently pub- their linked year were more likely to be retained in HIV care than lished reviews of largely clinic-based samples, however linkage fol- those who saw a non-HIV-related GP. This missed opportunity for lowing HBCT among those newly diagnosed was low. Additional engagement in HIV care suggests a potential for GP-related interven- efforts are needed to engage men, young adults, and individuals test- tions to improve time to retention in care. ing via HBCT.

42 9th International Conference on HIV Treatment and Prevention Adherence 409 Based HealthCenter Implementation HybridinaCommunity- individual settingsandcommunities. including concretestepsthatcanbeadaptedtothespecificneedsof suggests animplementationblueprintforotherhealthcenters, potential toavertnewHIVinfectionsamongMSM.Thispresentation within communityhealthcentersisacriticalsteptowardrealizingits Sharon Marazzo based healthcenterinNewYork City. integrated comprehensiveHIVpreventionpackageatacommunity- develop aprograminwhichPrEPisprovidedandsupportedasfully prevention programs.Thisproject(R01AA022067)wasfundedto few community-healthcentershaveintegratedPrEPintotheirHIV uptake amongmenwhohavesexwith(MSM)remainslow, and Recommendations: that mightimpactPrEPuptakeandpersistence. include PrEP-specificelements;and6)Confrontingissuesofstigma cols, 5)Developingvisitprotocolsthatspanprovidertypesand Deciding abouteligibility, targeting,referralandself-referralproto- sexual healthclinic,STItesting,HIVcounselingandtesting);4) related protocolsthroughoutthehealthcenterimpactedbyPrEP(i.e., variety ofstaffskepticismandresistance;3)Creatingoradapting extend beyondriskreductiontowardsexualhealth;2)Managinga areas: 1)Developingmessagesandcounselingstrategiesthat Lessons Learned: Description: Introduction: 2 Sarit Golub 1 staff atalllevels,writtenprogrammaterialsandawebsite. adherence. We alsodevelopedspecifictrainingforhealthcenter PrEP initiation,theotherdesignedtosupportandimprove ventions -onedesignedtoassistpatientsindecision-makingaround prehensive PrEPimplementationprogram,includingtwobriefinter- use/adherence. Basedonthisformativework,wedevelopedacom ty-level factorsthatmightinfluencePrEPaccess,uptake,and holders (n=44)regardingindividual-,organizational-,andcommuni- views withhealthcenterproviders,patients,andcommunitystake- 9th InternationalConference on Callen-Lorde CommunityHealthCenter, NewYork, NY, USA Hunter College,CUNY, NewYork, NY, USA 1 (presenting) T PrEP Demonstration/ Developing andImplementinga Although PrEPreceivedFDAapprovalinJuly2012, o developtheprogram,weconductedin-depthinter 2 , AnthonyCatalanotti Our workpointedtocriticalconsiderationsinsix Creating PrEPdeliveryandsupportprograms , AnitaRadix HIV T reatment andPrevention Adherence 2 2 , AmyHilley 2 , - - 413 HIV CareinRuralVirginia with multiplechallengestolinkageandengagementincare.The in care.PeoplerecentlydiagnosedwithHIVruralVirginia struggle nosed withHIVarelinkedtocareandasmallerproportionisretained Description: Positive Links Introduction: 2 1 George Reynolds Colleen Laurence Recommendations: Lessons Learned: Services (ARTAS) programandcoordinatedthemwiththe counseling sessionsdevelopedbytheAntiretroviralT PLp lenges withatailoredsmartphoneappandcounselingprogram. low uptakeofin-personcounselingsessions. but crisismessagesarefrequentandmustbeaddressed.4)Thereis board providesanimportantspacetoaskquestionsandshareideas, dashboard arethemostusedfeatures.3)Thecommunitymessage way eachweek.Thecommunitymessageboardandtheindividual On average,70%oftheparticipantsinteractwithappinsome ipants is32.54%areunder30.70%male,andMSM.2) pants intheprogram,74%ofthoseapproached.Meanagepartic- From September25,2013toJanuary11,2014,weenrolled26partici- patients, anditisreachingakeydemographicofyounggaymen. Wendy Cohn tional materials;andmonitorsadherenceotherkey active reminders;offersavirtualanonymouscommunityandeduca- highly secureapplication(app).Theappdeliverspersonalized,inter sion ofano-costcellphonethatisloadedwithcustom-designed, referred totheprogram.Participationinprogramincludesprovi- prospectively andsystematically. be collected.Clinicalandengagementoutcomeswillcollected and quantitativedataaboutappusageexperienceovertimewill Health DecisionTechnologies, Sheboygan,WI,USA University ofVirginia, Charlottesville,VA, USA has adaptedthestrengths-basedcasemanagement(SCBM) 1 Nationally, modelssuggestthat59-69%ofpeoplediag- , Support forSuccessfulLinkageto Positive Links:SmartPhonesand program ( PLp Rebecca Dillingham 2 1 , Tabor Flickinger 1) The , ErinWispelwey is offeredtoallnewly-diagnosedindividuals One yearofparticipationisplanned.Qualitative PLp PLp ) describedstrivestoaddressthesechal- is highlydesirabletonewlydiagnosed 1 1 1 , KarenIngersoll , ErinPlew-Ogan (presenting) 1 1 , , reatment and . Inaddition, PL app. - ORAL ABSTRACTS 43 Comparing Adherence Items of Houston Primary Care Providers’ 414 Missed Doses with Different 416 Perceptions of and Willingness to Timeframes and their Associations with Prescribe HIV Pre-Exposure Prophylaxis Viral Load in Routine Clinical Care Charlene Flash1 (presenting), Monisha Arya1, Amber Amspoker1, Kenneth Mayer2, Douglas Krakower3, Micha Zheng4, Robin Nance1, J Delaney1, Rob Fredericksen1, Anna Church1, Thomas Giordano1 Anna Harrington1, Paul Crane1, Heidi Crane1 (presenting) 1 Baylor College of Medicine, Houston, TX, USA 1 University of Washington, Seattle, WA, USA 2 Harvard University, Boston, MA, USA 3 Beth Israel Deaconess Medical Center, Boston, MA, USA Background: Medication adherence items often ask how many 4 University of California, Berkeley, CA, USA doses have been missed over a time period. However, questions remain regarding the optimal timeframe for asking about medication Background: Although HIV pre-exposure prophylaxis (PrEP) uptake adherence in clinical care settings. To determine which timeframe is ABSTRACTS requires providers to be knowledgeable and willing to prescribe or most useful, we compared 4-, 7-, 14-, 30-, and 60-day self-reports refer, little is known about their attitudes and practices. with viral load levels. Methods: From January-April 2013, community health center primary Methods: Observational study of 920 patients on antiretroviral thera- care providers in Houston, TX, participated in an anonymous on-line py from the University of Washington. Patients complete touch- survey, analyzed via multiple logistic regression. screen assessments including adherence at ~4-6-month intervals as part of clinical care. Adherence items asking about numbers of Results: 210 providers completed the survey. Mean participant age ORAL missed doses were included with timeframes of 4, 7, 14, 30, and 60 was 36 years and 63% were female. Almost half (43%) were faculty days. Using patients’ most recent assessment, and each timeframe and 56% trainees; 42% were internists, 26% family physicians, 19% item, we calculated missed doses per day and examined the correla- ob/gyns, and 11% HIV specialists. Fifty-seven percent believed PrEP tions with each other and with viral load. We conducted logistic and to be proven safe and effective; most of the rest were unsure. linear regression analyses examining associations between different Controlling for demographic and medical practice variables, HIV spe- timeframes and viral load as a binary and log-transformed continu- cialists had 7.4 times greater odds of agreeing that PrEP is safe and ous outcome. effective (p = .002), compared to other providers. Forty-one percent believed they could identify PrEP candidates in their practices. HIV Results: Longer timeframes capture increasing numbers of patients specialists were at nearly 4 times greater odds of being confident in as having missed doses (23% for 4-day vs. 58% for 60-day items). their ability to identify patients who needed PrEP than other Correlations between calculated missed doses for each timeframe providers (p = .003). Most providers would be willing to refer patients was strong (e.g.14- vs. 30-days 0.871) however became increasingly for PrEP (94%) or to prescribe PrEP (86%) if trained to do so. less strong for time periods farther apart (e.g. 4- vs. 60-days. 0.560). Nonetheless, 60% preferred PrEP be managed by a specialist. Calculated missed doses from the 14-day item had the largest corre- Although only 18% of providers had received a patient inquiry about lation with viral load. In logistic and linear regression models of viral PrEP, 80% would be motivated to prescribe PrEP by patient requests. load including all timeframes, only missed doses calculated from the However, 29% believed PrEP might promote risky behavior. 14-day adherence item was significant (p <0.05). In a stepped approach, adding the other timeframes did not improve the R2 of Conclusions: More than 40% of primary care providers in Houston models including the 14-day item. were unsure that PrEP is safe and effective. Many were concerned about increased risky behaviors in PrEP users, and many were Conclusions: Adherence measured by all missed dose items corre- uncertain they could identify eligible patients. Most primary care lated with viral load and with each other to varying degrees. Items providers prefer HIV specialists manage PrEP. These findings high- with longer timeframes captured more patients with missed doses. light the need for additional training for primary care providers to The 14-day item seemed to be best able to capture adherence behav- enhance their confidence in prescribing PrEP and willingness to ior as measured by viral load. engage patients in the use of PrEP.

44 9th International Conference on HIV Treatment and Prevention Adherence 427 in HIVCare injection druguse(IDU)history, AIDSdiagnosis,andHIVriskfactor. between groupsandlogisticregressiontocontrolforage,race,sex, to indexvisit.W est toindexvisitand2)#ofdayslatefordrugimmediatelyprior index drug(30-daysupply/daysbetweenfills)*100%fortwofillsclos care indexvisit.Adherencewasdefinedas:1)%dosestakenforan come was‘show’vs.‘no-show’atarandomlyselectedprimaryHIV adult HIVclinicbetweenOctober2012andApril2013.Primaryout- Conclusions: founding. the ‘noshow’than‘show’group(p=0.003].Therewasnocon- being lateforthemostrecentprescriptionwassignificantlyhigherin 75, >100%)vs.81%(95%CI:54%,>100%),p<0.005].Theprobabilityof taken washigherforthe‘show’than‘noshow’group[93.8%(95%CI: 46% onboostedproteaseinhibitor. 89(30%)‘noshowed.’%doses male, 63%black,20%withIDUhistory, 53%withAIDSdiagnosis,and Results: Methods: Background: 3 2 1 Robert Gross Robert Bonacci early markerofnon-retention. ed constructs.We evaluatedwhetherARVrefilladherenceisan months). Retentionandadherencetoantiretrovirals(ARVs)arerelat- with thelongtimeneededfordefiningadequateretention(6-12 time oflaterefilltopreventfuturenon-retentioninHIVcare. use thisearlymarkertoidentifyandaddressbarriersarisingatthe flag individualsinneedofretentioninterventions.Futurestudiesmay concept thattrackingofpharmacydataforbeinglaterefillsmay was significantlylowerin‘noshows’than‘shows.’Thisisproofof Katherine Frasca 9th InternationalConference on University ofPennsylvania,Philadelphia,PA, USA Philadelphia FIGHT, Philadelphia,PA, USA Philadelphia, PA, USA University ofPennsylvaniaPerelmanSchoolMedicine, 297 participantsweremedianage48years(IQR41-53),69% We enrolledaretrospectivecohortfromPhiladelphia 3 ‘No show’toclinicwascommonandARTadherence Non-retention inHIVcareisdifficulttodetectearly is anEarlyPredictorofRetention Antiretroviral RefillAdherence (presenting) 1 , Warren Bilker 1 , Lyles Swift e usedranksumteststocompareadherence 3 , JingRen 1 HIV , LauraBamford T reatment andPrevention Adherence 1 , DaohangSha 2 , BalighYehia 1 , 3 , - 428 in theSouth Background: 5 4 3 2 1 Mirjam-Colette Kempf Kelly Smith Conclusions: Results: defined atathresholdof<400VLcopies/mL. gate markersforretentioninHIVcare.virologicsuppressionwas Availability ofCD4countsorHIVviralload(VL)wereusedassurro- trimester andevery6monthspostpartumfrom6-36months. each intervalofprenatalandpostpartumcareasavailable; in BirminghamorMontgomery ed from226HIV-infected womenwhodeliveredbetween1998-2008 Methods: Martin Rodriguez women tomaintainclinicaloutcomesachievedduringpregnancy. ence tocareduringpostpartumamongpregnantHIV alarming. Effortsneedtobemadeaddressretentionandadher- outcomes suchasretentionincareandVLsuppressionratesare cessfully retainedincareduringpregnancy, postpartumHIVcare Alabama. care amongHIV This study’s objectiveistoprovideinsightonpostpartumretention in attention hasbeengiventotheirpostpartumcontinuumofHIVcare. been afocusofprenatalcareamongHIV-infected women,less at timeofdelivery (81.5%), withamedianageof26.8yearsandrange15-45 24 monthsand33%at36months. period forwomenwhowereretained,with40%at12months,42% suppression. VLsuppressionratesdeclinedoverthepostpartum for 198pregnanciesduringthe3rdtrimester, with68%showingVL tum and42%at36monthspostpartum).VLdataonlywasavailable over time(59%at12monthspostpartum,49%24postpar HIV care(92%)priortodelivery, followedbydecreasedretention nancies duringthethirdtrimesterofpregnancy, indicatingoptimal Birmingham, AL,USA Nursing andHealthBehavior, UniversityofAlabamaat Birmingham, AL,USA Department ofPediatrics,UniversityAlabamaatBirmingham, Birmingham, AL,USA Department ofBiostatistics,UniversityAlabamaat Birmingham, AL,USA Department ofMedicine,UniversityAlabamaatBirmingham, Birmingham, AL,USA Department ofEpidemiology, UniversityofAlabamaat Women inourcohortwerepredominantlyAfrican-American In thisretrospectivereview, 266pregnancieswereinclud- 1 , JodieDionne-Odom While preventionofmothertochildtransmissionhas Care amongHIV-Infected Women Postpartum RetentioninHIV While mostpregnantHIV-infected womenaresuc- -infected womenseekingcareatfiveHIVclinicsin 2 . VLorCD4countdatawasavailablefor245preg , MarshaSturdevant 5 (presenting) , AL.Datawerecollectedoncefor 2 , InmaculadaAban 4 , 3 , -infected - - ORAL ABSTRACTS 45 Six-Month Outcomes from a Engagement in HIV Care Following 430 Medical Care Coordination 431Release from Jail Program at Safety Net HIV Clinics in Los Mi-Suk Kang Dufour1, Janet Myers2 (presenting), Angeles County Timothy Buisker3, Mark Morewitz2, Rebecca Packard2, Milton Estes4, Kate Monico-Klein4 Wendy Garland1, Rhodri Dierst-Davies2 (presenting), Sonali Kulkarni1 1 UCSF CAPS, San Francisco, CA, USA 2 University of California, San Francisco, CA, USA 1 Los Angeles County Department of Public Health, Division of 3 University of California, Berkeley/University of California, San HIV/STD Programs, Los Angeles, CA, USA Francisco, CA, USA 2 Los Angeles County Department of Public Health, Los Angeles, 4 Forensic AIDS Project, Department of Public Health, San CA, USA Francisco, CA, USA

Background: Medical care coordination (MCC) models have Background: HIV-infected individuals receive health care in jail and ABSTRACTS improved health outcomes for a number of chronic diseases, howev- are often medically stabilized while incarcerated but face challenges er, there is limited application of these models in HIV care settings. to establishing continuing care in the community following release. To improve health outcomes at HIV clinics in LAC that serve predom- Patient navigators can support engagement in care and adherence inantly low-income, minority populations, we implemented a MCC to treatment, resulting in viral suppression and reduced HIV trans- service model that is integrated into routine clinical services. mission risk. Preliminary 6-month viral load (VL) and service data are presented. Methods: The Navigator Project is a randomized controlled trial for

ORAL Methods: In 1/2013 MCC was implemented at 18 Ryan White-funded HIV-infected inmates leaving jail in San Francisco. Participants were HIV clinics to identify and serve patients at risk for poor health out- randomized to either 1 year of active patient navigation or 90 days of comes as follow: HIV diagnosis 6m; VL 200 copies/mL; not on anti- as needed case management. To examine adherence to ARVs after ² ³ 3 retrovirals (ARTs) but CD4 ²500 cells/mm ; active substance user; release, we used data from jail and city-wide electronic medical incarcerated in past 6m; and last medical visit >7m. Baseline assess- record systems maintained by public clinics. Audio computer assist- ment and service delivery data were collected. Change in median VL ed self-interviews were conducted at 2, 6 and 12 months following and viral suppression (VL <200 copies/mL) at 6m were the main out- release. Results are presented here for the first 2 months post- comes. Comparisons from baseline to 6m were performed using release. Wilcoxon Signed Rank and McNemar’s tests for paired data. Regression methods were used to calculate odds ratios (ORs) and Results: Baseline pre-release data were available for 271 individuals. 95% confidence intervals (CIs) and to identify predictors of viral sup- In the 2 months following release, interview and medical record data pression at 6m. were available for 215 and 254 participants respectively. Most (78%) participants reported having seen an HIV provider for their regular Results: From 1/2013-10/2013, 271 patients were assessed for MCC care. Undetectable viral loads were reported by 50% and confirmed (51% Latino, 21% African American; 86% male; 76% ³ poverty level; in clinical data for 46% in the two months post release. Undetectable 54% ³40 years; 34% foreign-born; and 73% on ARTs). At 6m, patients viral load lab results were associated with higher baseline scores on received 10.8 median hours of MCC (interquartile range (IQR) = 12.0). the coping and self-efficacy scale (OR 1.01 95% CI 1.00, 1.02) and with From baseline to 6m, median VL decreased from 4,060 copies/mL older age (OR 1.03 95% CI 1.00, 1.06). There were no statistically sig- (IQR=55,158) to 68 copies/mL (IQR = 9,046) (p <0.0001) and the propor- nificant differences in self-reported adherence or service utilization tion of patients with viral suppression increased from 28% to 58% (OR or undetected viral load by alcohol or substance use severity, psychi- = 5.1, 95%C I= 2.6, 9.8). Viral suppression at 6m was not associated atric diagnoses or between study arms during the first 2 months fol- with hours of MCC received (OR = 0.9, 95%CI = 0.9-1.0). lowing release. Conclusions: Preliminary findings suggest that a clinic-based MCC Conclusions: Self-reported data indicated that for participants model has the capacity to identify and serve patients at risk for poor receiving either type of case management, engagement in care was health outcomes and improve viral load suppression at 6m. high among HIV-infected individuals in the first 2 months following release from jail.

46 9th International Conference on HIV Treatment and Prevention Adherence 436 care patientsisunknown patients. Howbesttousethesesourcestogetheridentifyoutof provide differentinformationregardingcarestatusforthesame Clinic-based datasourcesandpublichealthsurveillancemay tion, andincarcerationcanresultinmisclassificationofcarestatus. viduals whoarenotincareisapriorityissue.Silenttransfers,migra- Ascertainment ofHIVCareStatus Revery Barnes Conclusions: Results: as incare.W with aCD4orviralloadresultinthe210-dayperiodwereclassified Department ofPublicHealthHIV/AIDSsurveillanceregistry 2013. We thenmatchedthesamplewithSanFrancisco not seenanHIVprimarycareproviderinthe210dayspriortoApril6, in-person tracking.Patientswereconsideredoutofcareiftheyhad tained throughchartreviewandoutreachbyphone,email,mail, of April6,2013,forclinic-basedtracking.Carestatuswasascer- “late” foraprimarycarevisitbyelectronicmedicalrecordqueryas a 10%randomsampleofactivepatientswhowereatleast210days Methods: Background: 4 3 2 Katerina Christopoulos 1 of regularprimarycarevisitsinadditiontolaboratorytesting. may beaneffectivestrategy, especially sinceeffectivecareconsists alone. Usingsurveillancedatatoinformclinic-basedoutreachefforts provide abetterunderstandingofcarestatusthaneithermethod of carebyclinictrackingasincare. care. Surveillanceclassified12/32(38%)ofthosenotlocatedorout veillance classifiedasoutofcare,trackingfound22(52%)werein care, 2(2%)incarcerated,and1(1%)haddied.Of42individualssur- tracking found60(63%)incare,23(24%)notlocated,9(10%)outof Edwin Charlebois 9th InternationalConference on San Francisco,CA,USA UCSF CenterforAIDSPreventionStudies, Los Angeles,CA,USA Harbor UCLADepartmentofFamilyMedicine, San Francisco,CA,USA San FranciscoDepartmentofPublicHealth, University ofCalifornia,SanFrancisco,CA,USA Of 940patientslosttofollow-up,95weresampled.Clinic At alargepublicHIVclinicinSanFrancisco,weselected e comparedresultsfrombothsources. Identification andre-engagementofHIV-infected indi- Public HealthApproachesto Examining Clinic-Basedand Clinic-based trackingandsurveillancedatatogether 3 , Wendy Hartogensis 1 , StephenMorin 1 (presenting) HIV 1 , Hong-HaTruong 1 T , SusanScheer , Wayne Steward reatment andPrevention Adherence 2 1 , , ElvinGeng 4 , . Patients 1 437 and AccesstoCareInitiative with HIV: TheHRSA/SPNSSystemsLinkage Retention InterventionsforPeopleLiving needs thanthosewhoareout-of-care. People whoarenewlydiagnosedwithHIVhavedifferent(lesser) as welltransitioningpatientsoutofthespecializedservices. entiation betweenlinkage/retentionspecialistsandcasemanagers caseloads andhomevisits/fieldwork.Challengesincluderolediffer- interventions includefocusedsupportservicesdirectedtosmall out-of-care andlooselyretainedpatients.Commonfeaturesofthe creating procedurestoprioritizetheidentificationandretentionof care specialists.Staterepresentativesandtheirclinicpartnersare on additionalhumanresourcese.g.,patientnavigators,linkage-to- age servicesand/orcreatingnewinterventions,whichrelyheavily ing. Interventionsincludeenhancingandstandardizingexistinglink- linking, engaging/re-engagingandretainingpatientsratherthantest- different interventions.Themajorityoftheinterventionsfocuson center statewide demonstrationprojectincludesamultistateevaluation linkage andretentioninterventionsovera4-yearperiod.This Six statesreceivedfundingtodevelopandimplementHIVtesting, (SPNS) launchedtheSystemsLinkageandAccesstoCareInitiative. Administration (HRSA)SpecialProjectsofNationalSignificance Background: 3 2 Kimberly Koester Conclusion: Results: following theframeworkanalysisapproach. interviews wereaudio-recorded,transcribedverbatimandanalyzed demonstration siterepresentatives,andprojectimplementers.All ventions. Stakeholdersincludedstate-levelprojectrepresentatives, document thedevelopmentandimplementationprocessofinter- observations ofstatewideplanningmeetings(n=18)in6statesto ducted keyinformantinterviewswithstakeholders(n=68)and tative component.Duringtheformativeevaluationphase,wecon- Methods: 1 Cara Safon the emphasistoretainingpatients. home visitsandare-conceptualizationofcasemanagementtoshift are intensive.Areasofinnovationincludeincreasedabilitytodo While theinterventionsarenotnecessarilyexpansiveinreach,they tions thatconcentratetimeandpersonnelonhardtoreachpatients. University ofCalifornia,SanFrancisco,CA,USA Prevention Studies,SanFrancisco,CA,USA University ofCalifornia,SanFrancisco/CenterforAIDS UCSF CenterforAIDSPreventionStudies,SanFrancisco,CA,USA . To date,thedemonstrationstatescollectivelygenerated19 The multistateevaluationincludesacomprehensivequali 3 , ShaneCollins The SPNSdemonstrationprojectsaretestinginterven Implementation ofLinkageand Evaluating theDevelopmentand In 2011,theHealthResourcesandServices 1 (presenting) 1 , Wayne Steward , ShannonFuller 1 , JanetMyers 2 , AndreMaiorana 3 3 , - - ORAL ABSTRACTS 47 To Take or Not to Take PrEP: in+care Campaign: Improving 440 Perspectives from Participants 441 Retention in HIV Care Enrolled in the iPrEx Open Label Extension Michael Hager1 (presenting) (OLE) in the United States 1 National Quality Center, New York, NY, USA Hailey Gilmore1 (presenting), Kimberly Koester2, Albert Liu1, K Rivet Amico3, Vanessa McMahan4, Pedro Goicochea4, Background: The in+care Campaign is designed to facilitate local, David Lubensky5, Susan Buchbinder6, Kenneth Mayer7, regional and national improvement on patient retention in HIV care Sybil Hosek8, Robert Grant4 and to evaluate the factors correlated with such improvements.

1 Bridge HIV, San Francisco, CA, USA Methods: The Campaign involves prospective analysis of participant- 2 UCSF Center for AIDS Prevention Studies, San Francisco, CA, USA submitted sequential cross-sectional data using standardized indica- 3 University of Connecticut, Brighton, MI, United States tors: three retention indicators and one viral suppression indicator. 4 Gladstone Institutes, UCSF, San Francisco, CA, USA Utilization of Campaign resources, frequency of data submission, and ABSTRACTS 5 Bagatto, Inc., San Francisco, CA, USA important facility characteristics were included in the analysis. The 6 San Francisco Department of Public Health, San Francisco, CA, USA Campaign collected data from over 200 HIV provider organizations 7 Fenway Community Health, Brown University, Boston, MA, USA located in 48 states and territories and representing a patient pool of 8 Stroger Hospital of Cook County, Chicago, IL, USA 160,000 HIV-infected individuals (non-de-duplicated).

Background: iPrEx OLE, the first global HIV pre-exposure prophylax- Results: Average performance on three Campaign indicators is (PrEP) demonstration project, launched in 2011. Uptake of PrEP in improved over the two years of data collection (gap measure from

ORAL OLE was 69% overall. We describe motivations to take or not take 15.6% to 12.29%, visit frequency measure from 63.9% to 69.5%, viral PrEP among participants in San Francisco (SF), Boston, and Chicago. suppression measure from 70.0% to 73.2%). Organizations serving 501 to 2000 HIV+ people had higher levels of performance than those Methods: During OLE participants who had previously participated in serving 500 or fewer or greater than 2000 HIV+ people across the blinded PrEP trials were given the option to take open label PrEP or same three indicators (e.g., viral suppression measure October 2013 not. A subset of US-based participants was purposefully selected for was 76.5% for middle-sized agencies, 71.2% for small agencies, and in-depth-interviews (IDIs). Questions included reasons for choosing 71.8% for large agencies). In addition, community health centers had to take/not take PrEP and how PrEP did/did not fit into their lives. higher levels of performance than hospitals and health department Interviews were transcribed verbatim, coded, and analyzed using clinics (e.g., visit frequency measure October 2013 was 75.2% for Framework Analysis. non-FQHC health centers, 74.2% for FQHCs, 68.5% for hospitals, and 64.7% for health department clinics that are not FQHCs). Results: Overall, 293 men who have sex with men (MSM) enrolled in Improvement was greater among those agencies that reported their OLE in SF (154), Boston (93), and Chicago (46). Between April and retention strategies to Campaign staff (e.g., gap measure perform- August 2012, 77 participants (28 SF, 26 Boston, 23 Chicago) participat- ance improved by 5.4 percentage points for intervention reporters vs. ed in IDIs; 54 (70%) were on-PrEP. Median age was 34; 49% white; 1.4 percentage points for non-reporters) and among agencies per- 36% Black; 15% other; Chicago participants were younger and more forming in the bottom quartile at baseline. likely to be black or Latino. Most on-PrEP IDI-participants chose to take PrEP for an “extra layer of protection” when having sex, such as Conclusions: Retention in HIV care improved among Campaign par- in the event of condom non-use or malfunction. A few reported using ticipants over the first 2 years of the in+care Campaign. Additional PrEP as their primary HIV risk-reduction strategy. Off-PrEP partici- follow-up with more rigorous design is needed to identify which spe- pants expressed prior experiences with or concerns about side cific retention strategies led to greater improvement in performance. effects; change in relationship status; utilizing other safer-sex strategies; or (less often) having medical issues precluding use of PrEP; as reasons not to take PrEP. Participants in both groups described study benefits including regular HIV-testing, and a desire to continue to contribute to the HIV-prevention field.

Conclusions: IDI-participants expressed a range of factors influencing their PrEP-uptake decisions. Accurate information regarding safety/tolerability may assist in the decision to take PrEP. Understanding and supporting motivation to use PrEP will be critical to its success as an HIV prevention intervention.

48 9th International Conference on HIV Treatment and Prevention Adherence 443 Antiretroviral Therapy:AMeta-Analysis Conclusions: Results: Methods: Background: David Finitsis bias, andoptimizeitstransitiontomHealthapplications. acteristics thatwillimproveresponseaccuracy ART adherence.Additionalresearchshouldinvestigatedesignchar- measuring antiretroviraltherapy(ART)adherence. currently availablestudiescomparingVAS withothermethodsof ever hasyieldedamixedliterature.Thismeta-analysissynthesized Research ontheapplicationofVAS asameasureofadherencehow- front-runner foruseinmobileapplicationsmeasuringadherence. graphical interface,thevisualanaloguescale(VAS) isaputative monitor andmeasureadherence.Withitssingle-itemstructure 1 load (k=9;r-0.32;95%CI:-0.14,-0.48). was amediumstrengthassociationbetweenVAS scoresandviral and electronicdatamonitors(k=4;r0.51;95%CI:0.25,0.69).There = 0.55;95%CI:0.37,0.69),pillcounts(k5;r0.77;0.64,0.86), ed highratesofconcordancewiththewellvalidatedAACTG(k=6;r 0.43, 0.47).Sensitivityanalysesbycomparisonmeasuretypeexhibit- by studyindicatedalargestrengthofassociation(r=0.45;95%CI: years and45.5%ofthetotalsample(N=5212)wasfemale.MeanES pill count),andviralloaddata.Themeanageofparticipantswas37 VAS withadditionalself-reportmeasures,objectivemeasures(e.g., pendently duplicated. effects assumptions.AllstudycodingandEScalculationswereinde tion reportedineachstudy. Effectsizeswereanalyzedunderrandom ance weightedFisher’ coded forparticipantanddesigncharacteristics.Aninversevari- least oneotheradherenceorbiologicalmeasure.Studieswere an HIVsampleusingVAS and(b)analyzedVAS concordancewithat 9th InternationalConference on University ofConnecticut,Storrs,CT, USA Sixteen studiesmetinclusioncriteria.Studiescompared Included studies(a)reportedonmedicationadherencein Analysis supportsthecontinueduseofVAS tomeasure 1 Mobile technology(mHealth)caninnovatehowwe to MeasureAdherence Using theVisual AnalogueScale (presenting) s Zr effect sizewascalculatedfrominforma , JenniferPellowski HIV T reatment andPrevention Adherence 1 , SethKalichman , reduceself-report 1 - - 444 were mostlikelytosubmitperformanceandinterventiondata. regional andlocalresponsestotheNHAS.Middlesizedagencies Campaign wasanaction-orientedvehiclearoundwhichtoorganize United Statesasdiscussionforumsforretentionissues.The improvement strategies.35localretentiongroupsformedacrossthe assigned QIcoachandfewerstill(21%)submittedinformationon almost everydatapoint.Only33%ofparticipantsmadeusetheir ance dataatsomepointandmorethan20%ofenrolleessubmitted Description: Introduction: 1 Michael Hager HIV Services This abstractwillalsobepresentedasaposter. Recommendations: Lessons Learned: context ofapublichealthcampaign. and toevaluatethefactorscorrelatedwithsuchimprovementsin regional andnationalimprovementonpatientretentioninHIVcare in theconsumeractivity de-duplicated). Morethan500HIV-infected individualsareenrolled ries andrepresentingnearly475,000HIV-infected individuals(non- enrolled over500HIVproviderorganizationsin48statesandterrito tion orthreatofremovalforpartialparticipation.TheCampaign select whichactivitiestojoina-la-carte,andtherewasnoexpecta not arequirementofRyanWhitefunding.Participantswereableto Participants wereinvitedtoenrollvoluntarily related tothemeasures;and5)forumsforconsumerinput. standardized retentionindicatorsandimprovementinterventions enrolled providersandotherkeystakeholders;4)datareportingon 3) opportunitiestoorganizelocalretentiongrouptaskforcesamong and subpopulations;2)accesstonationallyrecognizedQIcoaches; ters, andwebresourcesgearedtowarddifferentstakeholdergroups including: 1)aneducationalcurriculumincludingwebinars,newslet- ment potential. chronic diseaseprojectstorapidlyexpandlearningandimprove- public healthcampaignmodelcanbeemployedaspartofother care areneededtoserveasapeerlearningplatform.Ana-la-carte National QualityCenter, NewYork, NY, USA The Campaignhasseveralactivitiesforparticipants, The in+careCampaignisdesignedtofacilitatelocal, Vehicle forQualityImprovementin Public HealthCampaignsasa 1 (presenting) More than50%ofenrolleessubmittedperform- Future campaignsfocusedonretentioninHIV , Partnersin+care. , andparticipationwas - - ORAL ABSTRACTS 49 Comparison of Engagement in Care Seek, Test, Treat and Retain for 448 Measures Using Self-Report and 456 People Who Inject Drugs in Kenya: Clinical Records Data Findings from the First Intervention Period of a Stepped Wedge Study Irene Kuo1, Kossia Dassie1, Paige Kulie1, Wenze Tang1, Ifeoma Ikwuemesi1, Jillian Dunning1, Sean Allen1, Meriam Mikre1, John Lizcano1 (presenting), Eva Muluve2, Ann Kurth1, Amanda Castel1 (presenting), James Peterson1, Avani Patel1 Charles Cleland1, Peter Cherutich2 1 George Washington University School of Public Health and 1 New York University, New York, NY, USA Health Services, Washington, DC, USA 2 MOH Kenya, Nairobi, Kenya Background: Determining where a person exists along the continu- Background: HIV infections in sub-Saharan Africa increasingly um of HIV care is often determined using various sources including occur among people who inject drugs (PWID). Needle and syringe self-report and clinical data. The objective of this analysis was to programs (NSPs) and PWID-specific ART support have been nearly ABSTRACTS assess differences in care-status measures collected from patient non-existent, though Kenya is among the first to implement NSP at a report and clinical records data. country-wide level starting in 2013. Methods: Between June-December 2013, a convenience sample of Methods: Evaluation is being done using a stepped wedge cluster- HIV-infected persons receiving care at three clinics in Washington, randomized design. We are using respondent-driven sampling (RDS) DC was surveyed regarding perspectives on engagement in care. to reach PWIDs for HIV-1 prevalence and viral load determination Participants provided self-reported visit frequency and were [SEEK]. We will then collect study data in additional time periods as assigned a HRSA stage of care status. In this preliminary analysis, ORAL PWID service sites roll out, including behavioral data collected using we compared engagement in care measures based on self-reported tablets, rapid HIV testing [TEST], POC CD4 determination for HIV-pos- and clinical records (CR). itives, and assignment of peer case managers (PCMs) to those with CD4 <350 cells/␮L to link to ART with adherence [TREAT]. Both PCMs Results: Of 159 participants surveyed, 94 (59%) had CR abstracted; of and PWID will receive small conditional cash transfers for PWID those, 81% were male, 64% black, median age was 50 years, and 93% adherence to HIV care visits [RETAIN]. were insured in the previous 12 months. Twenty-seven (28%) self- reported visit patterns consistent with being in sporadic care; using Results: 1739 individuals were screened during the first intervention CR, 24% were determined to be in sporadic care; of those in sporadic period with 1489 found to be eligible and enrolled (85.6%). Most care, 100% perceived themselves to be “fully engaged in care” when enrolled participants were male (87.9%). Median age was 31 years; presented with a visual continuum of care diagram. Based on CR, a age ranged from 18 to 83 years. Median age at first injection was 26 median of 4 visits were scheduled during the study period, of which years. About one in five (19.5%) were HIV positive. About 22% (n = 64) a median of 3 visits were kept. Participants self-reported more of those with HIV infection (n = 290) were newly diagnosed by our missed visits than those documented in the CR (mean 1.23 vs. 0.95). study. 21 participants were eligible to be assigned to a PCM and ini- Forty-two percent self-reported the longest gap between care visits tiate ART. Of those, 19 initiated ART, 9 successfully continued on ART, as being ³6 months vs. 30% based on CR. 6 stopped taking ART, 2 were incarcerated, and 2 died. Thus 60% were retained in care (9/15 retained). Conclusions: Among this sample of clinic attending HIV-infected persons, patients’ perceptions of their care patterns and engagement in Conclusions: The combination of RDS and rapid testing is an effec- care differed from their actual receipt of care as measured by clini- tive strategy for finding PWID with HIV infection, including those not cal records. Patient education regarding the meaning of optimal care previously diagnosed. Linkage to care by PCMs has been very effec- engagement may assist in the measurement of care patterns and fur- tive for ART initiation, but 40% of those not incarcerated or deceased ther inform research on the HIV care continuum. do not continue ART.

50 9th International Conference on HIV Treatment and Prevention Adherence 471 Trials GroupStudyParticipants and HIVViral LoadamongAIDSClinical that theyexplained24-weekVLdetectability barriers at12-weekintheirorderofimportancebasedonthedegree used dominanceanalysisforvariableimportance,byarranging11 and VL(detectablevs.undetectable)at24weeks(+/-4weeks).W between 11individualandsummedbarriersat12weeks(+/-4weeks) In bivariatelogisticregressionmodels,weexaminedtheassociation ART studiesconductedintheUnitedStatesbetween1997and2013. Mallory Johnson unchanged whenaccountingforstudyprotocol. with VLandwas10thinrelativeimportance.Patternofresults tance inexplaining24-weekVL.“Simplyforgot”wasnotassociated detectable VLandwerethebarrierswithhighestrelativeimpor whelmed” weresignificantlyassociatedwithoddsofhavinga the drugwastoxic/harmful,”“feltsick/ill,”anddepressed/over- “had toomanypillstotake,”“wantedavoidsideeffects”,“feltlike “were awayfromhome”(24.9%).However commonly reportedmissingARTdueto“simplyforgot”(22.5%)or depression asopposedtofocusingonforgetfulness. educating aboutandmitigatingARTadverseeffects,managing ventions maybemoreeffectivebyprioritizingdecreasingpillburden, virologic detectabilityinACTGstudyparticipants.Adherenceinter depression weremostimportantandsignificantlyassociatedwith high pillburdenorperceived/actualmedicationadverseeffectsand Conclusions: Results: Methods: Background: 2 1 Parya Saberi 386 cells/mm with men,and54.9%treatment-naïvepre-enrollment.MeanCD4+was male, 53.4%White,29.5%Black,18.1%Latino,51.1%menwhohavesex barriers andVLtodiscerntherelativeimportanceofeachbarrier. ACTG studiestoexaminetheassociationbetweenARTadherence 9th InternationalConference on Northwestern University, Chicago,IL,USA University ofCalifornia,SanFrancisco,CA,USA At 12weeks,weexamineddatafrom2,918individuals,83.2% We assessedadherencebarriersandVLfrom11ACTG 3 1 We conductedalongitudinalsecondarydataanalysisof Adherence barriersrelatedtoARTregimenssuchas Antiretroviral TherapyAdherence Association betweenBarriersto (presenting) and 44.0%hadanundetectableVL.Participantsmost 1 , MargaretChesney , Torsten Neilands HIV T 1 reatment andPrevention Adherence , SusanCohn , onlymissingARTdueto . 1 , EricVittinghoff 2 1 , e - - 472 Results fromtheVeterans AgingCohortStudy (2001: 38%to2010:84%).Theproportionwith portion virallysuppressedevenamongthosewith<95%adherence low-up. Therewasasignificantincrease(ptrend<0.001)inthepro- 7 6 5 4 3 2 David Rimland Shilpa Viswanathan der prescribingnewHAARTregimens earlyinHIVinfection. achieve perfectadherence,concerns ofadherenceshouldnothin- lations. AlthoughallHIV-infected personsshouldbeinstructed to among NNRTIuserslikelyrepresentsmoreeffectiveHAARTformu Conclusions: virus in2006-2010with remained stableat38%amongPIusers.Theproportionsuppressing increased overtimeforusersofNNRTI-basedregimensbut Results: Methods: RNA suppressionbyregimentype. the minimumcutoffofadherencetonewerHAARTneededforHIV retroviral therapy(HAART)andHIVRNAsuppression,estimated Background: 8 tic regression. G. CalebAlexander appreciably differbetweenNNRTI-basedsingleandmultiplepilluse. for 90-94%and85-89%adherence,respectively. Associationsdidnot lower adherencelevels,oddsratios:1.1(0.89,1.36)and0.82(0.64,1.04) PI-based regimens,whereasNNRTIuserssuppressedviruswith tively. Suppressionwith<95%adherence waslesslikely(p<0.05)for 76.0% forusersofNNRTI-,PI-andINSTI-basedregimens,respec- that observedwith at whichtheoddsofsuppressionwasnotsignificantlydifferentthan year examined bythemostfrequentlyusedHAARTregimeninagiven dates. T of drugsupplytothenumberdaysbetweenfirstandlastrefill 2010, adherencewasdefinedastheratiooftotalnumberdays Cohort StudyVirtual CohortbetweenOctober2000andSeptember 1 Atlanta, GA,USA VA MedicalCenter, EmoryUniversitySchoolofMedicine, San Francisco,CA,USA Program/Department ofMedicine,UCSFSchoolPharmacy, San FranciscoGeneralHospital-UCSFPositiveHealth Emory UniversityRollinsSchoolofPublicHealth,Atlanta,GA,USA School ofPublicHealth,Baltimore,MD,USA Johns HopkinsSchoolofMedicine,Bloomberg New Haven,CT, USA Yale SchoolofPublicHealth/VA ConnecticutHealthcareSystem, Baltimore, MD,USA Johns HopkinsBloombergSchoolofPublicHealth, MD, USA Johns HopkinsBloombergSchoolofPublicHealth,Baltimore, VA MedicalCenter, BaylorCollegeofMedicine,Houston,TX,USA . Theminimumneededadherencecutoffwasdefinedasthelevel emporal trendsinadherenceandviralloadsuppressionwere 21,865 HAARTuserscontributed82,217person-yearsoffol- Using pharmacyclaimsdatafromtheVeterans Aging We examinedtrendsinadherencetohighlyactiveanti- Suppression intheCurrentEra- Adherence andHIVRNA The loweradherencelevelneededforsuppression 6 , MariaRodriguez-Barradas 1 ³ , Todd Brown 1 95% adherenceusingrepeatedmeasureslogis- (presenting) ³ 95% adherencewere84.1%,79.2%and 3 , AmyJustice , NeelGandhi 7 , LisaJacobson 2 4 , , IanMcNicholl ³ 95% adherence 8 5 , - ORAL ABSTRACTS 51 Every Dose Every Day: A Peek into Pulling the Network Together: A 473 CDC’s New e-Learning Training 482 Novel Social Network Intervention Toolkit to Improve Adherence to for Promoting Engagement in HIV Care on Antiretroviral Treatment Mfangano Island, Kenya

Rhondette Jones1 (presenting), Kathleen Green1, Janet Van Ness2, Matt Hickey1 (presenting), Charles Salmen2, Dan Omollo3, Angel Ortiz-Ricard1 Brian Mattah3, Elvin Geng4, Peter Bacchetti4, Cinthia Blat4, Gor Ouma3, Kathryn Fiorella5, Daniel Zoughbie6, Robert Tessler7, 1 Centers for Disease Control and Prevention, Atlanta, GA, USA Marcus Salmen8, Monica Gandhi4, Starley Shade9, 2 John Snow Inc., Boston, MA, USA Elizabeth Bukusi10, Craig Cohen4

Background: Despite improvements in the potency and tolerability of 1 UCSF/Organic Health Response / Microclinic International, San antiretroviral treatment (ART), more than 800,000 of the estimated 1.1 Francisco, CA, USA million people living with HIV in the United States do not have a sup- 2 UCSF/Organic Health Response, Minneapolis, MN, USA ABSTRACTS pressed viral load. The CDC’s Division of HIV/AIDS Prevention 3 Organic Health Response, Mfangano Island, Nyanza, Kenya Capacity Building Branch developed a media-rich and interactive e- 4 University of California, San Francisco, CA, USA learning training toolkit entitled - “Every Dose Every Day” to support 5 UC Berkeley/Organic Health Response, Berkeley, CA, USA providers as they counsel and encourage patients to attain maximum 6 Microclinic International, San Francisco, CA, USA adherence to their HIV regimen. The toolkit features four e-learning 7 UCSF East Bay/Organic Health Response, Oakland, CA, USA modules based on evidence-based behavioral interventions that 8 Organic Health Response, Minneapolis, MN, USA were found to improve HIV adherence among ART-naïve and/or ART- 9 UCSF Center for AIDS Prevention Studies, San Francisco, CA, USA 10 ORAL experienced patients. Kenya Medical Research Institute, Kisumu, Kisumu, Kenya

Description: We piloted the Every Dose Every Day toolkit among 30 Background: Despite progress in the global scale-up of antiretroviral providers (physicians, nurses, pharmacists, and health educators) to therapy (ART), sustained engagement in HIV care remains challeng- assess interest in the toolkit and obtain feedback on the content of ing. Social capital has been identified as an important factor for sus- the learning environment. Providers were asked to review the contained engagement, but interventions to harness this powerful social tent of at least 1 module and complete a 30-item survey for each force are uncommon. module reviewed. Each module provided the learner with background information on the intervention, videos demonstrating the Methods: We conducted a quasi-experimental study evaluating the intervention in action, and downloadable materials that can be given impact of a targeted social network intervention on engagement in to patients about the importance of adherence. HIV care at a rural health facility on Mfangano Island, Kenya. 369 (87%) of 426 eligible adult patients on ART were enrolled. The inter- Lessons Learned: The pilot study revealed that over 90% of providers vention was introduced into one of four similar communities served by thought the length and pace of the activity was appropriate. Nearly this clinic, and comparisons were made between communities using all of providers reported that the delivery method used helped them intention-to-treat. Microclinics, composed of patient-defined support learn the content. Over 95% of providers indicated they could apply networks, participated in 10 bi-weekly discussion sessions covering the knowledge gained as a result of the activity. Very few providers topics ranging from HIV biology to group support. The curriculum also reported needing technical assistance with implementation. included voluntary participation in a group HIV status disclosure ses- Providers took an average of 60 minutes to complete each module. sion. We report impact on disengagement from care, measured by the incidence of ³90 day gaps in care following a missed clinic appoint- Recommendations: The development of an e-learning training for HIV ment, using Cox proportional hazards regression. The model was providers facilitates a swifter movement in disseminating evidence- adjusted for potential clinical and demographic confounders and based interventions into practice. E-learning technologies may be a included robust standard errors to account for clustering. more feasible delivery method when compared to face-to-face training to assist providers with supporting optimal patient adherence. Results: 113 (74%) intervention community participants joined a microclinic group, 86% of whom participated in group HIV status disclosure. Over 22-months of follow-up, incidence rates of 90-day disengagement were 6.8 per 100 person-years in the intervention group (95%CI 4.2-10.9) and 12.9 (95%CI 9.6-17.3) in control. In the adjusted Cox model, intervention community participants experienced one- half the rate of 90-day clinic absence as those in control communities (adjusted hazard ratio 0.48, 95%CI 0.25-0.92).

Conclusions: The microclinic intervention holds promise as a feasible community-based strategy to improve long-term engagement in HIV care. Reducing treatment interruptions using a social network approach has important implications for individual patient virologic suppression, morbidity and mortality, and for broader community empowerment and engagement in healthcare.

52 9th International Conference on HIV Treatment and Prevention Adherence 483 ty tocareanduseofmultiplesites. Monitoring theHIVCareContinuum and Viral Suppression:ANewMethodfor designing interventionstoimproveHIVcare. a potentialnewtoolformonitoringtheHIVcarecontinuumand care, retentioninandviralsuppression.Geospatialanalysisis continuum aspossiblepointsforintervention. analysis toidentifygeographicareaswithpooroutcomesalongthe improving thehealthofpeoplelivingwithHIV. We usedgeospatial Baligh Yehia Conclusions: spot” wassignificantlyassociatedwitheachoutcome(p<0.05). regression modelscontrollingforpatientfactors,residenceina“hot were linked,retained,andsuppressed,respectively. Inmultivariable 64.6%, 33.0%,and26.3%ofindividuallivingoutsidea“hotspot” 24.0% retainedincare,and9.1%virallysuppressed.Incomparison, “hot spot.”Forthoselivinginaspot,”54.3%werelinkedtocare, 12-19 uniquetractsperoutcome;23.8%ofthesampleresidedina 59 of384(15.4%)Philadelphiacensustractswerein“hotspots,”with Results: within 90daysofdiagnosis),retentionincare( Methods: Background: 2 1 Kathleen Brady Outcomes ofinterestwerelinkagetocare( HIV inPhiladelphiaduring2008-2009,withfollowupthrough2012. race, HIVriskfactor ship between“hotspots”andeachoutcome,adjustingforage,sex, poor outcomes.Logisticregressionmodelsevaluatedtherelation- analyzed usingK-functionstoidentifygeographic“hotspots”with the endof24-monthfollowupperiod).Geospatialpatternswere linkage), andviralsuppression(HIV test ineach6-monthintervalofthe24-monthperiodfollowingdate Chelsea Voytek 9th InternationalConference on University ofPennsylvania,Philadelphia,PA, USA Philadelphia DepartmentofPublicHealth,Philadelphia,PA, USA 1,704 personswerediagnosedwithHIVin2008-2009.Intotal, Retrospective analysisofpersonsnewlydiagnosedwith 2 Geospatial patternsareastrongpredictoroflinkageto The HIVcarecontinuumisaneffectiveframeworkfor Linkage toCare,Retentionin Geospatial PatternsPredict 2 1 , MichaelBlank (presenting) , insurancestatus,incarcerationproximi , MichaelEberhart 2 HIV , IanFrank -RNA <200copies/mLclosestto T reatment andPrevention Adherence 2 , DavidMetzger ³ 1 CD4orHIV 1 ³ , AmyHillier 1 CD4orHIV-RNA -RNA test 2 , 2 , - 486 Adherence Motivations effects ofage,education,race,andgender. els ofHLwereassessedinaMANCOV The adherencemotivationsofparticipantswithlowerandhigherlev- make”) andextrinsic(“Iwantotherstoapproveofme”)motivation. tains twosubscales:intrinsic(“animportantchoiceIreallywantto encouragement. Otherunknownfactorsmayalsocontribute. rely moreonprovidersandothersforinformation,support, tions. Thismightindicatetheydon’ ly higherlevelsofexternalmotivationforadheringtotheirmedica- Methods: Background: 2 Drenna Waldrop-Valverde Marcia Holstad Conclusions: likely tobeextrinsicallymotivated. [15, 302]=3.34,p<0.001).PatientswithlowerlevelsofHLweremore significant multivariateeffectonpatients’adherencemotivations(F els was19.2(SD=1.5).LevelofHLassociatedwithanoverall with lowHLwas13.4(SD=2.6)whilethatforthosehigherlev 2 scorewas16.7(SD3.5,rangeto23).Themeanforthose was 49.3yearsandaverageeducation12.8years.MeanHIV-HL- gendered persons;105werewhiteand218black.Average age Results: adhere totheirantiretroviralmedications. ship ofpatients’levelshealthliteracytotheirmotivations ing medicationadherence.Inthisstudy, weevaluatedtherelation- number ofhealth-relatedvariables.Motivationisimportantinenact- 1 cy HL predictingthosewhohadlowervs.higherlevelsofhealthlitera- HIV-HL-2). Inotheranalyses,wevalidatedacut-offscorefortheHIV- ables, antiretroviralmedicationmotivation,andhealthliteracy(the study ofmedicationuse.Questionnairesassesseddemographicvari- . Themotivationscale,basedonself-determinationtheory Nova SoutheasternUniversity, FortLauderdale,FL,USA Emory University, Atlanta,GA,USA The sampleconsistedof221men,96women,and6trans 323 personstreatedforHIVinfectionparticipatedina Health literacy(HL)hasbeenshowntoberelateda HIV-infected personswithlowHLreportedsignificant- Positive PersonsMedication Health LiteracyisRelatedtoHIV- 1 (presenting) 1 , RaymondOwnby t fullyunderstandtheirillnessand A modelthatincludedthe 2 , , con - - - ORAL ABSTRACTS 53 The Impact on 5-Year Mortality and The China Adherence through 490 the Cost-Effectiveness of Proven 493 Technology Study: The Effect of Text-Messaging Programs for Improving Real-Time Feedback on Adherence to Antiretroviral Adherence in Kenya Antiretroviral Therapy

Anik Patel1 (presenting), Richard Lester1, Scott Braithwaite2, Lora Sabin1 (presenting), Mary Bachman DeSilva2, Li Zhong3, Jason Kessler2, Kim Nucifora2, Mia van der Kop3, Carlo Marra1 Taryn Vian1, Wubin Xie4, Feng Cheng5, Jessica Haberer6, David Bangsberg6, Keyi Xu7, Allen Gifford8 1 University of British Columbia, Vancouver, BC, Canada 2 New York University, New York, NY, USA 1 Boston University School of Public Health, Boston, MA, USA 3 Karolinska Institutet, Vancouver, BC, Canada 2 Center for Global Health and Development, Boston University, Boston, MA, USA Background: Use of mobile phones to improve health (mHealth) has 3 FHI360, Beijing, China demonstrated ability to improve the care of HIV-positive patients. 4 FHI360, Nanning, Guangxi, China ABSTRACTS Adherence to antiretroviral therapy (ART) is a critical issue that may 5 Global Health Strategies, Beijing, China be promoted through text-messaging interventions. Randomized tri- 6 Massachusetts General Hospital, Boston, MA, USA als from Kenya have shown that text-message programs improved 7 Ditan Hospital, Beijing, China adherence to ART over 1 year, and 1 trial suggested improved patient 8 Veterans Health Administration, Bedford, MA, USA retention in care. In this study, we examined the impact on 5-year mortality and the cost-effectiveness of text-messaging programs Background: China has rolled out antiretroviral therapy (ART) rapid- compared to standard care. ly, with 287,000 HIV-positive individuals currently on ART. However,

ORAL inadequate adherence, indicated by emerging drug resistance, is a Methods: Adherence improvements above standard care that were problem. The ‘China Adherence through Technology Study’ (CATS) observed in the clinical trials were transformed into costs and bene- assessed an innovative ‘real-time feedback’ adherence intervention fits using a 2nd-order Monte Carlo simulation model. CD4 count using web-linked medication containers (‘Wisepill’) that allow instan- drives changes in Quality Adjusted Life Years (QALY) and costs asso- taneous adherence tracking. CATS’ intervention employed text mes- ciated with HIV health states. Our secondary analysis explored sages triggered by failures in prompt dose-taking, supplemented by retention benefits observed in one trial in addition to adherence counseling using patients’ Wisepill-generated adherence data. improvements. One-way and multi-way sensitivity analyses were included to reflect the uncertainty in the results and to highlight the Methods: We provided patients in Nanning, China, with Wisepill for key drivers of the outcomes of 5-year mortality impact and the incre- one ART medication and monitored their adherence for three mental cost per quality adjusted life year gained (ICER: US$/QALY). months. In Month 3, we stratified subjects to high (³95%) or low (<95%) adherence groups based on Months 1-3 mean adherence, Results: We found that the text-messaging programs would reduce and randomized them within groups to intervention and control arms. 5-year mortality by 10% from 25.3% to 22.7%. The low programmatic In Months 4-9, intervention subjects received a text reminder when- costs also made the program cost-effective by WHO standards in ever their device failed to open within 30 minutes of schedule, and most scenarios. In our base case analysis, the ICER was $1,662/QALY. monthly counseling using Wisepill-generated adherence data; con- In the base case analysis, addition of the retention benefits observed trols received no reminders or adherence-informed counseling. We from one trial further reduced 5-year mortality by 14% from 25.3% to compared Month 9 mean adherence and proportions achieving opti- 21.7%, but increased the ICER to $1,815/QALY. mal (³95%) adherence between arms and within groups. Conclusions: Our simulation results suggest that the text messaging Results: A total of 120 subjects were enrolled; 116 (96.7%) completed interventions could provide meaningful improvements in HIV mortal- the intervention. At randomization, mean adherence was 91.7% vs. ity and longevity through its ability to improve adherence. Retaining 92.6% (p = NS). In month 9, mean adherence was 96.4% vs. 89.2% (p patients in care could further improve its value, though retaining = 0.003); 88.5% vs. 52.7% achieved optimal adherence in intervention patients without good adherence may have a paradoxical effect on vs. control arms, respectively (risk ratio (RR) 1.7, 95% Confidence cost-effectiveness. Interval (CI) 1.3-2.2, p (RR 2.3, 95% CI 1.2-4.6, p = 0.003). Among high adherers, mean adherence was 98.0 vs. 92.0% (p = 0.020); 92.1% of intervention vs. 60.5% controls achieved optimal adherence (RR 1.5, 95% CI 1.2-2.0, p = 0.001).

Conclusions: Real-time feedback significantly improved ART adherence, in both low and high adherers. This approach holds promise for management of HIV and other chronic diseases.

54 9th International Conference on HIV Treatment and Prevention Adherence SUNDAY, JUNE8,2014,6:00 9th InternationalConference on 246-256 —> 272-285 —> 299-311 —> 325-336 —> 349-360 —> 420-432 —> 447-460 —> 474-487 —> 388-399 —> Highest ScoredPosters= HIV T reatment andPrevention Adherence 499 <— 498-488 <— 269-258 <— 296-286 <— 324-313 <— 348-338 <— 385-361 <— 419-401 <— 446-433 <— 470-461 P . M . • AMERICANA 4

POSTER SESSION MAP 55 56 POSTER ABSTRACTS 246 e. f. c. tion amongyoungblackMSM. ing needlesforinjectiondrugsasstronglyassociatedwithHIVinfec- use, sexwhilehighoncrackcocaine,marijuanaandalcohol,orshar- influence ofdrugsoralcohol.Previousstudiescitecrackcocaine ed analintercoursebecauseeithertheyortheirpartnerisunderthe vive onthestreetsbybecomingsexworkers,engaginginunprotect- Young black MSMarerun-awaysandhomeless,forcingthemtosur- have sexwithmen(MSM)is3to4timeshigherthanwhiteMSM. and explainthesignificanceofsubstanceuseamongthispopulation. acquiring HIV/AIDSamongyoungblackMSMaswelldescribe in othercities.Participantswillidentifycontributingriskfactorsfor sions aspreviouslycitedstudieswithHIV Surprisingly, theresultsofthisstudydonotdrawsameconclu- substance useplaysinthelivesofHIV range offactorsandenhanceourunderstandingabouttherolethat d. Results: Methods: Background: 1 Austin Nation b. a. HIV transmission. this populationthataddressissuessurroundingsubstanceabusein opment ofpreventioneducationstrategiesspecificallytailoredto and thesenseofhope.Thisinformationwillcontributetodevel- lence andaccess,testingpositiveforHIVwillpower, coping about family contributing tothelimitedbodyofknowledgecurrentlyavailable ry withasummaryofthesignificantexperiencesthispopulation, tive narrativestudydescribeanacross-caseexperientialtrajecto- Francisco with MenbetweentheAgesof18-29inSan Positive Young BlackMenwhohaveSex willpower, coping,andasenseofhope testing positiveforHIV methamphetamine exposure,prevalence,andaccess relationships relocation family such asunprotectedreceptiveanalintercourse. eventually beingHIVinfected. thing theyarenotsurprisedabout;there isaresignationabout ment and shower. including foodandhousingsotheycanhaveaplacetosleep and theneedtofindacommunityhavesenseofbelonging. the themeof University ofCalifornia,SanFrancisco,CA,USA family lack ofacceptanceandearlyexposuretodrugssexinthe orientation alongwithrejection,judgment,discrimination,and their circumstances.Theydiscoverthedrug’ drug, includingforemotionalnumbingsothattheycandealwith among theirnewfoundcommunityandpeerpressuretousethe effect andthisleadsthemtoengage inhigh-riskbehaviors . The themesthatemergedfromthecodingofthisqualita- includes issueswithbeingstigmatizedduetotheirsexual The goalofthisqualitativestudyistoofferinsightaboutthe , relocation,relationships,methamphetaminepreva to SanFranciscotalkaboutHIV, beinghomelessand The prevalenceofHIVamongyoungblackmenwho Substance UseExperiencesofHIV- Pilot StudytoDescribethe 1 , whichincludefeelingsofabandonment,loneliness, (presenting) survival , needingmoneyformeettheirbasicneeds , describingasasenseofreliefand some- , HowardPinderhughes -positive youngblackMSM. -positive youngblackMSM for theirfuture s sexual enhance 1 happening - 9th InternationalConference on - on theseyoungblackmen. among thepredominantlywhiteMSMpopulationhashadanimpact lence ofuse,andaccesstomethamphetaminesinSanFrancisco behavior assessment.Itappearsthatthehighexposureto,preva- well astheimportanceofmakingathoroughsexualhealthandrisk to understandtheprevalenceofHIV/AIDSamongthispopulationas tice settingswillencounterHIV-positive youngblackmenandneed Conclusions: Clinicians andresearchersinallacademicprac- HIV T reatment andPrevention Adherence Comparison of Adherence Rates for tically significant (p = 0.390). For scheduled mental health medications in the comparator group, there were 666 patients filling 828 pre- 247 Antiretroviral Medications, Blood scriptions. The comparator group had a higher average PDC than Pressure Medications, and Mental Health PLWH with 85.1% vs. 81.7%, respectively (p = 0.009). Medications for HIV-Positive Patients at an Conclusions: Based on average PDC, PLWH filling prescriptions at Academic Medical Center Outpatient this UCH Infectious Disease Group Practice (IDGP) pharmacy had an adherence of 84.5% to ART. However, only 40% of patients were Pharmacy adherent at the needed 95% level for therapeutic effects. Additionally, PLWH were more adherent to single-tablet once-daily Jacob Langness1 (presenting), Paul Cook2, Jasjit Gill1, regimens than multi-tablet once-daily regimens or multi-tablet twice- Rachael Boggs1, Nathan Netsanet3 daily regimens. Adherence in PLWH to HIV ART was better than the same patients’ adherence to scheduled blood pressure and sched- 1 University of Colorado Hospital, Aurora, CO, USA uled mental health medications. Levels of optimal adherence still 2 University of Colorado School of Nursing, Aurora, CO, USA need to be improved to reduce rates of resistance and maximize 3 University of Colorado Skaggs School of Pharmacy and therapeutic durability of selected regimens. When compared to the ABSTRACTS Pharmaceutical Sciences, Aurora, CO, USA comparator group, adherence in PLWH to scheduled blood pressure medications was similar but adherence to scheduled mental health Background: Despite advances in safety, tolerability, and decreased medications was slightly worse. Further work is needed to address pill burden for HIV antiretroviral therapy (ART), non-adherence con- non-adherence among PLWH, including their adherence both to ART tinues to be a major cause of HIV-related morbidity and mortality. and to medications prescribed for other chronic diseases. Adherence to these medications presents particular challenges because persons living with HIV (PLWH) must use medication more consistently (³95%) than persons with other chronic diseases such as hypertension or diabetes mellitus (³80%). Methods: In this retrospective review, we examined outpatient pharmacy prescription fill records at University of Colorado Hospital POSTER (UCH) to quantify PLWH’s adherence rates to ART, scheduled blood pressure medications, or scheduled mental health medications. We also examined how pill burden and dosing schedule affected adherence. Finally, we analyzed a comparator group, patients who filled scheduled blood pressure medications or scheduled mental health medications (but not HIV medications) at other University of Colorado outpatient pharmacies, in order to compare adherence rates of PLWH and similar patients to these medication classes. Patients 18 years of age or older, on any prescribed HIV antiretroviral therapy, scheduled prescription blood pressure medication or scheduled prescription mental health medication filled between March 1, 2012, and March 31, 2013, were included in the analysis. The proportion of days covered (PDC) was calculated to measure adherence. Statistical analyses were performed to compare adherence rates between groups.

Results: 865 PLWH filled 1,943 antiviral prescriptions with an average PDC of 84.53%. However, only 40% of patients had high enough adherence to achieve therapeutic benefits, based on a criterion of at least 95% of scheduled doses taken. When separated by regimen, 282 patients filled single-tablet once-daily regimens, 295 patients filled multi-tablet once-daily regimens, and 288 patients filled multi- tablet twice-daily regimens. The calculated PDC of PLWH on a single-tablet once-daily regimen was 89.7% vs. 81.0% for PLWH on a multi-tablet once-daily regimen (p <0.001). The average PDC for PLWH (n = 269) who filled 460 scheduled blood pressure prescriptions was 82.68%. The average PDC for PLWH (n = 295) filling 467 scheduled mental health prescriptions was 81.70%. This was a statistically significant difference in adherence between ART and blood pressure medications of 2.5% (p = 0.013) and a difference between ART and scheduled mental health medications of 3.03% (p = 0.002). For the comparator group of similar patients without HIV, 895 patients filled 1398 scheduled blood pressure medications with an average adherence of 83.8%. The difference between groups was not statis-

9th International Conference on HIV Treatment and Prevention Adherence 57 Patient Perspectives on Factors Influencing the 249 Psychological Aspects of Chronic 251 Vulnerability of HIV Transmission Pain While Living with HIV by Injecting Drug Users’ Widows Living with HIV in Ukhrul District Manipur India Jessica Merlin1 (presenting), Melonie Walcott1, Ivan Herbey1, Stefan Kertesz1, Eric Chamot1, Michael Saag1, Janet Turan1 Kingson Shimray1 1 University of Alabama at Birmingham, Birmingham, AL, USA 1 UNICEF Reproductive, Maternal, Newborn, Child and Adolescent Health (RMNCH +A), Imphal East, Manipur, India Background: Chronic pain (pain >3 months duration) is common in HIV-infected individuals, and associated with substantial functional Background: Ukhrul district is 1 of the 9 districts in Manipur with a impairment. Despite effective psychological interventions, providers’ population of 183,115. With HIV prevalence rate of 2.10% as estimat- approach to chronic pain remains mostly biomedical. We recently ed from antenatal care clinic (ANC) thus being the highest HIV preva- adapted and tailored the Biopsychosocial (BPS) Framework to HIV- lent district in Manipur. The current estimate of injecting drug users ABSTRACTS infected patients with chronic pain. Our goal is to explore HIV-infect- (IDUs) in Manipur is about 30,000 and in Ukhrul district there are ed patients’ perspectives on psychological aspects of pain, using the about 3200 IDUs. The IDUs in Manipur contributes to 50% of the total psychological portion of the BPS Framework as a guide. HIV infection. About 40% of IDUs are married, and the rate of HIV infection among their wives increased from 6% in 1991 to 45% in 1997. Methods: Eighteen participants were sampled to include individuals with mood disorders and substance use. Perspectives on chronic Methods: A qualitative exploratory research was done as part of my pain were collected using qualitative interviews. Three investigators Master in Public Health 2013. engaged in an iterative process of independent and group coding. Interviews were conducted until theme saturation was reached. Key Results: Studies shows AIDS related death among the IDUs has themes were identified and classified within the psychological por- increase in the past years, thus ultimately increase the widows of tion of the BPS Framework.

POSTER IDUs. Studies also shows that IDU wives are increasingly becoming widows below the age. Widows of IDUs are frequently stigmatized at Results: Of the 18 individuals interviewed, 13 were male, 12 were 2 levels, being widows of IDUs and HIV-positive individuals. They African American, the median CD4+ T-cell count was 587 cells/mm3 face many social challenges including social violence, stigma and (IQR 421-792), and 14 had HIV RNA <200 copies/mL. The majority (15) discrimination. In India, about 90% of HIV-positive widows reported had a mood disorder, and half (9) actively used illicit substances. forced to stop living in their marital home only 9% gets financial sup- Psychological issues related to pain from BPS Framework categories port from their in laws. These put them into socioeconomic, health included 3 main themes: mood (depression, anxiety); emotions and psychological problem, poverty, loneliness and difficulties to (anger, frustration, fear) and stress; and alcohol/substance use. The care for children. So HIV prevention services were not a primary combined impact of psychological burdens of chronic pain and living concern they predominantly concerned about the livelihood issues. with HIV can make the pain experience difficult to bear. Participants Due to these difficulties, they are more prone to risky behavior articulated a bidirectional relationship between pain and mood, and including sex work and substance abuse. pain and emotions. Substance use was seen as a temporary, mal- adaptive way to deal with pain. Conclusions: The HIV epidemic has placed an additional burden on IDU’s widows living with HIV in Ukhrul district, as many of them Conclusions: The chronic pain experience among HIV-infected indi- become exacerbate to social difficulties and destitute due to viduals resonates with the BPS Framework. Participants showed socioeconomic factors including cost of living and main care takers insight into the importance of psychological factors, and the context for their families, poverty and marginalization that predispose them of living with HIV, in their chronic pain experience. This suggests that to risky behaviors. There is a lack of specific program for psychological approaches to chronic pain treatment may be well widows/spouses of IDUs. received by HIV-infected patients, and should be developed and tailored to this population.

58 9th International Conference on HIV Treatment and Prevention Adherence 252 R34MH097609. improve engagementandretention.FundedbyNIMHgrant vate clinicpatients.Strategiestoenhancehealthcaresystemsmay patient engagementanddisparitiesinhealthamongpublicpri- motivational interventions.Preliminarydatahighlightchallengesto private HIVhealthcaresettingsinBuenosAires,Argentina. most challengingpatients,thosenotengagedincarepublicand care arerequiredforoptimalHIVoutcomes.Thisstudyexaminedthe ditions. regarding ARVsandwererandomizedtointerventionorcontrolcon an assessmentofadherence,knowledge,motivationandattitudes and disengagedfromtreatmentinthelast3to6monthscompleted 1 Conclusions: Results: Methods: Background: 5 4 3 2 Ines Mattioli Deborah Jones addressed inbothclinics. pants failedtoattendanysessions.Healthcarebarrierswere was higherintheintervention(p=.02),thoughonethirdofpartici VL (r=.35,p.01).Preliminaryoutcomesindicatethatattendance (p =.03);self-reportedadherencetomedicationwasassociatedwith (67%) weremorelikelytohavediscontinuedtreatmentfor3months less likelytotakeARVs;thoseconfusedabouttheirhealthstatus tion deficits.Patientsnotengagedincarefollowingdiagnosiswere vation wassimilarbetweenclinicsbuthalfofallpatientshadmotiva all, 52%weretakingARVslessthan80%ofthetime.Treatment moti- p =.005)andmorelikelytohavequestions(X214.81,.002);over health statuswerelesslikelytoknowtheirCD4/VLcount(X2=7.67, than publicpatients(52%).Those(35%)whodidnotunderstandtheir ARVs (p<.005);privatepatientsreportedhigheradherence(85%) reported mildtomoderatedepressionand60%haddiscontinued + 7.92)anddrug/alcoholuse(33%);over52%ofpublicpatients Isabel Cassetti Care inArgentina Maria Lopez 9th InternationalConference on Emory University, Atlanta,GA,USA Helios Salud,BuenosAires,Argentina Fundacion Huesped,BuenosAires,Argentina University ofMiami,FL,USA University ofMiamiMillerSchoolMedicine,Miami,FL,USA Participants weresimilarbetweenclinics(age,x=39.22SD Men andwomen(n=60)prescribedantiretrovirals(ARVs) 1 4 , DiegoCecchini Opportunities existforeducational,mentalhealth,and Engagement, andRetentioninHIV COPA: EnhancingAdherence, Treatment engagement,retentionandadherenceto 3 1 , StephenWeiss (presenting) , LinaBofill 3 , OmarSued 1 HIV , DrennaWaldrop-Valverde T reatment andPrevention Adherence 2 , InesAristegui 3 , PedroCahn 3 3 , , 5 , - - - - 253 Background: Deborah Jones enhance motivationforadherenceamongwomenhopingtoconceive. logue betweenpatientsandprovidersregardingconceptionmay treatment-as-prevention topreventtransmission.Openingthedia- care, anddiscussstrategiessuchaspre-exposureprophylaxis infrequent. Providersshouldincorporatefertilityplanninginroutine partner’s opinionsonfertilityplanning,butdiscussionsthetopic are Conclusions: information onsaferconception. patient initiationofthetopic.Bothwomenandpartnershadlittle providers, partnersandfamilymembers;providerstendedtorelyon tial. Women expressedunwillingnesstodiscussfertilitydesireswith scores” 24.0%,23.8%,21.4%);riskoftransmissionwaslessinfluen the mostinfluentialfactorsinfertilitydesires(mean“importance provider’s opinion,partner’s opinion,andfamily’s opinionemergedas dren (69%)andplannedtohavemore(78%).Usingconjointanalysis, African American(72%,n=23).Themajorityofparticipantshadchil Results: Methods: 1 Nahida Chakhtoura providers (n=14)completedqualitativeinterviews. ual factors.Asubsetofwomen(n=19),partners11)and desires usingaconjointquestionnaireand10-pointscaleofindivid HIV transmissiontopartners,andtheirdoctor’s opiniononfertility importance offinances,opinionspartnersandfamily Miami, Florida. ductive decisionmakingamongwomen,partnersandprovidersin knowledge, attitudesandpracticesregardingconceptionrepro- sions aboutsaferconceptionmethods.Thispilotstudydescribes intervention toaddressthisgaphelppatientsmakeinformeddeci- perspectives onfertilityplanningcanenablethedevelopmentofan or practicesduringroutinevisits.Understandingpatientandprovider Yet, eventhose engagedincareoftenfailtodiscussfertilitydesires to partnersrequiresignificantplanningandtreatmentengagement. and SaferConception University ofMiamiMillerSchoolMedicine,Miami,FL,USA Women (n=32)wereaged18-4536±7)andmost Women (n=32)wereassessedregardingtherelative Attempts toconceivethatdonotriskHIVtransmission Results suggestthatwomenvaluetheirdoctor’s and Living withHIVinFamilyPlanning PATHWAYS: EngagingWomen 1 (presenting) 1 , AlisonColl , RyanCook 1 , AndrewSpence 1 , JoNellPotter 1 , JennyJean 1 , theriskof , 1 - - - POSTER ABSTRACTS 59 Postpartum Retention in Care Reasons Clinicians Delay 254 among Women with HIV 256 Antiretroviral Therapy for Clinically Eligible HIV-Infected Patients Judy Levison1, Natalie Williams2 (presenting), Yvette Peters2, Monique Green2, Yanette Alvarenga2, Cynthia Deverson1 Linda Beer1 (presenting), Eduardo Valverde1, Raiford Jerris1, John John1, Becky White2, Jacek Skarbinski1 1 Baylor College of Medicine, Houston, TX, USA 2 Harris Health System, Houston, TX, USA 1. Centers for Disease Control and Prevention, Atlanta, GA, USA 2. University of North Carolina at Chapel Hill, NC, USA Introduction: Retention in care after pregnancy among HIV-positive women has been identified as a national challenge. Within Harris Background: Guidelines for antiretroviral therapy (ART) initiation Health System, which provides indigent health care in Houston, have evolved, but have consistently noted that adherence issues Texas, a review of 213 HIV-positive women who delivered from 2006 should be considered and addressed, as suboptimal adherence can through 2011 elucidated a 40% loss to care 1-year postpartum. lead to treatment failure and viral resistance. While the prevalence

ABSTRACTS of ART use is often studied, little is known regarding reasons Description: As prenatal care providers for HIV-positive women, our providers delay ART initiation in clinically eligible patients. team identified obstacles to communicating consistent messages to women about the importance of retention in care after delivery. We Methods: In 2009, we surveyed a probability sample of HIV care often focus more on the current pregnancy and immediate adher- providers in 582 outpatient facilities in the United States and Puerto ence to medications than long term follow-up. We sought interven- Rico with an open-ended question about reasons for delaying ART tions to encourage postpartum retention in primary care. A pilot proj- initiation in clinically eligible patients. Thematic codes were devel- ect engaged 31 women over 12 months in a support group scheduled oped through a standardized iterative process and multivariable on days they had appointments. Discussions often addressed issues logistic regression was used to identify provider and practice char- spontaneously brought up by clients; some women chose to attend acteristics associated with reasons to delay ART. even on days they did not have appointments.

POSTER Results: Of 1,743 eligible providers, 734 (42%) completed the survey. Lessons Learned: From the positive response to support groups tied Among 640 who responded to the ART delay question, 15 (2%) said to obstetrical visits, we embarked on merging the infrastructure of they never delayed initiating ART. Reasons for delaying ART were: CenteringPregnancy, a trade-marked program for group prenatal provider concerns about patient adherence (68%), patient accept- care, with HIV education. CenteringPregnancy proposes 10 sessions, ance of ART (60%), and provider concerns about structural barriers each with a specific focus ranging from discomforts of pregnancy to to ART use (33%). Provider age <50 years, non-Hispanic white labor to newborn care. We have developed a parallel 10 session cur- provider race, and having >50 HIV-positive patients per month were riculum for HIV education which includes coping with an HIV diagno- independently associated with adherence concerns. Being a nurse sis, adherence to medication, and multiple messages on retention in practitioner or physician assistant compared to a medical doctor, care. There are no published studies of CenteringHIV programs. having >5 years HIV care experience, >50% patients who were men who have sex with men, and never referring for initial ART prescrip- Recommendations: Our plan is to assess the number of primary care tion were independently associated with acceptance concerns. visits at 1 year among women who have attended group care (com- Non-Hispanic white provider race was independently associated pared to our historical baseline). We will also evaluate pre- and post- with structural concerns. test surveys of knowledge, adherence, stigma, and depression among women in group vs. individualized care. We hypothesize that Conclusions: Reported reasons for delaying ART were consistent women attending this 10-session program are more likely to remain with clinical guidelines and were patient-level and structural; hence in care after delivery. multi-level strategies to address patient barriers to taking ART are likely to be most effective. Providers may benefit from training and access to referrals for ancillary services to enhance their ability to monitor and address these issues with their patients.

60 9th International Conference on HIV Treatment and Prevention Adherence 258 aspects ofindividualizationamongstorganizations. tion incaresolongasthatapproachtakesintoconsideration regional approachtoresolvingissuesrelatedlinkageandreten sive dataprocessthatrequiresaminimalamountoftimetouse. munity. To bemosteffectivethereneedstoaflexibleandrespon- regional areasthatalsoneedattention.Itallowsforsynergyandcom- internally thatshouldbefocusedonaswellidentifyingbroader work towardtheirstrengthswhileallowingthemtoidentifyareas graphics. Theuseofregionalcascadesallowsfororganizationsto such asstaffing,resources,readiness,locationandpatientdemo- change, responsivetoadiversityofindividualorganizationalinputs they operatein.Thisapproachalsoallowsforawidevarietyof change bothwithinindividualorganizationsaswellthesystems reporting offindings. development, implementationandassessmentofinterventions, approach involveddatagenerationandanalysis,QI/QMprocesses, care withthegoalbeingtogetthemallviralsuppression.The dropped outofcare,andretainingallHIV+individualswhowerein HIV+ individualstocare,re-engagingthosewhohad comes relatedtoHIVwithaspecificfocusonlinkingnewlyidentified health outcomesaswellcommunityandpublicout- uums ofHIVcare),begantheprocessimprovingbothindividual laborative groupswhich,utilizingregionallybasedcascades(contin- cific geographicregionsinNewYork Stateweregatheredintocol- Recommendations: Lessons Learned: Description: care withinthosespecifiedgeographicregions. service providersinordertoimprovelinkagecareandretention York Linksto developgeographicallyregionalcollaborativesofHIV Introduction: 1 Steven Sawicki Regionally BasedCollaboratives 9th InternationalConference on Albany, NY, USA New York State DepartmentofHealthAIDSInstitute, All HIVproviders(clinicalandnon-clinical)withinspe- A reviewoftheprocessesandthinkingutilizedbyNew through theDevelopmentof Impacting LinkageandRetention 1 (presenting) This approachisaneffectivemeansofgenerating Systems shouldconsideradoptingamore HIV T reatment andPrevention Adherence - , 259 Background: 6 5 4 3 2 1 Adam Carrico Conclusions: Results: Methods: Multicenter AIDSCohortStudy faster all-causemortalityfollowingHAARTinitiationinthe use predictsmorerapidprogressiontoAIDS-definingillness(ADI)or retroviral therapy(HAART)initiation.Thisstudyexaminedifstimulant but itremainsunclearifthesepersistfollowinghighlyactiveanti- phetamine, crack-cocaine)experienceprofoundhealthdisparities, Michael Plankey Antiretroviral TherapyInitiation HAART initiation. ence healthoutcomesthatarecomparabletonon-usersfollowing gression toADIorall-causemortalityfollowingHAARTinitiation. was nosignificantassociationofanylevelstimulantusewithpro- = 4.9-7.8%)developedanADIoverthecourseoffollow-up.There ticipants withoutahistoryofanADIatHAARTinitiation,6.2%(95%CI all-cause mortalityfollowingHAARTinitiation.Amongthe1,138par There wasnosignificantassociationofanylevelstimulantusewith years). Thecrudemortalityratewas14.5%(95%CI=12.6%-16.5%). nate) duringthe8.5-yearmedianfollow-upperiod(IQR=4.2-11.5 deaths (50%AIDS-related;38%non-AIDSrelated,and12%indetermi- atHAARTinitiation.Therewere190 count lessthan350cells/µl (interquartile range[IQR]=38-48years),andhalf(59%)hadaT ed inthisstudy, 56%wereCaucasian,themedianagewas43years tality followingHAARTinitiation. ing predictorofall-causemortalityaswellADIormor- ed (i.e.,0%,1-49%,50-99%,and100%)wasexaminedasatime-vary cumulative proportionofvisitswhereanystimulantusewasreport- depressive symptoms,bingedrinking,andcigarettesmoking).The markers, otherhealthstatusindicators,self-reportedphysicalhealth, that adjustedforkeyconfounders(i.e.,demographics,HIVdisease Ronald Stall Georgetown University, Washington, DC,USA Chicago MACS,Chicago,IL,USA University ofPittsburgh,PA, USA Johns HopkinsUniversity, Baltimore,MD,USA University ofCalifornia,LosAngeles,CA,USA University ofCalifornia,SanFrancisco,CA,USA Among the1,313menwhohavesexwith(MSM)includ Marginal structuralmodelingproducedaweightedmodel 4 , XiuhongLi HIV-positive personswhousestimulants(e.g.,metham- Following HighlyActive Stimulant UseandMortality 1 HIV-positive, stimulant-usingMSMappeartoexperi- (presenting) 6 3 , DavidOstrow , StevenShoptaw . 5 , DavidVlahov 2 , ChristopherCox 1 , -helper 3 , - - - POSTER ABSTRACTS 61 Retained and Poorly Retained Social Network Characteristics 261 Patients with HIV had Similar Total 262 Moderate the Effects of HIV Stigma Costs in the First 2 Years of Diagnosis on Non-Adherence among African Americans Living with HIV Lauren Richey1 (presenting), Jason Halperin2 Laura Bogart1 (presenting), Glenn Wagner2, Harold Green2, 1 Medical University of South Carolina, Charleston, SC, USA Matt Mutchler3, David Klein4, Bryce McDavitt3, Marc PunKay2 2 New York University, New York, NY, USA 1 Boston Children’s Hospital/Harvard Medical School, Background: Multiple studies have shown that patients with HIV who Boston, MA, USA are retained in care have improved clinical outcomes and survival; 2 RAND Corporation, Santa Monica, CA, USA however, improved retention may result in increased costs for physi- 3 California State University Dominguez Hills/AIDS Project Los cian visits and diagnostic tests. The purpose of this study is to com- Angeles, Los Angeles, CA, USA pare the differences in costs among patients who are retained in 4 Boston Children’s Hospital, Boston, MA, USA ABSTRACTS care vs. those who are poorly retained in care. Background: Research suggests that HIV stigma contributes to dis- Methods: A retrospective cohort study was conducted using the parities, including lower antiretroviral treatment adherence among medical records of patients who had a positive rapid HIV test in the African Americans. Research has not examined whether social net- emergency department in 2008 and were linked to care. Inpatient, work characteristics moderate stigma’s effects. outpatient, and emergency costs as well as number of visits were collected for the first 2 years after initial HIV diagnosis. The Kruskal- Methods: At baseline and 6-months post-baseline, 147 HIV-positive Wallis test (SPSS) was used for analysis. Retained in care was African Americans (71% male) listed 20 social network members defined as 2 visits with an HIV provider divided by 90 days each year ² (alters) and indicated alters’ relationships to them and each other; for 2 years. social network structure measures were derived (proportion of iso-

POSTER lated alters without connections to others; average interaction fre- Results: 56 patients met the inclusion criteria; they were predomi- quency between participant and alters). Enacted stigma was opera- nantly uninsured (73%) and African American (89%). The median total tionalized with an indicator of whether participants had heard any costs per patient for the retained patients over 2 years was $45,793 alter say at least 1 of 2 prejudicial beliefs (“most people with AIDS (range $14,349 to $305,380) and for poorly retained patients $24,491 are responsible for having their illness”; “a person with AIDS must (range $2,685 to $137,489) (p = .11), driven predominantly by outpa- have done something wrong and deserves to be punished”). Social tient costs, median $26,600 for retained patients and $8,478 for poor- network changes were derived by subtracting baseline from 6-month ly retained patients (p = <.00). Inpatient and emergency department values of network characteristics. Daily adherence was measured costs for retained vs. poorly retained patients were similar, $8,100 vs. electronically for 6 months with the Medication Event Monitoring $10,311 (p =.59) and $1,945 vs. $2484 (p =.29), respectively. The medi- System and dichotomized at 85% of prescribed doses taken at an number of outpatient visits over the first 2 years was 30 for ³ baseline and follow-up. retained patients and 7 for poorly retained patients; inpatient days 1.5 vs. 3; emergency room visits 2 vs. 2. Results: On average, 33% reported that any alters said prejudicial beliefs, and 41.9% took ³85% of doses over 6 months. In multivariate Conclusion: Patients with HIV have high healthcare costs, but logistic regressions controlling for socio-demographics and network retained patients, who are known to have better health outcomes changes, baseline stigma was significantly related to decreased and decreased mortality, did not statistically cost more than patients adherence over time, OR(95%CI) = 0.33(0.13-0.85), p <.05. Significant who were poorly retained, with the exception of outpatient costs, interactions of network characteristics with baseline stigma indicat- which was expected. ed that increased interaction frequency with alters over time was protective, and increased proportions of isolated alters over time were detrimental: stigma’s effects on decreased adherence were attenuated among participants who had more frequent interactions with alters over time, OR(95%CI) = 4.59(1.39-15.11), p <.05, and augmented among those with a greater proportion of isolated alters over time, OR(95%CI) = 0.03(<0.01,0.86), p <.05.

Conclusions: Social networks characterized by strong relationships and inter-connections may buffer against stigma’s deleterious effects. Interventions to reduce stigma and support adherence need to take into account social network factors.

62 9th International Conference on HIV Treatment and Prevention Adherence 263 becoming adherent. reminders suchasvisualormeal-timecues,whichwascriticalto stand….” Participantsdevelopedaroutine,includingmedication to talkwith,someonelisten,notjudge,butunder- becoming adherent.Thepeerfacilitatorwasperceivedas“someone things Icouldn’ to HIVmedications.Participantsfeltthat“beabletell(thepeer) depression andlonelinesshadcontributedtopastinabilityadhere adherent, andstigma.Allparticipantsstatedthatemotionaldistress, ed thelackofaroutine,littleconfidenceaboutabilitytobecome 1 notes fromtheinterventionsessionsofthose10participants. time point.Contentanalysiswasusedtoexaminethedetailedfield 90% becameadherentandremainedsoatthe6-monthfollow-up domized toapeer-facilitated medicationadherenceintervention, medication regimensduetonon-adherence.Of10participantsran- care butfailingtreatment.Participantshadfailedanaverageof3HIV controlled pilotstudytargetingadults(n=20)linkedtoHIVmedical how tobeadherentbyexample. matized, thepeerservedasarolemodelandconfidantwhoshowed adults whohadlittlesocialsupport,lackedconfidenceandfeltstig- important intheprocessofbecomingadherent.Forthisgroup ment ofaplantomaketakingmedications‘becomeroutine’were HIV treatmentadherenceandwhocouldthenhelpwiththedevelop- Conclusions: Results: Methods: Background: 2 Maithe Enriquez adherent toHIVtreatment. undertaken tobetterunderstandtheprocessesinvolvedinbecoming with HIVhavedifficultyadheringtotheirmedications.Thisstudywas Repeated Failure 9th InternationalConference on University ofMissouri-Columbia,Nashville,TN,USA University ofMissouri,Leawood,KS,USA Findings indicatedthatmajorbarrierstoadherenceinclud This qualitativedescriptivestudywaspartofarandomized It isunfortunatethatalargenumberofpeopleliving Interaction withapeerwhounderstoodtherigorsof Adherent toHIVMedicationsafter “Someone toListen”:Becoming t tellotherpeople”helpedthemintheirprocessof 1 (presenting) , KimberlyHart HIV T reatment andPrevention Adherence 2 - 267 stimulated discussioninFGDs. and toelicitnewthemes.Whenpresent,peersparkssuccessfully tematically exploresaliencyandrankingofadherencechallenges grasp. Thethemecards,usedduringIDIsandFGDs,helpedtosys- cussion ofdrugdetectionpatterns,adifficultscientificconceptto interviewed todate,theteapottoolfacilitatedunderstandinganddis bers ateachsite.Among47ZimbabweanandSouthAfricanwomen research participants,localclinicstaff,orcommunityboardmem- strategies. Theteapottoolwasfinalizedafterpretestingwithformer researchers andsitestaffwasinstrumentalinthedevelopmentofall Introduction: 4 3 2 Barbara Mensch Miriam Hartmann Recommendations: Lessons Learned: Description: 1 about adherencechallengestoactaspeer“sparks”duringFGDs. population; and4)identificationofIDIparticipantswhowereopen challenges identifiedthroughpreviousqualitativeresearchinthis of overall“adherence”;3)usethemecardslisting20adherence 3 plasmasamples)usingillustrationsofteapots/cupsdepictinglevels participants’ drugdetectionpatternsduringVOICE(basedonatleast tion oflocalmediaclippingsonVOICEresults;2)representation sions ofadherencechallengeswithVOICEparticipants:1)presenta Elizabeth Montgomery Adherence amongWomen inVOICE reporting ofadherencebehaviorsinthecontextclinicaltrials. sion ofscientificallycomplexconcepts,andstimulatemorecandid and innovativestrategiescanhelpfacilitateunderstandingdiscus- adherence injuxtapositiontoindividualdrugdetectiondata. Zimbabwe, SouthAfricaandUganda,exploringreasonsfornon- depth-interviews (IDIs)andfocusgroupdiscussions(FGDs)in participants. MTN-003Disafollow-upqualitativestudyusingin- determined bylowplasmaTFVdetectioninasubsetofactivearm HIV acquisition.Thiswaspotentiallyduetopooradherence,as based pre-exposureprophylaxis(PrEP)wasnoteffectiveinreducing Population Council,NewYork, NY, USA Division ofAIDS,NationalInstitutesHealth,Bethesda,MD,USA FHI360, Durham,NC,USA RTI International/WGHI,SanFrancisco,CA,USA We developed4strategiestoencourageopendiscus- In theVOICEtrial,dailyoralorvaginaltenofovir(TFV)- Impacting StudyProduct Strategies toExploreFactors 4 1 , Use ofpre-testedculturallyappropriatevisualtools Ariane vanderStraten Input gainedthroughconsultationwithPrEP 1 , LisaLevy 2 , JeannaPiper 1 (presenting) 3 , , - - POSTER ABSTRACTS 63 Assessment of VOICE Adherence Validation of a New 3-Item ARV 268 Support Program (VASP) in the 269 Adherence Self-Report Scale: MTN-003 Trial Comparison with Electronic Drug Monitoring Ariane van der Straten1 (presenting), Ashley Mayo2, Elizabeth Brown3, K Rivet Amico4, Helen Cheng1, Nicole Laborde1, Ira Wilson1 (presenting), Yoojin Lee1, Aadia Rana2, Jeanne Marrazzo3, Kristine Torjesen2 Joanne Michaud1, Eric Lamy3, Judith Bentkover1, William Rogers4 1 RTI International WGHI, San Francisco, CA, USA 1 Brown University, Providence, RI, USA 2 FHI360, Durham, NC, USA 2 Alpert Medical School of Brown University/The Miriam Hospital, 3 University of Washington, Seattle, WA, USA Providence, RI, USA 4 University of Connecticut, Brighton, MI, USA 3 The Miriam Hospital, Providence, RI, USA 4 Tufts Medical Center, Boston, MA, USA Background: Given suboptimal adherence to pre-exposure prophy-

ABSTRACTS laxis (PrEP) in some HIV prevention trials, the VOICE Adherence Background: In previous work, we used detailed cognitive testing to Strengthening Program (VASP) was implemented in May 2011, to develop a new 3-item adherence scale. Herein we describe valida- improve adherence counseling in MTN-003, an ongoing placebo- tion work that compares self-reports with an objective adherence controlled study of daily oral or vaginal tenofovir (TFV)-based PrEP. measure, electronic drug monitoring (EDM). Methods: To assess VASP acceptability and effect, anonymous base- Methods: We have enrolled 65 persons in a 4-month observational line (N = 82) and one-year follow-up (N = 75) surveys were conducted validation study. Eligibility criteria include adults taking antiretrovi- with counseling staff (counselors/pharmacists) at 15 VOICE sites, and rals (ARVs) and at least 1 other regular non-ARV medication who compared using GEE. Baseline (N = 18) and final (N = 26) qualitative have had a recent non-suppressed viral load (i.e., some evidence of interviews were also conducted with purposively sampled staff at 13 a possible adherence problem). Participants were enrolled at a base- sites and analyzed for key themes. Of VOICE participants in the active

POSTER line visit and trained in the use of the MedSignals Device, which has arms (N= 365), the proportion with detectable plasma TFV within 6 4 medication bins and thus allows up to 4 medications to be electron- months pre/post VASP, was compared using McNemar’s test. ically monitored simultaneously. They had 3 follow up visits at approximately monthly intervals, at which time they were asked to Results: Overall, staff were more likely to ‘strongly like’ VASP com- self-report on their adherence to the medications in each of (up to) 4 pared to the previous counseling approach (OR = 2.66, 95% CI = 1.32- bins. In this abstract we report the findings from the first bin (an ARV) 5.35). Counselors expressed greater satisfaction with VASP compared for the 26 patients who had completed 3 follow up visits as of to pharmacists throughout. Perspectives of successful counseling December 2013. The time frame for the self-report items is the past shifted towards participant engagement, identifying or overcoming 30 days, and we therefore matched the EDM data to the same 30 barriers rather than focusing on perfect adherence. Further, 72% of days for these analyses. We used a previously published measure of staff thought participants preferred VASP to the previous approach. “covered time” to summarize the EDM data, and compare the 2 Nevertheless, no difference in prevalence of detectable plasma TFV measures by subtracting the SR from the EDM measure. was noted among participants pre and post VASP (p = 0.8). Results: Mean participant age was 49 years, 54% were non-white, Conclusions: While the majority of staff preferred VASP and thought and 35% were female. Mean SR adherence was 83.4, and mean EDM that participants preferred VASP, no changes in participants’ biolog- adherence was 78.3 (100 = perfect for both). The mean of the differ- ical product use were observed. Interpretation of results is compli- ence scores (SR-EDM) was 4.5 points (p = 0.07). cated by the proximity of VASP implementation to early closure of oral and vaginal TFV study arms due to futility. Modifying adherence Conclusions: In these preliminary analyses the overestimation support approaches in the midst of clinical trial follow-up is challeng- shown by the SR methods was relatively small (4.5 points), providing ing, and careful consideration to include all staff should be made. preliminary evidence of the validity of the SR scale. Analyses of the full data set will be available by June 2014.

64 9th International Conference on HIV Treatment and Prevention Adherence 272 1 Moussa Sarr improved adherence. Larger studiesarerequiredtoseeiftheseactuallymediatorsof Differences indepressionduringtheinterventiondidnotpersist. sisted foranadditional24weeksoncephonecallswereterminated. study demonstratedsignificantreductionsinsubstanceusethatper- Conclusions: Results: “any druguse.” and 48weeks.Substanceusewascombinedtocreateavariableof emergency services)collectedbyACASIatbaseline,6,12,24,36, utilization (includingmedical,mentalhealth,casemanagement, reported onsubstanceuse(ASSIST),depression(BSI)andservice In additiontoHIVRNAandself-reportedadherence(VAS), subjects 10 consecutivecallswerediscontinuedfromtheintervention(N=7). missing morethan20%ofcallsfor2consecutivemonthsor could utilizetheirownphone(subsidizedwith$45monthly).Subjects Intervention subjectswereprovidedfreecellphonesandplansor brief callsfor24weeksbyunlicensedadherencefacilitators. 18) orMondaythroughFridaycellphonesupport(n=19) adherence (mean33%)wererandomizedtoreceiveusualcare(n= Methods: Background: 4 3 2 Marvin Belzer Abuse, andServiceUtilization Intervention onDepression,Substance Youth Non-AdherenttoART: Impactofthe self-reported serviceutilization. Intervention andcontrolgroupsshowednosignificantdifferencesin week intervention(P=0.0222)butnotat48weeks0.1666). reported lowerdepressionthancontrolsubjectsattheendof24- = 0.0151).Controllingforbaselinedepression,interventionsubjects intervention subjectsthancontroloverthe48weekstudy(p Controlling forbaselineuse,anydrugusewassignificantlylowerin American, and54%acquiredHIVsexually(46%perinatally). service utilization. the impactofthisinterventionondepression,substanceuseand viral loadamongHIVinfectedadolescents.Theseanalysesevaluate significantly improvedself-reportedadherenceandreductionsin 9th InternationalConference on Wayne StateUniversity, Detroit,MI,USA WESTAT, Rockville,MD,USA National InstituteonDrugAbuse,Bethesda,MD,USA Los Angeles,CA,USA Children’s HospitalLosAngeles/UniversityofSouthernCalifornia, Mean agewas20years,62%weremale,70%African Thirty-seven youthages15-24withmarkedbaselinenon- 3 An experimentalpilotstudyofcellphonesupportfound , SarahThornton The provisionofcellphonesupportinthissmallpilot Study ofCellPhoneSupportfor A Randomized,ControlledPilot 1 (presenting) , LeslieClark 3 , JohannaOlson HIV T reatment andPrevention Adherence 1 , ShoshanaKahana 1 , SylvieNaar-King 2 , 4 278 1 Joshua Metlay cal outcomes. when designinginterventionstoimprovehealthbehaviorsandclini- Andersen’s BehavioralModel.Thismodelshouldbeconsidered adherence thatcanbeclassifiedwithinthebroaddomainsof from thedataastheyarecollected. methodology thatinvolvesiterativedevelopmentofthemesemerging were analyzedforthemesusingagroundedtheoryapproach, perceived qualityofthepatient-providerrelationship.Interviewdata and attendingclinicvisits,obtainingadheringtoART, andthe open-ended questionsassessingbarriers/facilitatorstoscheduling Philadelphia betweenAprilandNovember2013.Interviewsincluded Background: 4 3 2 Carol Holtzman Conclusions: Results: Methods: ( (89%) werenonwhite.Twenty (56%)patientswereretainedincare ed adults( Robert Gross in HIVcareandantiretroviraltherapy(ART)adherence. Andersen’s BehavioralModelto2HIV-specific behaviors-retention and providerfactors)influencehealthbehaviors.Weapplied need) andthecontextualhealthcareenvironment(clinic,system, acteristics (predisposingfactors,enablingandperceived Model mayprovideaframeworkforunderstandinghowpatientchar- Andersen’s BehavioralModel behaviors), andmentalillness(bothbehaviors). cal symptoms(bothbehaviors),cognitivefunction/memory behaviors), adversemedicationeffects(ARTadherenceonly),physi barriers/facilitators were:competinglifeactivities(relatedtoboth retention, 6toARTadherence,and13bothbehaviors.Thetop5 domains ofAndersen’ retention inHIVcareandARTadherencecorrespondingwithall7 to theinterviewdate.Participantsidentified24barriers/facilitators pressed (allHIV-1 RNA<400copies/mL)inthe12-monthperiodprior (94%) wereonARTfor ³ 1 visitineach6-monthperiodoftheyear, atleast60daysapart),34 Massachusetts GeneralHospital,Boston,MA,USA Philadelphia FIGHT, Philadelphia,PA, USA University ofPennsylvania,Philadelphia,PA, USA Temple University, Philadelphia,PA, USA Among patientsinterviewed,20(56%)weremaleand32 Qualitative semi-structuredinterviewswith36HIV ³ 2 18-years-old) enrolledincareat3HIVclinics We identified24barriers/facilitatorstocareandART HIV CareandARTAdherenceto Mapping BarrierstoRetentionin , JanetHines For HIV-infected populations,Andersen’s Behavioral 4 1 , BalighYehia (presenting) s BehavioralModel.Ofthese,5relatedto ³ 1 month,and24(67%)werevirallysup 2 , KaramMounzer , JudyShea 2 2 , KarenGlanz 3 , RafikSamuel 2 , 1 -infect , - - - POSTER ABSTRACTS 65 Correlation between Use of Barriers to Retention in HIV Care: A 279 Antiretroviral Adherence 280 Comparison of Patient and Provider Devices by HIV-Infected Youth and Plasma Perspectives HIV RNA and Self-Reported Adherence Baligh Yehia1 (presenting), Aaloke Mody1, Leslie Stewart1, Carol Holtzman2, Janet Hines1, Karam Mounzer3, Joshua Metlay4, Parya Saberi1 (presenting), Kenneth Mayer2, Eric Vittinghoff1, Judy Shea1 Sylvie Naar-King3 1 University of Pennsylvania, Philadelphia, PA, USA 1 University of California, San Francisco, CA, USA 2 Temple University, Philadelphia, PA, USA 2 Harvard University, Boston, MA, USA 3 Philadelphia FIGHT, Philadelphia, PA, USA 3 Wayne State University, Detroit, MI, USA 4 Massachusetts General Hospital, Boston, MA, USA Background: Our objective was to investigate antiretroviral (ARV) Background: Prior studies have described differences between adherence device use (i.e., labels, calendars, pillboxes, beepers, ABSTRACTS patient and provider-identified barriers to healthcare engagement, timers, MEMS, programmable wrist watches, and diaries) by HIV- including weight loss, prenatal care, and adherence to diabetes infected youth and assess the association of device use and viral treatment. We compared patient and provider perceptions of barriers suppression and self-reported ARV adherence. Additionally, we to retention in HIV care. examined this association among 3 sub-groups: 1) perinatally vs. behaviorally HIV-infected youth, 2) those reporting vs. not reporting Methods: Qualitative semi-structured interviews with 36 HIV-infect- forgetting to take ARVs as an adherence barrier, and 3) those reported adults ( 18 years old) and 13 providers at 3 HIV clinics in ing vs. not reporting problematic substance use. ³ Philadelphia between April and November 2013. Interviews included open-ended questions assessing barriers/facilitators to scheduling Methods: We conducted a secondary data analysis of cross-section- and attending clinic visits. Interviews were analyzed for themes al data collected as part of Adolescent Trials Network 086 and 106 using a grounded theory approach. We compared the relative impor-

POSTER and included data from 1,317 HIV-infected individuals 12-24 years old tance placed on themes between patients and providers. prescribed ARVs. Psychosocial information using audio-computer assisted self-interviews and biomedical data via blood draw or med- Results: Among patients interviewed, 20 (56%) were male, 32 (89%) ical charts from HIV-infected youth receiving care at 20 Adolescent nonwhite, and 20 (56%) retained in care ( 1 visit in each half of the Medicine Trials Units were collected from 2009 through 2012. ³ 12-month period prior to the interview date, ³60 days apart). Of providers interviewed, 12 were physicians and 1 was a nurse practi- Results: Mean 7-day adherence was 86.1%, 50.5% had an unde- tioner. On average, patients identified 7 (range 3-16) and providers 16 tectable viral load, and pillbox was the most commonly endorsed (range 13-24) barriers to retention in HIV care. Barriers frequently device. No specific device was independently associated with higher offered by both patients and providers were mental illness, cognitive odds of 100% adherence. Having an undetectable viral load was function/memory, competing life activities, and poor clinic experi- inversely associated with use of adherence devices (OR = 0.80; p = ences. Compared to patients, providers were more likely to identify 0.04) and among those with <100% adherence, higher adherence was substance abuse (100% vs. 22%), lack of health insurance (92% vs. associated with the use of at least 1 device (coefficient = 7.32; p = 33%), and structural issues - difficulty navigating the health system 0.003). The route of HIV infection, forgetting to take ARVs as an adher- (100% vs. 27%), transportation (92% vs. 35%), and inconsistent hous- ence barrier, and problematic substance use did not modify the asso- ing (77% vs. 19%) - as barriers. Patients were more likely than ciation between undetectable viral load and use of specific devices. providers to offer physical symptoms and negative patient-provider interactions as barriers. Conclusions: Our data suggest that youth who experienced virologic failure often used adherence devices, which may not have been suf- Conclusions: Perceived barriers to retention in HIV care differed ficiently effective in optimizing adherence. Therefore, future between patients and providers. Providers identified more barriers research should concentrate on approaches of improving adherence and more commonly cited substance abuse, lack of insurance cover- that are above and beyond adherence devices alone. Other tailored age, and structural issues as obstacles to care, while patients adherence-enhancing methods may need to be considered to maxi- focused on their physical symptoms and the patient-provider rela- mize virologic suppression and decrease HIV transmission. tionship. These differences should be considered when designing tools and strategies to improve retention in HIV care.

66 9th International Conference on HIV Treatment and Prevention Adherence 281 patients retainedandnotinHIVcare. broader understandingofrelevantfactorsthatdifferbetween retention inHIVcare.Amorequalitativeapproachwouldprovidea limited numberofpatientandhealthsystemfactorsassociatedwith Retained inHIVCare Conclusions: Results: Methods: Background: 4 3 2 1 Baligh Yehia patients incare. for designingmoreeffectiveinterventionstoretainHIV sistently engagedincare.Thesefindingshaveimportantimplications and identifiedmoreindividual-levelbarrierstocarethanthosecon- placed onthemesbetweenpatientsretainedandnotincare. grounded theoryapproach.We comparedtherelativeimportance relationship. Interviewdatawereanalyzedforthemesusinga attending clinicvisitsandtheperceivedqualityofpatient-provider ended questionsassessingbarriers/facilitatorstoschedulingand to theinterviewdate,atleast60daysapart.Interviewsincludedopen- Carol Holtzman Judy Shea substance abuse). of familyandsocialsupport,HIVstigma,physicalsymptoms, offer individual-levelbarrierstocare(competinglifeactivities,lack to retainedpatients,thosenotincareweremorelikely barriers weresimilarlyendorsedbybothgroups.However, compared provider ficulty schedulingappointments,challengingclinicexperiences)and range 3-16)thanretainedindividuals(5,3-10).Clinic-level(dif (range 3-16).Patientsnotretainedincareofferedmorebarriers(8, mission risk.Overall,patientsidentifiedanaverageof7barriers the majoritybeingmale,nonwhite,andwithheterosexualHIVtrans defined asattending Philadelphia betweenAprilandNovember2013.Retentionwas HIV 9th InternationalConference on Massachusetts GeneralHospital,Boston,MA,USA Philadelphia FIGHT, Philadelphia,PA, USA Temple University, Philadelphia,PA, USA University ofPennsylvania,Philadelphia,PA, USA -infected adults(20retainedand16notretained)at3HIVclinicsin -level (perceivedqualityofthepatient-providerrelationship) Both groupshadsimilardemographiccharacteristics,with Qualitative semi-structuredin-personinterviewswith36 1 1 Patients RetainedandNot Barriers toCareDifferfor Patients notretainedinHIVcarehadagreaternumber Multiple cohortandsurveystudieshaveevaluateda (presenting) 2 , JanetHines ³ 1 visitineachhalfofthe12-monthperiodprior , LeslieStewart 1 , KaramMounzer HIV T reatment andPrevention Adherence 1 , AalokeMody 3 , JoshuaMetlay 1 , -infected 4 , - - 282 1 Methods: HIV-positive individualsparticipatinginanationalonlinesurvey. care utilization(HCU)purposesandHIVclinicaloutcomesamong Internet useandthecorrelationofforhealth- Background: Parya Saberi Conclusions: Latino OR=0.67;overallp-value0.02,comparedtoWhite). purposes wasassociatedwithrace(AfricanAmericanOR=0.63, reporting excellentadherence(p=0.001).UseoftheInternetforHCU reporting anundetectableVL(p=0.009)anda1.49-foldoddsof who usedtheInternetforanyHCUpurposeshada1.52-foldoddsof (mean hoursonInternetusinganydevice/day=5.2hours).Those phone nearlydailyorforInternetaccessinthepastmonth Approximately 57.7%ofparticipantsstatedusingtheirmobiletele- phone accessand71.5%reportedowningasmartphone. ed InternetuseforHCUpurposes.96.2%statedhavingmobiletele- tectable VL,66.5%reportedexcellentadherence,and55.5%report American, and91.9%homosexual/bisexual.84.8%reportedunde- age= 45.6years,94.4%male,71.1%White,12.8%Latino,8.9%African Results: ical appointments). healthcare providers[HCPs],refillingmedications,ormakingmed asked abouttheirInternetuseforHCUpurposes(includingemailing viral adherencerating(excellentvs.

POSTER HIV+ 10-19 year olds in Botswana. Electronic monitor (MEMS) data additional supportive messaging, and assistance with problem solv- were downloaded and viral loads were obtained every 3 months. We ing by an interventionist. English-speaking patients with a cell phone calculated percent adherence over each 30-day interval for each capable of sending/receiving text messages, age ³ 18, who are subject and used a piecewise linear mixed effects model to evaluate either: newly entering into care, re-engaging with medical care after the linear adherence trend for each interval. We separately evaluat- a lapse >1 year, or have a history of non-adherence are eligible to ed the trend among adolescents who had <95% adherence during participate. any 3-month interval. Results: From January-August 2013, 58 participants were referred and 32 enrolled. The majority are male (72%),white (66%), non- Results: We enrolled 300 adolescents (158, 53% female) of median age Hispanic (81%), median age 38 (range 19-64), insured (72%), pre- 13.4 (IQR 11.8-15.6) years and median treatment duration of 91 (IQR 64- scribed antiretroviral therapy ( 94%) with 25% newly diagnosed, 6% 105) months. At baseline, 47 (16%) had detectable viral loads without re-entering care, and 69% with a reported history of treatment non- evidence of drug resistance. Of those, 22 (47%) suppressed their viral adherence. Participants chose confidential pre-selected (69%) or loads by 6 months without a regimen change. 143 subjects (48%) had self-created (31%) medication and appointment text-message adherence <95% during 1 or more 3-month intervals. First month medi- reminders. Eight participants opted for a third text message e.g., an adherence was 100% (IQR 96.7-100), declining to 98.3 (IQR 93.3-100) reminders for substance support meetings, smoking cessation, or by month 9 in the complete cohort. Among the 143 with adherence supportive messaging. Participant-initiated messages include ever <95%, month 1 median adherence was 98.3 (IQR 90-100), declin- inquiries on clinic appointments (5), study visits (13), clinic contact ing to 91.7 (IQR 81.3-96.7) by month 9. There was a significant decline information (2), and notifications of number change (4). To date, 5 par- in adherence for each month after the first 3 months (p <0.001) in both ticipants are lost to follow-up with 1 out-of-state transfer, 1 death, analyses and a significant inflection point during month 2. and 1 incarceration. Nine participants have pending final visits. Conclusions: In our cohort, adolescent virologic outcome was tem- Conclusions: The use of a low-cost bi-directional mHealth interven- porarily improved by enrollment in an observational adherence study. tion among patients at high risk for disengaging with medical care is This was likely due to the high rates of adherence observed early feasible at our clinic given prevalence of texting cell phone use and after enrollment. Due to early improvement in adherence caused by acceptability of intervention. the Hawthorne Effect, short-term studies showing high levels of adherence in adolescents may fail to demonstrate the poor adherence seen over time, particularly for those who are least adherent.

68 9th International Conference on HIV Treatment and Prevention Adherence 286 patients athighestriskfordisengagingwithHIVmedicalcare. messaging interventiontoimprovetreatmentadherenceamong pilot studywastoinvestigatetheacceptabilityofabidirectionaltext are engagedintreatmentandcareissuboptimal.Thepurposeofthis ment, thenumberofpersonslivingwithHIVinUnitedStateswho Aadia Rana 1 est riskfordisengagingwithHIVmedicalcare. potential toimprovetreatmentadherenceamongUSpatientsathigh Conclusions: Results: Methods: Background: 3 2 Jacob vandenBerg for thematicanalysis. quently enteredintoNVivo 9anddatawerereviewedinaggregate transcripts toresolveanycodingdiscrepancies.Codesweresubse researchers codedeachtranscriptseparatelyandthencompared audio-recorded andtranscribedverbatim.Two independent capable ofreceivingandsendingtextmessages.Interviewswere gaging withmedicalcare;English-speaking;andownedacellphone in 12monthsofentryintocare,ahistorynon-adherence,orreen- if theywere18yearsofageorolder;receivedanHIVdiagnosiswith- piloting abidirectionaltextingintervention.Participantswereeligible in thenortheasternUnitedStateswhocompleteda6-monthstudy 15 HIV ed messagingwasrare. tinue receivingmessagesuponstudycompletion.Participant-initiat- received bythestudyinterventionist,andexpressedadesiretocon experiencing anemotionalconnectiontothemessagesthey tiality asthemessagesdidnotmentionHIV. Manyalsoindicated the sametimeeveryday, andbeinglessconcernedaboutconfiden- text messages,likingthefrequencyofreceivingonly1messageat Most reportedlikingtheoptionofcreatingtheirownpersonalized medications dailyandtokeeptheirscheduledmedicalappointments. vention tobeacceptableasithelpedthemremembertaketheir with HIVMedicalCare Patients atHighestRiskforDisengaging Improving Treatment Adherenceamong 9th InternationalConference on Providence, RI,USA Alpert MedicalSchoolofBrownUniversity/TheMiriamHospital, The MiriamHospital,Providence,RI,USA Brown University/TheMiriamHospital,Providence,RI,USA -infected patientsfromacommunity-basedHIVcliniclocated Emergent datarevealedthatallparticipantsfoundthisinter- Semi-structured, in-depthinterviewswereconductedwith 3 Text messagingisaviablesupportivetoolthathasthe Despite considerableadvancesinantiretroviraltreat- Text MessagingInterventionfor Acceptability ofaBidirectional 1 (presenting) HIV , EricLamy T reatment andPrevention Adherence 2 , CurtBeckwith 3 , - - - 288 Glenn Wagner depression anddosetimingadherence(path Background: 3 2 Jessica Magidson Conclusions: Results: Methods: relationship betweendepressivesymptomsandARTnon-adherence. and ARTnon-adherence-asapotentiallymodifiablemediatorofthe study testedlifestylestructure-afactorrelatedtobothdepression important toinformbehavioralinterventiondevelopment.Thecurrent studies havetestedfactorsthatunderliethisrelationship,whichis depression andantiretroviraltherapy(ART)non-adherence,few .026, -.003]). between depressionanddosetimingadherence( and lifestylestructuresignificantlymediatedtherelationship longer significantwhenlifestylewasincludedinthemodel(p=.61), Mediation Analysis of LifestyleStructureinaLongitudinal ence todailyactivitiesmayimprovebothadherenceanddepression. structure, increaseroutinizationofdailyactivities,andlinkadher- behavior. Interventionsthataimtominimizedisruptionslifestyle tor intherelationshipbetweendepressionandARTadherence lifestyle structureanddosetimingadherence(path strapping. tested usinggeneralizedlinearmixed-effectsmodelingandboot depressive symptomswereassessedviaself-report.Mediationwas the degreeoforganizationandroutinizationdailyactivities- antiretroviral monitoredbythecap.Lifestylestructure-definedas of prescribeddosestakenwithinthespecifiedtimewindowfor using electronicmonitoringcapsandcalculatedasthepercentage controlled forinallanalyses.Dosetimingadherencewasmeasured HIV evaluating acognitivebehavioralinterventionforadherenceamong over 48weeks.CCTG578wasa3-armrandomizedcontrolledtrial California CollaborativeT 1 tionship betweendepressionandlifestylestructure(path the studyperiodwas78.2%(SD=24.5).Thereasignificantrela- 30% wereHIVantiretroviralnaïve.Meandosetimingadherenceover RAND Corporation,SantaMonica,CA,USA Boston, MA,USA Massachusetts GeneralHospital/HarvardMedicalSchool, Massachusetts GeneralHospital,Boston,MA,USA -infected individualsstartingorre-startingART Participants (n=199)were20%femaleand49%Latino(a). HIV-infected individualsstartingorre-startingARTinthe Lifestyle structuremaybeanimportantmodifiablefac- Dose Timing Adherence?TheRole How DoesDepressionImpactART Despite thewell-documentedrelationshipbetween 3 1 (presenting) reatment Group(CCTG)578wereassessed , AaronBlashill c , p<.05).Path 2 , SteveSafren ␥ . Studyarmwas =-.013, 95%CI[- b , p<.01),and a , p<.0001), c was no 1 , - POSTER ABSTRACTS 69 HIV and Maternal Mortality: Depression as a Predictor of 4-Day 289 Factors Responsible for Deaths 290 ART Treatment Interruptions and Virologic Detectability in an Online Social Clara Appiagyei Nkyi1 (presenting), Daniel Ansong1, Sampson Ofori2, Samuel Nguah Blay1 Network Survey of People Living with HIV

1 Komfo Anokye Teaching Hospital, Kumasi, Ghana John Sauceda1 (presenting), Mallory Johnson2, Parya Saberi2 2 Koforidua Regional Hospital, Koforidua, Ghana 1 Center for AIDS Prevention Studies, San Francisco, CA, USA Background: HIV/AIDS is among the leading causes of maternal mor- 2 University of California, San Francisco, CA, USA tality in sub-Saharan Africa and infectious diseases like tuberculosis, malaria, pneumonia, and sepsis are major contributing factors. We Background: The objective of our study was to examine the associa- sought to determine the trend in HIV related maternal mortality and tion between self-reported depressive symptomatology, 4-day anti- factors associated with it in a regional hospital in Ghana. A retro- retroviral treatment (ART) interruptions, and viral suppression in a spective study of maternal deaths in a cohort of HIV-infected women national cohort of online social media users. ABSTRACTS was carried out between January, 2010 and December 2012. A facility-based maternal death review using case records and mortality Methods: Using a cross-sectional study design we collected data from summaries was systematically conducted. Maternal deaths which HIV-positive individuals via an online survey advertised on Facebook. were associated with HIV/AIDS and its complications that occurred The 2 primary outcomes were a 4-day treatment interruption (none vs. within the 3-year study period were critically reviewed. Factors asso- at least 1 in the past 3 months), and reporting a detectable or unde- ciated with the maternal deaths were identified. tectable viral load (VL). Depression, via the Patient Health Questionnaire-9, was our primary predictor. Participants were catego- Results: Maternal mortality for the 3-year period (per 100,000 live rized as “minimum,” “mild,” “moderate” and “moderate/severe” by births) was 763 (95%CI: 628-918). Year-specific mortality rates were diagnostic cutoff scores. Age, race/ethnicity, gender, sexual orienta- 2010, 2011 and 2012 were 828 (95%CI: 583-1141), 664 (95%CI: 454-937) tion, and substance use were controlled for in 2 logistic regression

POSTER and 797 (577-1074) respectively. There was neither a statistically models (1 per outcome). detectable difference between the yearly morality rates (p = 0.616) and detectable trend (p = 0.904). The proportion of the maternal mor- Results: The mean age of our participants (N = 1,218) was 44 years, with tality that was associated with HIV infection was 5.4% (2/37), 12.5% 78% being White, 12% Latino, 10% African American, and 93% (4/32), 13.9% (6/43), and 10.7% (12/112) for 2010, 2011, 2012 and all gay/bisexual. Approximately 14% of the sample reported at least 1 years combined. There was no significant changes in the proportion treatment interruption and 10% reported a detectable VL. Groups based of HIV-associated deaths (p = 0.435) as well as a significant trend on self-reported depression severity were as follows: 37% “minimum,” seen (p = 0.244). The mean age of the HIV maternal deaths was 30.25 30% “mild,” 15% “moderate,” and 17% “moderate/severe.” Each suc- years (range 18-42 years). Out of a total of 12 HIV maternal deaths, 6 cessive depressive group (i.e., from “mild,” to “moderate/severe”) was had undergone prevention of mother-to-child transmission (PMTCT). associated with greater odds of treatment interruption (ORs = 2.4, 2.2, All the HIV-related maternal deaths were admitted to the hospital at 2.7, respectively, ps <.01), compared to the “minimum” group. The an advanced. 4 persons in the study population had received anti- “moderate” and “moderate/severe,” but not “mild” group, had greater retroviral treatment (ART) whilst 6 were never started on ARTs. odds of having a detectable VL (ORs = 1.7, 1.8, respectively, ps <.05), Tuberculosis was the commonest cause of death (4). The other caus- compared to the “minimum” group. Increased odds for interruptions es of death were cerebral toxoplasmosis (3), septicemia (1), chronic also emerged for Latinos and African Americans (ORs = 2.0, 2.4, respec- diarrhea (1), severe anemia (2), and pneumonia (1). tively, ps <.05), compared to Whites and participants reporting drug use (OR = 1.8, p <.01). Conclusion: Inadequate implementation of the PMTCT and delays in ART in qualified women was observed to be common in HIV mater- Conclusion: Depression, as well as race/ethnicity and substance nal-related death. Stringent measures should be put in place to ade- use, was predictive of missing 4 consecutive days of ART over a 3- quately address the implementation of PMTCT HIV preventive pro- month period. Addressing mental health needs in HIV care is vital to grams to help decrease the burden of the disease. reducing the risk for virological rebound and drug resistance via treatment interruptions.

70 9th International Conference on HIV Treatment and Prevention Adherence 291 1 Norma Ware A QualitativeAnalysis Treatment PrograminNorthCentralNigeria: Conclusion: Results: Methods: Background: 4 3 2 Grace Kolawole treatment settingsinsub-SaharanAfrica. different fromobstaclesreportedpreviousforlarge,centralized cles topersistingincare.Someofthesearesimilarandsome barriers toretentionbutpresentsotherstructuralandsocialobsta- ing particularlyonretentionbarriersandfacilitators. study ofpatientexperiencesdecentralizedcareinNigeria,focus- receiving careindecentralizedfacilities.We conductedaqualitative know littleofpatients’experiencestransferringtonewsites,or for severalyearsinAfrica,andretentionrateshaveimproved.Y health caresettings.Decentralizationeffortshavebeenunderway expanding deliveryofARTfromtertiarytosecondaryandprimary care seekstolinkpatientsservicesandimproveretentionby Becky Genberg resulting stigma. long clinicwaitingtimes,and(d)heightenedfearofdisclosure that hadbeenprovidedfreeofchargeatthepreviouscaresite,c) points (b)havingtopayformedicinesreceivednon-HIVailments including: (a)traveldifficultiesduetopoorroadsandsecuritycheck retention facilitators.Anumberofbarrierswerealsopointedout clinic appointmentsandexperiencesofimprovedhealthfromARTas analyzed toidentifythemes. ceived qualityofservicesreceived.Datawereinductivelycontent riences ofdecentralizedcare,keepingclinicappointmentsandper- tion topicsincludedreasonsfortransferringtothelocalclinic,expe- individuals participatedinfocusedgroupdiscussions.Datacollec community hospitalstookpartinindividualinterviews.Another16 larger, centralHIVspecialtyclinicandwerereceivingARTatlocal north centralNigeria.56(N=56)adultswhohadtransferredfroma based sitesinadecentralizedHIVcarenetworkPlateauState, 9th InternationalConference on Harvard MedicalSchool,Boston,MA,USA AIDS PreventionInitiativeNigeria,Abuja,Nigeria Brown University, Providence,RI,USA Jos UniversityTeaching Hospital,Jos,Nigeria Participants reportedsavingmoneyandtimeontransportto We reportdatacollectedat1public,privateand2faith Decentralization alleviatestransporttimeandcostas 4 Retention inaDecentralizedHIV Barriers andFacilitatorsof Decentralization ofantiretroviraltreatment(ART)and 2 , OcheAgbaji 1 (presenting) 1 , NancinDadem , AtieneSagay HIV T reatment andPrevention Adherence 1 , ProsperOkonkwo 1 , PatriciaAgaba et we 1 , 3 , - - 292 encouragement fromreligiousfaith. ness andadheringtoART ing consciouscommitmentstoremaininghealthybyacceptingill- ing literaterelativesforconfirmationofappointmentdates;(3)mak- tance fromfriendsandrelativesfortransportationcosts;(2)consult ties. Patientsrespondedtochallengesby:(1)seekingfinancialassis- during therainyseason;andcivilinstabilityinneighboringcommuni of patients’homestocaresites;impassableruralroadsandbridges increased fearofdisclosureandsubsequentstigmaduetoproximity lenges includedilliteracy, resultinginmistakenappointmentdates; lenges. Forsome,transportationdifficultiespersisted.Otherchal- Background: 4 3 2 Nancin Dadem Conclusions: Results: scribed andinductivelyanalyzedtoconstructdescriptivecategories. with continuous,long-termcare.Interviewswererecorded,tran- missing scheduledclinicappointments,andstrategiesforcoping experiences ofcareatinitialanddecentralizedsites,reasonsfor for seekingtransfer, perceptionsofaccessingcareclosertohome, ing hospital.Interviewandfocusgroupquestionsaddressedreasons decentralized sitesfromalargerspecialtyclinicatanurbanteach- patient focusgroups.Participatingpatientshadalltransferredto ed of42individualinterviews(patientsN=30;staff12),and3 care sites-1public,1privateandfaith-basedfacility Methods: Norma Ware persist indecentralizedcarePlateauState,centralNigeria. in Africa.Thisqualitativestudydescribesstrategiespatientsadoptto experiences ofkeepingappointmentsatdecentralizedcarefacilities ing travelbarriers.We knowlittle,however, aboutpatients’actual seeks toimproveretentionbybringingcarecloserpatients,reduc- HIV/AIDS requirelifelongmedicalmanagement.Decentralization to overcomethesebarriersandpersistincare. ly exploitcollectivesocialreserveandpersonalreligiousinclination inated throughdecentralizationofHIVtreatment.Patientsdeliberate 1 Health FacilitiesinNigeria Persisting inLong-Term CareinSecondary Becky Genberg Harvard MedicalSchool,Boston,MA,USA AIDS PreventionInitiativeNigeria,Abuja,Nigeria Brown University, Providence,RI,USA Jos UniversityTeaching Hospital,Jos,Nigeria Despite decentralization,patientsfacedretentionchal- The presentationreportsdatacollectedat3decentralized 4 Barriers topatientretentionarereduced,butnotelim- Treatment: PatientStrategiesfor Retention inDecentralizedHIV Patients receivingantiretroviraltreatment(ART)for 1 2 , OcheAgbaji (presenting) ; and(4)derivinginnerstrength , GraceKolawole 1 , ProsperOkonkwo 1 , PatriciaAgaba 3 , AtieneSagay . Dataconsist 1 , 1 , - - - - POSTER ABSTRACTS 71 High Levels of Stress, Depression Patient-Centered Survey in a 293 and other Psychiatric Disorders 296 Hepatitis C Virus (HCV) and Human among HIV-Infected Individuals Immunodeficiency Virus Coinfected Population Lotte Rodkjaer1 (presenting), Maria Slot2, Charlotte Gabel1, Tinne Laursen1, Nils Christensen3, Jens Holmskov4, Rebecca Sumner1 (presenting), Eugene Morse1, Joshua Sawyer1, Peter Leutscher1, Morten Sodemann5 Anthony Martinez2 1 Department of Infectious Diseases, Aarhus University Hospital, 1 University at Buffalo, Buffalo, NY, USA Aarhus, Denmark 2 University at Buffalo/Erie County Medical Center/Buffalo General 2 Department of Infectious Diseases, Odense University Hospital, Medical Center - Kaleida Health, Buffalo, NY, USA Odense, Denmark 3 The Research Clinic for Functional Disorders and Background: Among individuals living with HIV in the United States, Psychosomatics, Aarhus University Hospital, Aarhus, Denmark approximately one-quarter are coinfected with Hepatitis C virus ABSTRACTS 4 Department of Psychiatric Diseases, Odense University Hospital, (HCV). HIV accelerates the natural history of HCV infection with rates Odense, Denmark of fibrosis progression nearly 3 times greater in the coinfected popu- 5 Department of Infectious Diseases, Odense University, Odense, lation. Incorporation of HCV treatment to antiretroviral therapy com- Denmark plicates medication regimens, introducing additional adverse effects and drug interactions which may have a potential impact on patient Background: HIV-infected individuals are at high risk of depression quality of life. Integration of patient-centered goals and values into and other psychiatric disorders. In Denmark there is a lack of infor- medical decision-making allows individuals to assess the risks and mation about mental health in the clinical records of HIV-infected benefits associated with treatment outcomes, allowing for patient- individuals, hence we know very little about the prevalence of psy- specific provider education and intervention prior to treatment initia- chiatric disorders in this patient group. Previous studies have shown tion. Patient-centered outcomes research assessing HCV knowledge that psychiatric disorders are associated with non-adherence to

POSTER and treatment expectations in the coinfected population is warrant- antiretroviral therapy (ART). The aim of this study was to investigate ed to identify potential adherence barriers and improve successful the prevalence of depression and other psychiatric disorders among treatment outcomes. HIV-infected in 2 outpatient clinics in Denmark and to detect factors associated with risk of depression. Methods: Approximately 44 HCV/HIV-coinfected adults in an urban, immunodeficiency clinic will be administered a survey during an ini- Methods: In 2013, a population of 501 HIV-infected individuals were tial pre-HCV treatment evaluation appointment. Twenty-two survey included in a questionnaire-based study. The Beck Depression questions will be independently completed by the patient and will Inventory II (BDI-II) was used to assess the prevalence and severity address previous HCV treatment experiences, expectations and of depressive symptoms. HIV-infected individuals with a BDI-II 20 ³ anticipations regarding HCV treatment and its adverse effects, as were offered a clinical evaluation by a consultant psychiatrist. well as regimen complexity with current antiretroviral treatment. A Results: Symptoms of depression (BDI >14) was observed in 167 subset of patients initiated on HCV treatment during the study period (33%) patients, and symptoms of moderate to major depression (BDI will be administered an abbreviated survey at treatment weeks 4 and 12 to compare current health status to initial treatment perspectives. ³20) in 111 patients (22%). Of all patients 93 (19%) reported having other psychiatric disorders. Among the 111 patients at risk of depression 65 (59%) patients consulted a psychiatrist and the psychiatrist Results: This study is in the active enrollment phase and identifica- found that 71% had a psychiatric disorder other than depression. In tion of coinfected patient subjects for study participation is ongoing. multivariate logistic regression self-reported stress, self-reported Study results will be finalized with abstract modification for poster bad health, not satisfied with life, previous psychiatric history and presentation. substance abuse, non-adherence to ART, were independently associated with symptoms of depression (BDI 20). Conclusions: The results of this study will identify general knowledge ³ deficits on HCV disease and current treatment modalities. The data Conclusions: The study found that Danish HIV-infected individuals are gathered will be utilized to create a patient-centered educational at high risk of depression and have other psychiatric disorders. module delivered prior to HCV treatment initiation, derived directly Depression and stress reduces adherence, hence HIV clinics should from patient-identified responses and focused concerns. routinely screen and treat depressive symptoms to provide full evaluation, and accurate psychiatric diagnosis. We suggest that simple questions such as self-reported health and satisfaction with life could be used in clinical practice to identify patients at risk of depression.

72 9th International Conference on HIV Treatment and Prevention Adherence 299 variables andthechisquaredtestforcategoricalvariables. were comparedusingtheWilcoxonranksumtestforcontinuous vs anxietysymptoms/panic).Personswithandwithoutchronicpain 18 months);depression(PHQ-9 mometer vs somewhat/unabletoperform);overallhealthstate(EuroQOLther clinic. Thesefactorswereimpairedphysicalfunction(EuroQOL,none chronic paininasequentialsampleofpatientsanacademicHIV least mildchronicpainfor>3monthsandfactorsassociatedwith HIV-infected patients. pain for>3months.ThisstudyquantitativelyevaluatedtheBCPSin severity inthepastweekandanotherquestiononpresenceof Brief ChronicPainScreeningTool (BCPS)includes1questiononpain the generalpopulation,orindividualswithHIV. Anew2-question widely usedscreeningtooltoidentifypatientswithchronicpainin tant syndromeofchronicpainisunknown.To date,therehasbeenno infected patientsiscommon,theprevalenceofclinicallyimpor- beyond theperiodofnormaltissuehealing.WhilepainamongHIV- 1 vs. 10%,p<0.001). 17%, p<0.001),depression(30%vs.15%,=0.08),andanxiety(43% overall healthstate(70vs.84,p=0.002),consistentpain(65% p <0.001),difficultywithusualactivities(47%vs12%,lower chronic painweremorelikelytohaveimpairedmobility(43%vs.12%, the BCPSasascreeningtoolforchronicpaininHIVcaresettings. depression, andanxiety. Thisstudy providespreliminaryevidencefor measures offunctionalimpairment,healthstate,consistentpain, Conclusions: Results: Methods: Background: 2 Jessica Merlin Stefan Kertesz Michael Saag in HIV-Infected Individuals 9th InternationalConference on University ofCalifornia,SanFrancisco,CA,USA University ofAlabamaatBirmingham,AL,USA , 0-100);consistentpain(painon2/2EuroQOLsoverthepast Of 100participants,30hadchronicpain.Participantswith We evaluatedtheassociationbetweenpresenceofat The resultsoftheBCPScorrelateashypothesizedwith Brief ChronicPainScreeningTool Quantitative EvaluationofaNew 1 Chronic painisofatleast3monthsduration, 1 , MelonieWalcott 1 (presenting) , AndrewWestfall ³ HIV 1 10); andanxiety(PHQ-Anxiety, none , ChristineRitchie T reatment andPrevention Adherence 1 , EricChamot 2 , 1 , - 300 Methods: Background: 2 Kellie Watkins Re-Linkage toCare Conclusions: provider referralswereout-of-care. referrals, 68.4%ofhealthdepartmentand50.0%medical eligible forSL (3.3%) wereineligible,and50persons(55.6%)out-of-care returned tocareontheirown,8(8.8%)wereout-of-jurisdiction,3 and 20frommedicalproviders.Twenty-nine individuals(32.2%) rals camefromsurveillancedata,19healthdepartmentstaff, Results: 1 Biru Yang assistance toreturncare. out intervention,wereout-of-jurisdiction,incarcerated,orneeded conducted todetermineifthesepersonsreturnedHIVcarewith- Houston/Harris CountyTexas werelosttocarein2012.Astudywas will beexaminedinfutureanalyses. while inandoutofcare,thedatesresultsCD4viralloads impact ofSLWs andfurtherunderstandthehealthstatusofpatients while providerreferralsidentifiedthelowest(50.0%).T rals identifiedthehighestproportionofout-of-carepatients(68.4%) towards individualin-needofassistance.Healthdepartmentrefer- confirm thecurrentcarestatusofpatientshelpstargetSLW efforts jurisdictions. Utilizinglaboratorydatafromsurveillancerecordsto care patientsisaninnovativeapproachthatcanbeappliedinother for re-linkagetocare. appointment, theywereassignedtoServiceLinkageWorkers (SLWs) previous 6monthsand/orweremissingevidenceofamedical their own.IfpatientslackedCD4countorviralloadresultswithinthe bases determinedwhetherthesepersonshadreturnedtocareon tion orwereincarcerated.Searchesin4surveillanceandcaredata- determined whetherthesepersonshadrelocatedtoanotherjurisdic health departmentfieldstaff.Informationcollectedfrom2databases tive patients:surveillanceprogramdata,medicalproviders,and TX, USA University ofTexas HealthScienceCenteratHouston, TX, USA Houston DepartmentofHealthandHumanServices,Houston, Ninety patientswereincludedinthestudy. Fifty-onerefer- 1 Three sourcesidentifiedpotentiallyout-of-careHIV-posi- , MarciaWolverton Identifying Out-of-CarePersonsfor Innovative Approachesfor The creationof3referralmethodstoidentifyout-of- 1 Forty-nine percentofHIV-diagnosed personsin W intervention.Fifty-threepercentofsurveillance , CamdenHallmark 1 , MarleneMcNeese 1 , Richard Grimes 1 2 (presenting) o measurethe , - POSTER ABSTRACTS 73 Applying Motivational Interviewing HIV-Related Stigma and ART 302 to Enhance Engagement in Care in 304 Adherence in a Clinical Cohort: Patients with HIV in Buenos Aires, Argentina Does the Impact of Stigma Differ for Men and Women? Lina Bofill1 (presenting), Stephen Weiss2, Alejandra Bordato3, Graciela Fernandez-Cabanillas3, Mar Lucas4, Analia Dorigo4, Bulent Turan1, Sheri Weiser2, Mallory Johnson2, Greer Burkholder1, Deborah Jones2 Anne Zinski1, James Raper1, David Batey1, Michael Mugavero1, Janet Turan1 (presenting) 1 University of Miami, Miami, FL, USA 2 University of Miami Miller School of Medicine, Miami, FL, USA 1 University of Alabama at Birmingham, Birmingham, AL, USA 3 Fundacion Helios Salud, Buenos Aires, Argentina 2 University of California, San Francisco, CA, USA 4 Fundacion Huesped, Buenos Aires, Argentina Background: HIV-related stigma may affect women differently than Background: Motivational interviewing (MI) has been shown to pro- men, contributing to disparities in adherence and health outcomes by ABSTRACTS mote behavioral change in different health related situations. gender. However, the use of MI among physicians providing HIV health care have has not previously been conducted in Argentina. This study Methods: We analyzed observational data from a cohort of patients examines provider responses to MI training, uptake and ongoing use who initiated HIV care at a clinic in the US South after 2011. Data on of this specialized application of MI to improve engagement and HIV-related stigma (internalized, enacted, community, and disclosure retention in care for challenging patients who have not responded to concerns) were collected during patients’ clinic orientation session usual care. (new-to-care or re-engagement). A validated self-report measure of ART adherence in the past month was administered at each subse- Methods: An MI intervention was offered as a 1-day workshop for quent clinic visit over 30 months. ART adherence data were available physicians (n = 12) in public and private settings in Buenos Aires, for 349 men and 83 women at their first clinic visit, and for 269 men Argentina. A questionnaire was completed immediately post-training

POSTER and 48 women at subsequent visits. Logistic regression analyses pre- and 1 week post-training, to evaluate MI uptake and implementation. dicting adherence were adjusted for gender, race, age, baseline CD4, Physicians also video recorded a routine consult session with a chal- and years since HIV diagnosis. lenging HIV-positive patient pre- and post-MI training, to evaluate the utilization of the MI techniques. Results: The proportion of clients reporting less than excellent ART adherence was 44% for the first visit and 50% for subsequent visits Results: The health care providers expressed strong interest and sat- (average of up to 10 visits). Internalized, enacted, and total stigma isfaction immediately following their participation in the work shop. scores were negatively related to ART adherence assessed at first One week post-training, all were committed to using MI strategies in and subsequent visits. Compared to men, women had higher enact- their consultation sessions. Based upon their post-workshop ed, community, and total stigma, and lower ART adherence. Effect responses, most physicians (n = 8) described using several MI tech- sizes for the relationships between stigma dimensions and ART niques, 2 were ambivalent about the techniques; 1 declined to adherence measured at first visit were larger for women (OR range = respond. All wanted to receive additional MI training. Preliminary .32-.80) compared to men (OR range = .75-.95), with significant inter- analyses (5 of 7 video recordings) illustrated appropriate utilization of actions between gender and total and enacted stigma (p = .03). For MI techniques (active listening, short summaries, positive feedback, ART adherence measured at subsequent visits, effect sizes remained guiding more than directing). Two providers used MI scales (1 to 10) larger for women. For women, internalized and total stigma were to quantify their patient’s willingness to be adherent. negatively associated with both ART adherence measured at first visit (OR = .54, 95%CI = .30-.999; OR = .32, 95%CI = .13-.77, respective- Conclusions: MI was well received among HIV health care providers. ly) and at subsequent visits (OR = .39, 95%CI =.15-.998; OR = .28, However, the process of HIV management is lengthy and providers 95%CI = .08-1.00, respectively). may need more intensive training and practice using MI techniques to fully incorporate their routine use in care. MI appears to be a Conclusions: HIV-related stigma at baseline has stronger effects on viable strategy to enhance engagement and retention in care among ART adherence for women as compared to men, which may be a challenging patients. driver of more negative outcomes for women living with HIV.

74 9th International Conference on HIV Treatment and Prevention Adherence 307 all primaryvariablesexceptmissed visit proportionacrosstimethat measures analysesofvariancerevealed significantimprovementsin and moodqueriesweresimilar, rangingfrom62%-67%.Repeated covariate inanalyses.Responseratestosubstanceuse,adherence, dition, requiringusingbaselinealcoholdisorderdiagnosisasa much higherrateofalcoholproblemsanddiagnosesintheText con- Randomization failedtocontrolforalcoholusedisorders,witha portion ofdaysusingalcoholordrugsatbaselinewas53%. visit proportionforthepast6monthswas30%atbaseline.Thepro previous 90dayswas64%andbyTLFB63%atbaseline.Missed measured bytheMINI.Adherencepharmacyrefillratefor bidity inthesample,predominantlydepressiveandanxietydisorders and 30randomizedtoT reported ontheTLFB. measured bymissedvisitproportion,andsubstanceusingdays and T outcomes includedadherencemeasuredbypharmacyrefillrate weeks, followedbya3Mand6Mfollowupassessment.Primary cies duringthetrainingsession.Theinterventionperiodwas12 use. Theycreatedpersonalizedmessagesforvariouscontingen- (TAU). Text participantsweregivenastudyphoneandtrainedonits randomized tothetextingcondition(T ing informedconsent,completedabaselineassessment,andwere to protectpatientprivacy. Participantsenrolledinthestudyfollow- times, andsubstanceuseinthepast24hoursusingacodedquery medication takingatdosingtimes,moodtwicedailyrandom and substanceuse/nouse.Thetoolsentdailyqueriesof own reportsofadherence/non-adherence,goodmood/poormood, which includespatient-createdpersonalizedmessagesfortheir Nonurban SubstanceUsers Messaging ARTAdherenceTool for Results: Methods: Background: 3 2 1 George Reynolds Karen Ingersoll using peoplelivingwithHIV tional textmessagingARTadherencetoolfornonurbansubstance study, ouraimwastodevelopandpilottestapersonalizedbidirec- adherence justintimefordeliveryofinterventions.Inthis bidirectional messagingcouldprovideopportunitiestodetectnon- to textmessagingservices.T Internet ormobilewebservice,theyusuallyhaveadequateaccess challenges. Whilemanynonurbanpatientshavelimitedaccessto adherence, includingstigma,isolation,poverty, andtransportation reside innon-urbanandruralareashavespecificchallengesto Jennifer Hettema 9th InternationalConference on University ofNewMexico,Albuquerque,NM,USA Health DecisionTechnologies, Sheboygan,WI,USA University ofVirginia, Charlottesville,VA, USA imeline Followbackselfreport(TLFB),treatmentengagment 63 peopleparticipatedinthetrial,with33randomizedtoText Following formativeresearch,wedevelopedthetool, Personalized BidirectionalText Feasibility andImpactofa Patients withHIVrecentsubstanceusewho 1 (presenting) 2 3 , JasonFreeman AU. Therewassignificantpsychiatriccomor . , RebeccaDillingham ext messaginghasabroadreachand HIV 1 , Avi Hosseinbor T reatment andPrevention Adherence ext) ortotreatmentasusual 1 , 1 , - - these returnnearlytobaselinelevelsat6M. improvement inadherence,engagement,andsubstanceuse, and powerwaslimitedinthissmallsample.Follow-upat3Mshows of changefortheotherprimaryoutcomesfavoredText group, have atimeeffectoraninteractionovertime.However, rates ures ANOVA showedthatMVPdifferedbetweengroups,butdidnot Text and.31fortheTAU condition;t=2.58;p<.02.Therepeatedmeas- did notdifferbytreatmentassignment.MVPat3Mwas.09forthe Conclusions: powered subsequenttrial. ise toimprovebothadherenceandsubstanceuse,meritsafully responded toall3typesofqueries.TheText conditionshowsprom- adherence toolengagesparticipantsforthefull12weeks,andthey A personalized,bidirectionaltextmessagingART

POSTER ABSTRACTS 75 Retention in HIV Care among CD4 Response in Pediatric Patients 308 a Medicaid Insured Population, 309on HAART with Sustained Virologic 2006-2010 Control

Kathy Byrd1 (presenting), Melissa Furtado1, Tim Bush1, Lytt Gardner1 Janice Piatt1 (presenting), Rukmalee Vithana2, Laura Clarke-Steffen1 1 Centers for Disease Control and Prevention, Atlanta, GA, USA 1 Phoenix Children’s Hospital, Phoenix, AZ, USA Background: Using surveillance and other available data, the 2 Children’s National Medical Center, Washington, DC, USA Centers for Disease Control and Prevention estimates that 51% of HIV-positive persons are retained in medical care. Health services Background: Previous studies of CD4 response have followed mixed claims data may also be useful for estimating levels of retention in cohorts of adherent and non-adherent children; pre- and post-highly care, particularly in insured HIV populations. active antiretroviral therapy (HAART). This study evaluated the CD4 response in highly adherent children with sustained viral suppres-

ABSTRACTS Methods: We used the 2006-2010 Truven Health MarketScan® sion on HAART. Medicaid claims databases to identify HIV/AIDS cases, and to determine levels of retention in care and gap in care. We used ICD-9 Methods: Retrospective chart review of HIV-positive patients, ages Clinical Modification diagnosis codes to identify cases, restricted 0-21 years, on HAART with CD4 counts and viral loads available prior analyses to persons who: 1) were ³18 years (in the case identifica- to HAART initiation, as well as at least 2 data points per year. Patients tion year) 2) had ³10 months of continuous enrollment in Medicaid must have maintained viral suppression with only 1 viral load dis- during each 12-month period of a 24-month measurement period and crepancy >1,000 copies allowed for the duration of the study. CD4 3) had ³1 office visit claim in the first 6-month period of follow-up. response was followed from initial treatment for as long as patients Procedure and provider codes were used to identify office visits. remained suppressed. Retention in care was defined as ³1 office visit claim with a clinical provider during each 6-month period of the 24-month measurement Results: Thirty-five patients remained adherent and virologically sup- POSTER period, and gap in care as the absence of an office visit claim in the pressed after initiation of HAART, 17 male, 18 female, ages 6 weeks last 6 months of the first 12-month period. Procedure and drug codes to 17 yrs (mean 4.58 yrs) with 6 acquired, 29 vertical transmissions. were used to determine if cases with a gap in care received other Patients were followed at least 2 yrs up to 15 yrs (mean 6.44 yrs). HIV care related services. Eight patients (22.8%) had greater than 10 years of viral suppression. Baseline CD4 counts ranged from 25-4400 (mean 1045), CD4% 2.8-28.7 Results: Between 2006-2010, 4,894 cases met the study criteria of (mean 26.68); viral loads 4,700 to >750,000 (mean 283,048). Peak CD4 whom 54.4%, on average, (range 47.0%-56.6%) were retained in care counts were dependent on initial CD4 count and age at initiation of during the 24-month measurement periods and 23.4%, on average, therapy. Plateaus in the CD4 graph began between 2.5 and 5.5 years (range 20.6%-30.1%) failed to be retained in care due to a gap in care after the start of HAART. Independent of age and initial CD4 count, in the first 12 months of follow-up. Of cases who experienced a gap CD4 % stabilized and increased throughout 15.5 years of therapy. All in care, 5.8% had claims for HIV viral load and/or t-cell tests, or 22 patients suppressed >5 years of therapy achieved >25% CD4. All 8 claims for antiretroviral therapy (ART) during the gap. patients suppressed >10 years achieved 30% and 4 patients with data from 12-15.5 years were above 32%. Conclusions: Our estimates for retention in care among a Medicaid- insured HIV population were similar to national estimates. Among cases Conclusion: Some patients had virologic control over greater than experiencing a gap in care, approximately 6% continued to be moni- 10-15 years of treatment. Unlike previous studies, this study suggests tored through laboratory tests, or continued to fill ART prescriptions. that children who remain adherent to HAART can experience ongoing immune healing for as many as 15 years.

76 9th International Conference on HIV Treatment and Prevention Adherence 311 and identifyadditionalhigh-riskcommunities. ect interventionstrategies,HIVrisk-transmissionbehaviorchange system. Dynamicallygeneratedgraphsandtablesinformedonproj- tracking cardsandanovelweb-basedhealthinformationexchange for outreachtoassesstestingandchangeincareengagementvia forums. Innovativemonitoringandmeasuringstrategieswereused were developedanddeliveredviacommunity-basedevents to otherraces(OR=0.536,CI:0.439-0.655,p Black/African raceswerelesslikelytobetestedovertimecompared conducted andengagementincarebetween20062009. T Lives (P-TAS) Collaborative byPeopleTake ActionSave Conclusions: year-old groupovertime. Outreach effortstofemalesshowedanincreaseintestingthe Those mostlytestingpositivewerenon-Hispanicandage tested wasachievedduetoHispanic-communityoutreachefforts. CI:1.207-1.729,p were morelikelytogettestedcomparednon-whites(OR=1.44, paratively largerdeclinesintestingratesovertime.Whiteraces Results: Methods: Background: 3 2 Dean Wantland 1 implemented toimproveoutreach,access,andengagementcare. faith-based andminorityorganizations,localcolleges,P-T sortium oflocal/statehealthdepartments,industryrepresentatives, routine HIV care. P-TAS isaninnovativemodeldevelopedbyacon- who over-utilized emergencydepartmentsandwhodidnotaccess developed inNewJerseytodecreasethenumberofPLWH of color based participatoryresearch(CBPR)voluntaryhealthcollaborative, “treatment asprevention”efforts. tion andthereductionof“communityviralload,”promoting through acomprehensive,community-wideapproachtoHIVpreven- tion thattesting,treatmentinitiationandadherencecanbeachieved approach toHIVprevention.Thestudyprovidesempiricalinforma 9th InternationalConference on AS demonstratedanincreaseinnumbersandratesofHIVtests Queens, NY, USA NYC HealthDepartmentof&MentalHygiene, RDE Systems,Clifton,NJ,USA Rutgers CollegeofNursing,Newark,NJ,USA Of 1,197tested,52(4.3%)newHIVcaseswereidentified.P- Targeted messagingtomembersofhigh-riskcommunities People Take People ActionSaveLives(P-TAS) isacommunity- Promotion-Disease Prevention A Five-Year HIV/AIDSHealth The P-TAS modeliscriticaltoatest-and-treat 1 ² , 0.001). Inyear4,anincreaseinHispanicsbeing Jesse Thomas 2 HIV (presenting) T reatment andPrevention Adherence ² 0.001) alsoshowingcom- , CatherineMiranda AS was ² ³ 24- 45. 3 - 313 1 Sunday Abidoye as clientattendingatleast1clinicvisitin2012). follow-up visit,clientARTstatusandretention(defined basic demographicinformation,CD4count,WHOstage,numberof SPSS version16.Atotalof621clientrecordswerereviewedfor health facilityfrom2005to2012werereviewedandanalyzedusing who wereenrolledinARTCareandsupportunit(HIVClinic)ofthe Federal MedicalCenter, Ido-Ekiti,Nigeria.Medicalrecordsofclients Methods: Background: 2 Oluwole Babatunde Federal MedicalCentreIdo-Ekiti,Nigeria tion ARTprogram. other factorsaffectingretentionfromdiagnosistolong-termreten- adopted tomaximizeARTbenefits.Afollow-upstudyisneededon strategies tosustainandimprovementretentionrateshouldbe condition formaintainingorrestoringhealthinthelongrun.The Conclusions: cells/mm pre-ART client,HIVstage1andIVclientbaselineCD4above400 value 0.027).Poorretentionwasassociatedwithdecreasingage, baseline CD4(P-value0.004),yearofdiagnosisandARTinitiation(P- cally significantwithage(P-value0.031),ARTstatus0.000) (37%) werenotretainedinthisstudy2012.Retentionwasstatisti- Results: possible determiningfactorsinPLHIVonART. (ART). Theobjectiveofthisstudyistoassessretentionratesand program despitetheexpansionofaccesstoantiretroviraltherapy poor performanceandincreasedmorbiditymortalityinHIV/AIDS Akure, Ondo,Nigeria Ondo StateBranchofWorld HealthOrganization, Federal MedicalCenter, Ido-Ekiti,Ekiti,Nigeria A totalof347(63%)patientswereretainedincareand208 3 . It isadescriptive,cross-sectionalstudyconductedin Poor retentionofpatientsincareisamajordriver HIV CareandTreatment inthe Seven-Year ReviewofRetentionin Retention incareofPLHIVisaminimumnecessary 2 1 (presenting) , LawrenceAyodele 1 ,

POSTER ABSTRACTS 77 High Rates of Early Linkage to HIV Self-Reported Adherence to 314 Care in a Facility-Based HIV 316 Pre-Exposure Prophylaxis and Counseling and Testing Unit in Jos, Nigeria Sexual Behavior by Text Messaging: Preliminary Findings from the Partners Olawumi Dinchi1 (presenting), Patricia Agaba2, Fadeware Hauwa1, Oche Agbaji2, Imade Godwin1, Ugoagwu Placid1, Muazu Awal1 Mobile Adherence to PrEP Study Jessica Haberer1 (presenting), Jared Baeten2, Renee Heffron2, 1 Apin Juth, Jos, Plateau, Nigeria Nelly Mugo3, Elly Katabira4, Kenneth Ngure5, Njambi Njuguna6, 2 Jos University Teaching Hospital, Jos, Plateau, Nigeria TImothy Muwonge4, Lara Kidoguchi2, Courtney Ng1, David Bangsberg1

Background: Globally, a substantial number of HIV-infected persons 1 Massachusetts General Hospital, Boston, MA, USA are unaware of their status and some of those who are aware are 2 University of Washington, Seattle, WA, USA unable to establish linkage to care services or remain engaged in care. 3 KEMRI, Nairobi, Kenya Our objectives were to assess rates of early linkage to HIV care from 4 Infectious Disease Institute, Kampala, Uganda ABSTRACTS a facility-based HIV counseling and testing (HCT) unit and to determine 5 Jomo Kenyatta University of Agriculture and Technology, associated factors in a tertiary health institution in Jos, Nigeria. Nairobi, Kenya 6 Partners in Prevention, Nairobi, Kenya Methods: We reviewed electronic medical records of HIV-positive adults referred from the facility-based HCT unit to the antiretroviral Background: Effectiveness of PrEP is highly dependent on adher- (ARV) clinic between January 2010-December 2012 ( n =1096). We ence, which has been associated with sexual activity. Self-report of defined early linkage to care as having a clinic visit or laboratory test these behaviors is often limited by recall and social desirability bias- (CD4 count or viral load assay) within 30 days of diagnosis. Gender es, which may be overcome via text messaging (SMS). differences were explored for variables and regression analysis with crude and adjusted odds ratio (OR) and 95% confidence interval (CI) Methods: The Partners Mobile Adherence to PrEP Study is a sub- was used to identify factors associated with early linkage to care. study within the Partners Demonstration Project, which involves pro-

POSTER vision of comprehensive HIV prevention services to high-risk Results: Of 1,095 HIV-positive referred patients enrolled, there were serodiscordant couples in East Africa. In the substudy, daily SMS 703 females (64.7%) and the median age was 32 years (IQR = 16-75 surveys are sent to HIV-uninfected partners taking PrEP for 14 days years). The males were significantly older than the females (37+9 vs. around each scheduled study visit. An airtime incentive (~US$0.50) is 32+8, p <0.001). 604 (56.4%) patients were married and 146 (13.6%) provided for each completed survey. had no formal education. The median CD4 and viral load were 177 Results: As of December 2013, 225 (65%) of the 348 HIV-uninfected 3 cells/mm and 41,532 copies/mL, respectively at baseline, with 578 participants initiating PrEP have been screened and 187 (83%) (55.4%) having CD4 count ²200. There was no difference in CD4 (177 enrolled in the substudy. Reasons for ineligibility are no personal vs. 184, p = 0.73) and viral load (42 782 vs. 40 765, p = 0.97) levels phone with reliable charging source (N = 8), illiteracy/inability to between female and male patients respectively. 955 (87.2%) were send SMS (N = 16), not taking or planning to continue PrEP at sub- linked to care and had both CD4 and viral load testing within 30 days study recruitment (N = 10), and other/refusal (N = 4). Technical chal- of HIV diagnosis. 32 (2.9%) and 13 (1.1%) were linked within 31-60 and lenges include temporary server outages and variable network avail- 61-90 days respectively. 10 (0.9%) were linked beyond 91-180 days of ability. This table presents data per participant: diagnosis. 86 (7.8%) were missing linkage data. Median time to linkage was 8 days (IQR: 1-189), there was no difference in time to link- SMS surveys age between females and males (8 days, IQR: 1-164 vs 8 days, IQR: 1- N completing >1 survey to date 125 189, p = 0.05). Following uivariate analysis using age (>35 years or Mean N (SD) surveys 13.3 (6.6) Mean % expected surveys completed 76.1% younger-OR-0.93, 95% CI: 0.65-1.35), marital status (married vs other- Adherence OR-1.65, 95% CI: 1.15-2.35), level of education (none vs other- OR-0.66, Median (IQR) 100% (77.8-100%) 95% CI:0.41-1.07), occupation (unemployed vs other-OR-0.66, 95% CI: Mean (SD) 83.0% (27.4) 0.66-1.07), gender (male vs female-OR-1.18, 95% CI:0.81-1.74) and CD4 N (%) with <80% adherence 33 (26.4%) count (<200 vs. >200-OR-0.95, 95% CI:0.67-1.36) as independent vari- Mean % non-adherence and no sex 76.4% ables, a regression model fitting the same variables was developed. Sexual activity in previous 24 hours Marital status (married vs. other-aOR: 1.70, 95%CI: 1.19-2.44) was the Mean % surveys reporting sex 36.2% Of surveys reporting sex, % also reporting only predictor of early linkage to care. PrEP use 83.0% Condom use 68.0% Conclusions: Early linkage to HIV care and laboratory testing was PrEP and condom use 60.7% high in this cohort of HIV-positive patients. Further follow-up to No PrEP or condom use 9.7% assess retention in long-term care is necessary for sustained benefits to the patients and their communities. Conclusions: Preliminary findings suggest periodic collection of daily adherence and sexual behavior data from most participants is feasible. Self-reported adherence to PrEP is generally high, although some adherence is suboptimal. Most sex acts are protected by PrEP and/or condoms. Future analyses will determine the validity and factors influencing these data.

78 9th International Conference on HIV Treatment and Prevention Adherence 318 (e.g., identificationsupportmechanismforeachpartner). (PEP), andtreatmentasprevention);coupleproblem-solving (e.g., pre-exposureprophylaxis(PrEP),post-exposure rules andsexualdecision-making);biomedicalpreventionstrategies styles); relationshipstrengthening(e.g.,identificationofunwritten self-care plan);communication(e.g.,useofeffective including self-care(e.g.,informationaboutHIV/AIDS,stimulantuse, ally appropriate.Theadaptedinterventioncoversseveraltopics couple-based approachescanhaveonLEPLatinoMSM. the increasingrelevanceanddemonstratedimpactthatinnovative are associatedwithsexualriskbehavior. Theproposedstudyaddsto research hasidentifiedthatrelationshipdynamicsinmalecouples shown tobeathighriskforHIVandSTIs.Epidemiologicalbehavioral Latino MSMandtheirsame-sexpartners,agroupthatresearchhas ners foranewtargetpopulationoflimitedEnglishproficient(LEP) Black menwhohavesexwith(MSM)andtheirsame-sexpart- providers. shop whichincludedsessionswithkeyinformants:14couplesand10 ing qualitativemethods,includinganinterventionadaptationwork- Zach Shultz Sex withMenandtheirSame-SexPartners English ProficientLatinoMenWhoHave ples isfeasibleandattractive. that acouple-basedHIVpreventioninterventionforgayLatinocou- Conclusions: Results: Methods: Unite” Background: 9 8 7 6 5 4 3 2 1 Omar Martinez Scott Rhodes preventive interventionthatisculturally Juan DelaCruz 9th InternationalConference on Lehman College,NewY Community Activist,NewYork, NY, USA University ofTexas atElPaso,TX,USA USA W Bloomington, IN,USA Indiana UniversityBloomingtonSchoolofPublicHealth, AID forAIDSInternational,NewYork, NY, USA Latino CommissiononAIDS,NewYork, NY, USA Columbia UniversitySchoolofSocialWork, NewYork, NY, USA NY, USA State PsychiatricInstituteandColumbiaUniversity, NewYork, HIV CenterforClinicalandBehavioralStudiesattheNewYork ake ForestUniversitySchoolofMedicine,Winston-Salem,NC, (CNU) We developed This studyfollowedasystematicadaptationprocessutiliz- 1 , HugoOvejero , anexisting,evidence-basedHIV/STIinterventionfor 6 The purposeofthestudywastoadapt Prevention InterventionforLimited Adaptation ofaCouple-BasedHIV The processculminatedwithpreliminaryindications , SilviaChavez-Baray 9 1 (presenting) Latinos enPareja ork, NY 3 , JonathanCapote , ElwinWu HIV , USA 7 , EvaMoya T reatment andPrevention Adherence 2 , linguistically , TheoSandfort , ascience-basedHIV/STI 4 7 , BrianDodge , AntonioPorras , andcontextu 1 “Connect ‘n , 5 , 8 , - 319 Background: 1 Samuel Nwafor Oshodi LagosStateNigeria Nigeria -CaseStudyofLagosIslandand Results: Methods: survey. QuestionnaireswereanalyzedusingSPSSversion16.0. focus groupdiscussionsand2in-depthinterviewsduringanend-line dents wererecruited.Alsoqualitativedatawasobtainedfrom2 ed. Two hotspotswereselectedfromeachLGA,where6to7respon- government areaswhereUSAIDfundedSHiPSprojectisimplement among 120MSMrandomlyselectedfrom7hotspotsacross2local means ofliving(71.9%),andsatisfaction(6.9%). include stigmatization(43.4%),increasedhomophobia(73.2%), ground. Thestatedreasonsforpracticinganalsexunderground country while17.2%MSMwillstayandpracticeanalsexunder- 88.3% ofrespondentsagreedthatMSMmightbeforcetofleethe 99.4% oftherespondentsurgegovernmenttoreviewlaw is inflictingfearsamongMSMandhealthserviceproviders, will increaseHIVprevalence.62.0%ofrespondentsbelievethelaw led amongMSMamidstthelaw with men(MSM).ThisstudyaimstodescribehowHIVcouldbetack among thegeneralpopulationand17%menwhohavesex 2010, thenationalHIVprevalenceinNigeriawasestimatedat4% provide HIVpreventionandtreatmentservicestoLGBTpeople.In gender identity. Thelawalsoprohibitsorganizationsfromworkingto ices, andviolencebasedonrealorperceivedsexualorientation had ledtoincreasedhomophobia,discrimination,denialofHIVserv- which wouldreduceHIVinfection.Thepassageoftheanti-gaylaw anti-gay lawwillincreasemoralsandreduceanalsexpractice Centre fortheRighttoHealth,Ilupeju,Lagos,Nigeria About 91.7%oftheMSMsstronglybelieveanti-gaylaw A population-basedcross-sectionalsurveywasconducted Passage ofSame-SexLawin Public HealthImplicationof The perceivedknowledgebymostNigeriansisthat 1 (presenting) . - - . POSTER ABSTRACTS 79 Strengthening HIV Counseling and The Health, Development, and 320 Testing among Men who have Sex 321Human Rights Implication of the with Men in Lagos through Integration of Same-Sex Law in Nigeria Other Health Services and Home-Based Christy Ekerete-Udofia1 (presenting) Service Delivery 1 Centre for the Right to Health, Ilupeju, Lagos, Nigeria Grace Hygie-Enwerem1 (presenting) Introduction: The news of the passage of the same-sex bill came to 1 Centre for the Right Health, Lagos, Nigeria most Nigerians as a rude shock especially those in development work, the bill passed not minding the fact that Nigeria has the second Introduction: The purpose of this project is to increase demand for largest HIV epidemic globally [UNAIDS] - 2012, there were an esti- uptake of HIV counseling and testing (HCT) by MSM to reduce the mated 3.4 million people living with HIV in Nigeria. In 2010, national incidence of HIV infection among them. The strategy was developed HIV prevalence in Nigeria was estimated at 4% among the general

ABSTRACTS over a period of time in partnership with Department Of Defense population and 17% among men who have sex with men. The recent- funded clinic- a local HIV service centre for MSM community estab- ly signed same-sex law in Nigeria will retard the already made lished by the Population Council Nigeria. progress as concerns the fight against HIV.

Description: It is estimated that there are about 1,496 MSM in Lagos Description: In the 2011 United Nations Political Declaration on state. Focal group discussion conducted across two LGA in Lagos HIV/AIDS, all UN Member-States committed to removing legal barri- shows that 8.7% of MSMs are willing to access HCT on the field ers and passing laws to protect populations vulnerable to HIV and while 92.3% are not willing to access HCT. Most MSMs living in urge governments to protect the human rights of lesbian, gay, bisex- Lagos Island received other medical treatment willingly in a desig- ual and transgender people through repealing criminal laws against nated clinic. About 78% of the survey respondents believed that adult consensual same sex sexual conduct; implementing laws to home-based HCT service delivery is good for MSM due to stigma protect them from violence and discrimination; promoting campaigns POSTER associated with clinic. 84.7% of the respondents believes that MSM that address homophobia and transphobia; and ensuring that ade- will appreciate uptake of HCT if other health services are integrated. quate health services are provided to address their needs. 75% of the respondents believed that there must be something wrong with any MSM going to access HCT. Lesson Learned: The provisions of the law could lead to increased homophobia, discrimination, denial of HIV services and violence Lessons Learned: Most MSM do not like to uptake HCT openly due to based on real or perceived sexual orientation and gender identity the internal stigmatization and discrimination among community which could further increase the HIV epidemic prevalence among members. They think there must be something wrong with whoever this group from present 17.3%. It could also be used against organi- seeks HCT. This situation discourages HCT uptake and contributes to zations working to provide HIV prevention and treatment services to increases in HIV prevalence among LGBT. LGBT people thereby discouraging those in development work.

Recommendations: Home-based HCT which will be integrated with Recommendations: There is an urgent need to review the constitu- other related health services is recommended. This strategy will also tionality of the law in light of the serious public health and human curb the associated effect of anti-gay law which is currently affect- rights implications. The Nigerian authorities and civil society organi- ing public service delivery to LGBT community. zations must strive to ensure safe access to HIV services for all people in Nigeria.

80 9th International Conference on HIV Treatment and Prevention Adherence 324 tice andoutcome. ing knowledgeandskills.Thishasimplicationsforcounsellingprac- hence theneedforcounselorstoacquireandimprovetheircounsel lenges whichareclientspecificrequirecounsellingskills; particularly inPMTCTprogram.HIV-positive statusanditschal- success inHIVpreventionandmanagementofthesero-positives, assessing care,treatmentandsupport.Thisiscriticalinachieving 0% atpost-trainingrating. was alsoadownwardmovementfrom20.5%atpre-trainingratingto post-training ratingwas6.6%.Forthepoorlevelofknowledge,there For moderatelevelofknowledge,pre-trainingratingwas71.1%while knowledge atthepre-trainingratingto84.2%post-trainingrating. movement from0.3%oftheparticipantswithgoodcounselingskills Atiene Sagay Jonah Musa Central, Nigeria Skills’ andCounsellingOutcome,Jos,North Conclusions: Results: Methods: Background: 3 2 Alero Babalola-Jacobs 1 adherence totreatmentandpatientroleincontinuitycare. acceptance ofserostatusanddecisionforpositivelivingincluding effect needstobeexplored.Counsellingoutcomehasimplicationsfor tice andpracticewillaffectoutcome.However, thenatureof improved knowledgeofcounselingskillswillinformprac edge hasimplicationsforcounselingpracticeandoutcome;thatisan were collatedandanalyzed. rated and27participantswereinvolved.Only24completedproforma for good,moderate,andpoor. Atotalof15counsellingskillswere a counsellingskillsproformawithscoringguide.Theratingswere an in-trainingsession.Thiswasdoneforpre-andpost-trainingusing participant ratingofcounsellingskillsknowledgewasconductedin 9th InternationalConference on Jos UniversityTeaching Hospital,Jos,Nigeria University ofJos/JosTeaching Hospital,Jos,Nigeria Teaching Hospital(Juth),Jos,Nigeria AIDS PreventionInitiativeinNigeria(Apin)/JosUniversity Counselling skillsknowledgeratingindicatesanupward This isa2time-pointassessmentsurvey 2 3 , GloriaAngyo The generalimprovementinthecounselingskillsknowl on theKnowledgeofCounselling Impact ofHIVCounsellingTraining HIV testingandcounselingremainsthegatewayto 1 (presenting) 1 , ChineduEkwempu HIV T , RosemaryOmoregie reatment andPrevention Adherence 2 , OcheAgbaji . Forthisphase, 1 , 3 , - - - 325 and drug/alcoholuse. outcome expectancy, futureorientation,psychologicalreactance, cent adherencestudy. We includedassessmentsofself-efficacy, psychosocial assessmenttoolstobeusedinalongitudinaladoles- 1 Karen Glanz Background: 4 3 2 Elizabeth Lowenthal Positive AdolescentsinBotswana may yieldmoreaccurateresults. social desirabilitybias(e.g.,audiocomputer-assisted self-interview) treatment adherence,testingproceduresthatseektominimize In evaluatingtherelationshipbetweenthesemeasuresandHIV difficult foradolescentstoadmit,particularlyaknownquestioner. Conclusions: cigarettes, oranydrugsofabuseintheirlifetimes. second timepoint.Nosubjectsadmittedtoeverhavingtriedalcohol, significantly (p=0.03)moresocially-desirableanswersgivenatthe did notknowthetesterwell(N=25),ICCwaslower(0.21)with ly-desirable reactancescoresatbothtimepoints.Amongthosewho who knewthetesterwell(N=15),ICCwashigh(0.84)withsocial- tance scalehadlowertest-retestreliability(ICC=0.41).Amongthose expectancy andfutureorientationtools.Thepsychologicalreac Test-retest reliabilitywassimilarlyhighfortheself-efficacy, outcome assessed inthetoolswasgoodtoexcellentrange(0.74-0.83). (55%) werefemale.Theinternalconsistencyofallconstructs Results: Methods: Keboletse Mokete scores atthe2timepoints. Kruskal Wallis testwasusedtoassessfordifferencesbetween coefficients (ICC)werecalculatedtodeterminetest-retestreliability. multiple itemsmeasuringthesameconstructs.Interclasscorrelation Cronbach’s alphawasusedtoassessinternalconsistencyof the Individual assessmentsrangedfrom4to11questionsinlength. nurse asthoseshedidornotknowwellattheinitialassessment. cents twice,2-4weeksapart.Participantswereclassifiedbythe University ofPennsylvania,Philadelphia,PA, USA Gaborone, Botswana Botswana-Baylor Children’s ClinicalCentreofExcellence, Botswana, Gaborone,Botswana Excellence/Centers forDiseaseControlandPrevention- Botswana-Baylor Children’s ClinicalCentreof Philadelphia, PA, USA University ofPennsylvania,PerelmanSchoolMedicine, Median ageofsubjectswas12.4years(IQR11.4-13.6),22 A studynurseadministeredallassessmentsto40adoles 4 , RobertGross We culturallyadaptedandsoughttovalidateaseriesof Assessment Tools UsedwithHIV- Social DesirabilityBiasin Psychological reactanceandsubstanceusemaybe 3 , OntibileTshume 1 (presenting) 4 , Tafireyi Marukutira 3 , GabrielAnabwani 2 , 3 , - - POSTER ABSTRACTS 81 Prediction of Virologic Failure Self-Report and Parent-Report 326 among Adolescents Using the 327 Measures Over-Estimate Pediatric Symptom Checklist Adherence among Adolescents in Botswana

Elizabeth Lowenthal1 (presenting), Tafireyi Marukutira2, Elizabeth Lowenthal1 (presenting), Rachel Hammond2, Ontibile Tshume3, Gabriel Anabwani3, Robert Gross4 Knashawn Morales1, Tafireyi Marukutira3, Ontibile Tshume4, Gabriel Anabwani4, Robert Gross5 1 University of Pennsylvania Perelman School of Medicine, Philadelphia, PA, USA 1 University of Pennsylvania Perelman School of Medicine, 2 Botswana-Baylor Children’s Clinical Centre of Philadelphia, PA, USA Excellence/Centers for Disease Control and Prevention- 2 Children’s Hospital of Philadelphia, Philadelphia, PA, USA Botswana, Gaborone, Botswana 3 Botswana-Baylor Children’s Clinical Centre of 3 Botswana-Baylor Children’s Clinical Centre of Excellence, Excellence/Centers for Disease Control and Prevention- Gaborone, Botswana Botswana, Gaborone, Botswana

ABSTRACTS 4 University of Pennsylvania, Philadelphia, PA, USA 4 Botswana-Baylor Children’s Clinical Centre of Excellence, Gaborone, Botswana Background: Psychosocial dysfunction is a risk factor for non-adher- 5 University of Pennsylvania, Philadelphia, PA, USA ence. The Pediatric Symptom Checklist (PSC) is a simple psychosocial screening tool which has previously been associated with a his- Background: Adolescents have lower rates of adherence and higher tory of virologic failure. We sought to determine whether high scores rates of virologic failure (VF) than adults and younger children. Self- on the PSC could predict which adolescents will have virologic fail- report and parent-report adherence measures are commonly used, ure in the next 6 months. but their validity for adolescents with HIV remains poorly studied.

Methods: Parents/guardians of 300 adolescents (age 10-19 years) Methods: 300 adolescents (age 10-19) utilized microelectronic moni- from a longitudinal adherence study were asked to complete the PSC tors (MEMS) and were followed every 3 months for self-reported POSTER at baseline. PSC scores were dichotomized (at ³20 = positive) based adherence, parent-report of adolescents’ adherence and viral load. on prior data. Inability to complete a PSC was classified as a positive Adherence was scored using both the Reynolds Index and last score. Viral loads were obtained at least every 3 months. Virologic missed dose (never, >3 months, 1-3 months, 2-4 weeks, 1-2 weeks, 1 failure was defined as any viral load ³400 copies/mL within 6 months week ago). VF was defined as viral load >400 copies/mL. The associ- of enrollment. Chi-squared test evaluated the difference in virologic ation of each measure with VF was evaluated using generalized esti- failure rates among those with and without positive PSC scores. Total mating equations. PSC scores were evaluated continuously for their association with virologic failure using logistic regression. Results: 3-month data are available for all subjects and 6 and 9 month data for a subset. There are 70 VFs among 59 adolescents. Median Results: 290 parents/guardians completed the PSC. Sixty adoles- MEMS adherence was 95.3% (IQR 84.4-98.4) in those with VF and 98.8 cents (20%) had virologic failure with 14 (23%) having a positive PSC. (IQR 94.6-100) in those without VF. 96.4% of adolescent and 98.2% of 28 (12%) adolescents without virologic failure had a positive PSC, p = adult Reynolds Indices were 100%. 76.6% of adolescent and 73% of 0.02. The odds of failure among those with higher PSC scores were adult reports indicated no missed doses. 3.4% of adolescents and 1.03 (95% CI 0.99-1.07). The positive and negative predictive value of 3.5% of adults reported the last missed dose was >3 months ago. The a positive PSC score predicting virologic failure were 31% and 82%, odds of VF were 2.2 (95% CI 1.3-3.8) for those with MEMS adherence respectively. <95% compared with >95%, p = 0.004, but neither the adolescent nor adult Reynolds Index was associated with VF (OR 1.0, 95% CI 0.6-2.3 Conclusions: In settings in which viral load testing is limited, a posi- and OR 1.0, 95% CI 0.95-1.1, respectively). Compared with those who tive PSC score could help identify adolescents who might benefit reported no missed doses, the odds of VF were 0.2 (95% CI 0.1-0.4) most from virologic testing due to a higher risk of virologic failure. and 0.1 (0.0-0.3) for those indicating missing >3 months ago via self- However, most patients with virologic failure would not be identified or parent report, respectively. by a positive PSC. Conclusions: Self-report and parent-report measures of adherence over-estimated true adherence. Admitting distant imperfect adherence may paradoxically indicate better current adherence. Compared with both subjective measures, MEMS performed better as an adherence measure in this setting.

82 9th International Conference on HIV Treatment and Prevention Adherence 328 based model. training andsupportareintegraltothecontinuedsuccessofthisMI- found itdifficulttotransitionintotheMIstyleofinteraction.Ongoing tion. PNsthathavehistoricallyuseddifferentmodelsofcare, niques whenaddressingclientissuesthatrequireimmediateatten- Linkage andRetentioninCare Recommendations: Lessons Learned: Description: health behaviorsinadiversegroupofclients. drinking, MIhasexpandedtofocusonmedicaladherenceandpro- clients exploreandresolveambivalence.Initiallydirectedatproblem tered styleofcounseling,designedtoelicitbehaviorchangebyhelping trained andcertifiedinMotivationalInterviewing(MI),aclient-cen- and maintainclientaccesstoHIVcarerelatedservices.PNsare Navigation (PN)programsforHIV-positive individuals.PNsfacilitate Virginia DepartmentofHealth(VDH)currentlysupportsPatient (HRSA) SpecialProjectsofNationalSignificance(SPNS)grant,the Introduction: 2 1 Anne Rhodes retaining clientsinHIVcareandservices. tion deliveryandindevelopingsuccessfulapproachesforlinking and resources,itisessentialindocumentingthefidelityofinterven- While theMI-focusedPNprogramrequiresaninvestmentoftime measured toensuresuccessfultranslationfromtheorypractice. effectiveness oftheseapproachesinreal-worldsettingsmustbe care aredevelopedaspartoftheNationalHIV/AIDSStrategy skills withshiftstowardsprotocoladherenceandskillmaintenance. group feedbacktoPNs.TheinitialfocusisontheestablishmentofMI T reviewed byevaluationstaffusingtheMotivationalInterviewing tion inHIVcare.PNsaudiotapeclientsessions,whicharethen and evaluatePNuseofMIskillstosupportclientlinkagereten- University’ 9th InternationalConference on reatment IntegrityCode.Thiscodingsystemguidesindividualand Virginia CommonwealthUniversity, Richmond,VA, USA Virginia DepartmentofHealth,Richmond,VA, USA s InstituteforDrugandAlcoholStudiestotrain,monitor Through aHealthResourcesandServicesAdministration 1 Navigation InterventionforHIV Fidelity MonitoringofaPatient , VDH workedwiththeVirginia Commonwealth Thomas Moore PNs havefounditchallengingtoutilizeMItech- As interventionstolinkandretainpersonsinHIV 2 (presenting) HIV T reatment andPrevention Adherence , the , 331 Adolescents onAntiretroviralTherapy Hilda Mujuru ence riskfactorshavebeenunderstudiedinthedevelopingworld. vention oftransmission.Adolescentsareathighrisk,butnon-adher- Background: 6 5 4 3 2 1 Robert Gross Conclusions: Results: Methods: are desperatelyneeded. adolescent non-adherencemeritsattention;effectiveinterventions groups needmoretestingbeforefurtherinvestment.Highratesof Professional groupcounselingmaybejustified.Peer-led support portive parent/guardianinvolvementmaybeaviableintervention. logistic regressiontocontrolforconfounding. responses withnon-parametrictestsoftrendmultivariable Dichotomous responseswerecomparedwithX2testsandordinal logistics ofclinicvisits,HIVcareattitudes,andservicesprovided. dichotomized asoptimal(excellent)orsuboptimal.We queriedabout good ajobdidyoudowithtakingyourmedicationeveryday;”itwas reported adherencewasratedexcellenttopoorinresponse“How Zimbabwe. Studystaffprovidedanonymousquestionnaires.Self- 19 yearsinHIVcareatParirenyatwaandHarareHospitals 78% traveling<1hour female, 94%livingwithfamily, with74%having>4visitsannually, and not associated.Therewasnoconfounding. X2 p=0.02].Age,sex,andpeer-led supportgroupparticipationwere counseling groupparticipations[optimalvs.suboptimal:65%50%, not atall(0vs.4%),testoftrendp=0.02],andprofessionallyrun [optimal vs.suboptimal:very(70%57%),somewhat(20%29%), vs. 4%),testoftrendp=0.04],comfortaskingproviderquestions suboptimal: very(87%vs.76%),somewhat(14%19%),notatall(0 14%), testoftrendp=0.01],confidenceinadherence[optimalvs. timal: always(56%vs.38%),sometimes(31%48%),never(12% ed withparent/guardianintheroomduringvisits[optimalvs.subop- Zimbabwe London SchoolofHygieneandTropical Medicine,Harare, Philadelphia, PA, USA University ofPennsylvaniaPerelmanSchoolMedicine, Zimbabwe University ofZimbabweCollegeHealthSciences,Harare, REPSSI, Harare,Zimbabwe Biomedical ResearchandTraining Institute,Harare,Zimbabwe University ofPennsylvania,Philadelphia,PA, USA 262 participantsweremedian15.5years(IQR14-17),61% W e conductedacross-sectionalsurveyofadolescents10- 4 1 Reported AdherenceinHarare Factors AssociatedwithSelf- Adherence iskeytoHIVtreatmentsuccessandpre- , ElizabethLowenthal Suboptimal adherencewascommon.Increasingsup- (presenting) . Adherencewasoptimalin39%andassociat , Tsitsi Bandason 5 , RashidaFerrand 2 , LisaLanghaug 6 3 , - POSTER ABSTRACTS 83 Understanding Patients’ Illicit Drug Use and Unmet Needs 332 Experiences with PEPFAR- 333 Sub-Group Effects in the Retention Associated Transitions in c-Care in South in Care Intervention Trial Africa: A Qualitative Study Lytt Gardner1 (presenting), Lisa Metsch2, James Willig3, Richard Moore4, Mari-Lynn Drainoni5, Jessica Davila6, Ingrid Katz1 (presenting), Laura Bogart2, Christie Cloete3, Tracey Wilson7, Faye Malitz8, Allan Rodriguez9, Charles Rose1 Tamaryn Crankshaw4, Janet Giddy3, Tessa Govender3, Elena Losina5, Ingrid Bassett5 1 Centers for Disease Control and Prevention, Atlanta, GA, USA 2 University of Miami, Miami, FL, USA 1 Brigham and Women’s Hospital, Boston, MA, USA 3 1917 HIV/AIDS Clinic and Department of Medicine, University of 2 Boston Children’s Hospital/Harvard Medical School, Boston, MA, Alabama at Birmingham, Birmingham, AL, USA USA 4 Johns Hopkins University School of Medicine, Baltimore, MD, USA 3 McCord Hospital, Durban, South Africa 5 Boston University School of Public Health, Boston, MA, USA 4 Health Economics and HIV and AIDS Research Division (HEARD), ABSTRACTS 6 Baylor College of Medicine, Houston, TX, USA University of KwaZulu-Natal, Durban, South Africa 7 SUNY Downstate Medical Center School of Public Health, 5 Massachusetts General Hospital, Boston, MA, USA Brooklyn, NY, USA 8 Health Resources and Services Administration, Rockville, MD, USA Background: South Africa (SA) was the largest recipient of PEPFAR 9 Miller School of Medicine, University of Miami, Miami, FL, USA funding for antiretroviral therapy programs from 2004-2012. However, PEPFAR funds to SA will decrease by 50% by 2017. These funding Background: The Retention in Care (RIC) trial, a 3-arm randomized changes have led to transfers from hospital and non-governmental trial in 1,838 adults in 6 academic-medical center HIV clinics in the organization-based care to government-funded, community-based United States , was broadly efficacious, yet 2 sub-groups: 1) patients clinics. We examined the impact of clinic transfers on patients’ who actively used drugs, 2) patients with unmet service support healthcare-related attitudes and experiences. needs, appeared to benefit less from the intervention. We formally POSTER assessed these sub-group effects with an interaction term. Methods: We conducted semi-structured interviews with 36 participants (66% women) to assess patient experiences related to transfer Methods: Outcome measure: visit constancy rate: having 1 or more of care from a PEPFAR-funded, hospital-based clinic in Durban, to 3 kept HIV primary care visits in each of 3 4-month intervals. Trial arms: different levels of government facilities (primary health care clinic, a) enhanced contact (2 arms, N =1225; patients received reminder community health care clinic, hospital-based clinic). Data were tran- and missed visit calls, face-face contacts and basic HIV education); scribed, translated, coded, and inductively analyzed. and b) standard of care (SOC: 1 arm, N = 613, patients received usual care. Subgroups: Drug use: use of >1 of 7 illicit drugs (ID) (e.g., crack, Results: Participant narratives revealed the importance of connect- heroin, meth) in the 3 months before enrollment. Unmet needs (UN): edness between patients and clinic staff at the PEPFAR-funded pro- >1 unmet support service needs (e.g., housing, transportation, finan- gram, which was described as respectful and conscientious. cial assistance) in the 6 months before enrollment. Effect measures: Participants reported that transfer clinics were largely focused on Visit constancy risk ratios (RR) were calculated at each level of the dispensing medication and on throughput, rather than holistic care. subgroup variables, and compared with a study variable x interven- Although participants appreciated the free treatment at transfer sites, tion interaction term, fit with a log-binomial model. they expressed frustration with long waiting times and low quality of patient-provider communication, and felt that they were treated dis- Results: For illicit drugs: ID-yes, 44.9/46.5, intervention vs. SOC; ID-no, respectfully. These factors eroded confidence in the quality of the 58.1/45.5, intervention vs. SOC. RR 0.97 vs. 1.28, p = 0.044. For unmet care. The transfer was described by participants as hurried with an needs: UN-yes, 49.1/47.5, intervention vs. SOC; UN-no, 60.5/44.5. RR apparent lack of preparation at transfer clinics for new patient influx. 1.04 vs. 1.37. p = 0.007. The sub-group differences were significant for Formal (e.g., counseling) and informal (e.g., family) social supports, unmet needs and for illicit drugs. both within and beyond the PEPFAR-funded clinic, provided a buffer to challenges faced during and after the transition in care. Conclusions: Compared to standard care, our intervention of Experiences were generally consistent across transfer facility type. enhanced contacting significantly improved visit constancy, an important target for intervention in the treatment cascade. In our Conclusions: Transitions in care are a necessary component of shift- sub-group analysis, patients with unmet support service needs and ing PEPFAR funds in SA. Optimizing retention in care during such patients using illicit drugs were less likely to benefit from the inter- transitions may require moving beyond a response rooted in efficien- vention. These are important groups to now target for additional ele- cy, to one that focuses on sociobehavioral factors, including improv- ments over and above the current intervention approach to improve ing inter-clinic coordination and provider-patient communication, retention in care. while maintaining social support.

84 9th International Conference on HIV Treatment and Prevention Adherence 336 rupted accesstofreeservicesregimenisstronglyrecommended. tion strategiestargetedatreducingpillloadandensuringanuninter compared totheurban.Appropriateadherenceenhancinginterven- in bothtreatmentlocationsbutsignificantlylowertheruralsite Living withHIV/AIDSinSouthernNigeria Antiretroviral Therapy(HAART)amongPeople Conclusion: Results: Level ofsignificancewassetatp<0.05. adherence toHAARTweredeterminedbybinarylogisticregression. based ona1weekrecallbeforeactualinterview. Predictorsof were obtainedonself-reportedmedicationadherencewhichwas secutively recruitedpatientswhomettheinclusioncriteria.Data istered questionnairewasusedtocollectinformationfrom804con- T site) andtheSpecialT ducted amongpatientsonHAARTattendingaGeneralhospital(rural Methods: an urbanandruraltreatmentlocationinCrossRiverState,Nigeria. among peoplelivingwithHIV/AIDS(PLHIV)accessingtreatmentin compare theprevalenceanddeterminantsofadherencetoHAART, remains largelyunexploredinNigeria.Thisstudythereforeaimedto the variousgeopoliticalzonesofcountry, thesouth-southzone Although adherencetoantiretroviraltherapieshasbeenassessedin (HAART) isamajorpredictorofthesuccessHIV/AIDStreatment. Background: 2 1 Afiong Oku Emmanuel Monjok significantly decreasedthelikelihoodofadherencetoHAART. CI: 0.46-0.97]anduseofherbalremedies[OR0.7;95%0.46-0.89] instructions [OR1.5;95%CI:1.06-1.97].However es [OR2.2;95%CI:1.21-4.15.]andtreatmentregimendevoidofdietary ceived improvedhealthstatus[OR3.0;CI:1.50-5.89],freeARTservic ence werefewernumberofchildren[OR1.9;95%CI1.13-3.07],per depressed [urban20%vs.rural12.2%].Themajorpredictorsofadher- getting medications[urban31.1%vs.rural48.9%]andfeeling operating abusyschedule[urban43.8%vs.rural50.6%],simplyfor- rural site(p=0.007).Themajorreasonscitedforskippingdoseswere (59.9%) comparedto50.4%amongstPLHIVaccessingtreatmentinthe recall was55.2%.Thissignificantlyhigherintheurbanpatients 9th InternationalConference on eaching Hospital(urbansite).Asemi-structured,intervieweradmin University ofIbadan,Oyo,Nigeria University ofCalabar, Calabar, Nigeria The overallself-reportedadherenceratebasedona1-week A comparativecross-sectionalanalyticalstudywascon- 1 The adherenceratereportedinthisstudywasquitelow (presenting) on AdherencetoHighlyActive A Rural-UrbanComparativeStudy Adherence tohighlyactiveantiretroviraltherapy 1 reatment clinicoftheUniversityCalabar , OkuOboko HIV T 1 reatment andPrevention Adherence , EmeOwoaje , pillload[OR0.7;95% 2 , - - - - 338 1 Michael Mugavero Background: 4 3 2 Riddhi Modi Depression Treatment Study HIV-Infected PatientsRecruitedfora Conclusions: Results: Methods: International NeuropsychiatricInterview(MINI)amongPLWHA. sion screeningtoolrelativetothereferencestandardMini not well-defined.We evaluatedtheperformanceofPHQ-9depres- PHQ-9 todistinguishdepressionfromotherpsychiatricdisordersis clinical settingstoscreenfordepression.However for depression.ThePatientHealthQuestionnaire-9(PHQ-9)isusedin (ART) adherence.ARTadherenceguidelinesrecommendscreening population. Depressionisassociatedwithpoorantiretroviraltherapy prevalence of20-30%;a3-foldincreasecomparedtothegeneral ple livingwithHIV/AIDS(PLWHA). Depressionismostcommonwith ders. met MDDcriteriawithoutanyevidenceofbipolarorpsychoticdisor teria, 6(2%)didnotmeetMDDcriteria,andtheremaining304(84%) (9%) metbipolardisordercriteria,21(6%)psychoticcri- informed consentforstudyparticipation.Among363participants,32 depressive disorder(MDD),meteligibilitycriteriaandprovided African American)scoredPHQ-9 consent completedMINIfordiagnosticconfirmation. ticipants. PatientswithPHQ-9scores ment onHIVoutcomes.ThePHQ-9wasusedtoidentifypotentialpar and UNC);arandomizedclinicaltrialofeffectdepressionmanage- Antiretroviral ManagementatDuke,UAB,NorthernOutreachClinic, assessed fortheSLAMDUNC(StrategiestoLinkAntidepressantand Melonie Walcott Julie O’Donnell other psychiatricdisorders. screen onitsownmaynotbesufficienttodistinguishMDDfrom remains ausefuldepressionscreeninginstrument,positivePHQ-9 require verydifferenttreatmentapproaches.WhilethePHQ-9 mately 1in6hadindicatorsforbipolarorpsychoticdisorderwhich were confirmedtohaveMDDusingdiagnosticinstrument,approxi- Alabama atBirmingham,AL,USA 1917 HIV/AIDSClinicandDepartmentofMedicine,University Duke University, Durham,NC,USA University ofNorthCarolinaatChapelHill,NC,USA University ofAlabamaatBirmingham,AL,USA Overall 363patients(meanage=63years,70%female,63% Study participantswereHIVinfectedpatientswho 1 Psychiatric illnessoccursatarateof63%amongpeo- While themajorityofpatientsidentifiedbyPHQ-9 (presenting) Depression ScreeningTool among Performance ofthePHQ-9 3 , Teena McGuinness 1 , JamesRaper 1 , BrianPence 1 , AnneZinski 1 ³ , JamesWillig 10, indicatingalikelymajor 2 ³ , BradleyGaynes 10 whoprovidedinformed 1 , MalaikaEdwards 4 , , theabilityof 3 , 3 , - - POSTER ABSTRACTS 85 Psychosocial Factors, Depression, Stressful Caregiving and Coping by 340 Adherence, and Response to 341HIV-Infected Women Antiretroviral Therapy in Persons Living Evelyn Quinlivan1 (presenting), Lynne Messer2, with HIV in the Dominican Republic Peggye Dilworth-Anderson1, Tandrea Hilliard1, Katya Roytburd1

Consuelo Beck-Sague1 (presenting), Carlos Adon2, 1 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA Leonel Lerebours-Nadal3, Ellen Koenig2, Maria Pinzon-Iregui4, 2 Portland State University, Portland, OR, USA Mina Halpern3, Sharice Mc Donald4, Micaela Atkins5, Julia Siegel5, Jessy Devieux6 Background: Limited information exists regarding the caregiving experiences and coping strategies used by HIV+ women of color 1 Robert Stempel College of Public Health and Social Work, (HIV+WOC). Miami, FL, USA 2 Instituto Dominicano de Estudios Virologicos (IDEV), Methods: As part of a retention in care project (Guide to Healing), 106 Santo Domingo, Dominican Republic HIV+WOC attending an academic Infectious Disease Clinic were ABSTRACTS 3 Clinica de Familia La Romana, La Romana, Dominican Republic interviewed from December 2010 through June 2011. A 30-question 4 Robert Stempel College of Public Health and Social Work, assessment of caregiving activities (10 caretaking activities; 10 cul- Miami, FL, USA tural context; 10 coping strategies) was included. Associations 5 University of Massachusetts Medical School, Worcester, MA, USA between caretaking and personal characteristics are described. 6 Florida International University, Miami, FL, USA Results: Most HIV+WOC were African American (80%), >40y.o. (74%), Background: Antiretroviral therapy (ART) can reduce HIV mortality high school graduates (68%), unemployed (70%), and housed perma- as well as sexual and mother-to-child transmission but optimal nently (90%). Few HIV+WOC were married (21%) or insured privately “treatment-as-prevention” may require achievement of unde- (19%). Women provided care to minor children or grandchildren tectable viral loads (UVLs). Psychosocial factors may affect ART (25%); other adults (21%), both adults and minors (25%) or to neither

POSTER adherence and response. (30%). Women reported coping with caregiving using humor (87%), laughter (85%), faith (93%), by seeking help from others (82%), and Methods: To estimate the prevalence and predictors of clinically sig- advance planning (87%). Despite this, women commonly reported nificant depressive symptomatology (CSDS) in DR patients and iden- that they had to physically (25%) or emotionally (20%) remove them- tify psychosocial factors affecting ART adherence and response, the selves from the caregiving situation due to stress. The mean positive Center for Epidemiologic Studies Depression Scale (CES-D) scores, caregiver coping score was 33.0 (SD: 4.5; range: 19 - 40). Difficulties and demographic, psychosocial and clinical data were anonymously with caregiving were also reported: unable to fulfill their caregiving abstracted from records of patients receiving ART in a clinic in the responsibilities due to their health (25%); were unable to ask family DR capital and one in the interior and analyzed. for assistance with caregiving (24%) or skipped or delayed a medical appointment to provide care (16%). However, most reported that they Results: Data abstracted from 205 patients’ records showed that 103 could call someone for help if they were sick (86%). Family expecta- (50.7%) were female. Ages ranged from 19-72 (median = 39.3) years. tions to provide care regardless of health was associated with pri- Age did not vary significantly by sex. Sixty one (29.8%) patients met vate insurance (20% with private insurance vs. 47% without private CES-D criteria for CSDS. CSDS prevalence did not vary by sex or age, insurance (unadjusted OR: 3.48, 1.07 - 11.26) and positive coping but was higher in patients reporting food insecurity than those who (unadjusted beta-coefficient: 1.84, 0.09 - 3.6). did not (37/71 [52.1%] vs. 20/124 [16.1%]; p <.001) and in those resid- ing in “bateyes” (underserved sugarcane plantation worker bar- Conclusions: Despite high levels of coping, 1 in 5 HIV+WOC report racks) than others (8/8 [100%] vs. 48/177 [27.1%], p <.001). Patients significant difficulties in providing care to others. Plans for the aging with CSDS were more likely to describe ART use as “problematic” of this population must include efforts to support HIV+WOC with (13/61 [21.3%] than others 6/144 [4.2%]) (p <.001). Patients who report- chronic caregiving responsibilities. ed imperfect adherence (27/64 [42.2%]), or had CSDS (22/62 [36.1%]) were less likely to have an UVL than those reporting perfect adherence (80/141 [56.7%]; p = .053) and/or no CSDS (85/144 [59%]; p = .003). When controlled for perfect adherence and describing ART as “problematic” in a logistic regression model, CSDS continued to be associated with a >65% decrease in likelihood of UVL (p = .004).

Conclusions: CSDS is associated with food insecurity, problems with ART, living in bateyes, and with a lower likelihood of achieving an UVL in DR patients. Addressing psychosocial factors and depression in these patients may improve adherence and response to ART.

86 9th International Conference on HIV Treatment and Prevention Adherence 342 Salaam, Tanzania. collection duringimplementationofFamiliaSalamastudyinDares ability ofdatacollectedroutinely. We provideourexperienceindata tion intoroutinehealthcaresettings,itisimportanttoascertainreli- supervisors. ship isprovidedbythestudyteamincollaborationwithdistrict the month.Technical supportthroughtrainingandroutinementor- in physicalregisters,andsummaryreportsproducedattheendof ing datacollectedduringroutineservicedelivery. Dataarecollected tation studyisconductedinDaresSalaamfrom2012to2014,utiliz- Ester Mungure Theodora Mbunda Challenge Recommendations: nurses wereabletoutilizeCTC-2cardscomfortably. 2 cardsinlowerlevelMNCHfacilities,andaftertraining,cadre designed forhighercadrehealthworkers.OurstudyintroducedCTC- shows thatwhentrained,lowcadrestaffcantakeupM&Etasks we wereabletolinkabout65%ofpregnantwomen.Finally, ourstudy facilities. Onutilizingothervariablessuchasage,parity, residence, the facilitylevelhencedifficulttolinkpregnantwomenacross Identification numbersprovidedatregistrationwereuniqueonly linking pregnantwomenatdifferentstagesofPMTCTcascade. acceptable margininourstudysettings.Themajorchallengewas difference inreporteddatacomparedtophysicallyverifieddata,an collected dataarequestionable.However trainings, mentorshipandsupervisions.Dataaccuracyinroutinely changes inPMTCTguidelineandpolicies,leadingtoaneedformore lection toolsandreportingsystemhappenedsince2012asaresultof izing timeforpatientcare.Furthermore,3majorchangesindatacol data captureregisterscollectingsimilarinformationhencejeopard- Lessons Learned: Description: Introduction: 2 Lameck Machumi nant womenatdifferentlevels. use ofuniqueIDnumbersisrecommendedtoensurelinkagepreg- ables toavoidoverwhelmingpaperworksatMNCHclinics.Finally use ofstandardizedregistersincorporatingdifferentprogramvari- terms oftrainingsandmentorshiparewarranted.We recommend tion studyispossibleinTanzania, butonlyifadditionalsupportin 1 9th InternationalConference on Harvard SchoolofPublicHealth,Boston,MA,USA Dar esSalaam,Tanzania Management andDevelopmentforHealth, Familia Salama,a2-yearPMTCToptionA/Bimplemen- As HIV/AIDScareandtreatmentshifttowardsintegra- MNCH Settings:AnImplementation Documenting PMTCTServicesin 2 , IreneAnyango 1 In thisstudy, welearnedthepresenceofmultiple 1 , DavidSando , GuerinoChalamilla Using routinehealthfacilitydataforimplementa- 1 HIV 1 , MarySando , Lucy Magesa T reatment andPrevention Adherence 1 , wefoundalessthan10% 1 , AllyKaduma 1 (presenting) 1 , , - 343 had significantlymoredrinkingdays( condom use(CU),menreportingNCUduringlastreceptiveanalsex during mostrecentsexualencounters.Comparedtomenreporting drug use(47.1%)andalcoholconsumption(35.6%)wereprevalent anal sex,and38.3%reportedNCUduringlastinsertivesex.Illicit a third(37.2%)reportednocondomuse(NCU)duringlastreceptive drinks perdrinkingday, and5.7days“drunk”inthepastmonth.Over in theirlifetime.Menreported8.4drinkingdaysthepastmonth,2.9 marijuana (91.1%),methamphetamine(39.3%),andclubdrugs(38.8%) ty (55.8%)ofmenweregay-identified.Menreportedhavingused and riskysexualbehavior(i.e.condomuse)amongBMSM. ined theassociationbetweensubstanceuse(i.e.alcoholanddrugs) of sexmayincreasethelikelihoodsuchbehavior. Thisstudyexam- associated withHIVtransmission,andsubstanceuseinthecontext Black menwhohavesexwith(BMSM).Unprotectedis Background: 2 Vincent Allen Have SexwithMen Socioeconomic StatusBlackMenWho Sexual RiskBehavioramongLow among BMSMandcaninformHIVpreventionstrategies. of substanceuseinsexualbehaviorisimportantforaddressingHIV particular riskforsubstanceuseduringsex.Understandingtherole sex. Non-gay-identifiedBMSMandreceptivesexpartnersmaybeat among BMSMandhassignificantassociationswithunprotected Conclusions: hol duringlastreceptiveanalsex,(55.3%vs.44.7%) p Results: Methods: ( men, non-gayidentifiedmenhadagreaternumberofdrinkingdays analyses bysexualidentityrevealedthatcomparedtogay-identified and prevention. risk behaviors,substanceuse,andattitudestowardHIVtreatment likely toreportinconsistentCU(62.7%), 1 more daysusingmarijuanaoverthepastmonth( M <.05). Menusingdrugsduringlastinsertiveanalsexweremore Medicine, LosAngeles,CA,USA University ofCalifornia,LosAngeles,DepartmentFamily Los Angeles,CA,USA University ofCalifornia,LosAngeles,DepartmentPsychology, = 10vs. Men rangedinagefrom18to69( BMSM (N=224)participatedinastudyassessingsexual M These findingssuggestthatsubstanceuseisprevalent = 7.3,p<.05),andweremorelikelytoreportusingalco In theUnitedStates,HIVincidenceishighestamong 1 between SubstanceUseand Examining theRelationship (presenting) , RonaldBrooks M = 12.7vs. M ␹ 2 , StanleyJohnson 2 =5, p<.05.Exploratory = 33.5),andthemajori M M = 6.4, = 6.5vs. ␹ 2 = 6,p<.05. p <.01), and M 2 = 2.4, - - POSTER ABSTRACTS 87 Description of HIV-Disclosure and Does the Treatment Matter? 344 Relation with Medication Adherence 346 A Prospective Study of Post- Sexual Exposition Prophylaxis in Canada Aurelie Rotzinger1, Matthias Cavassini2, Matthias Reymermier3, Amico Sebastian1, Olivier Bugnon1, Réjean Thomas1, Nima Machouf1 (presenting), Sylvie Vézina1, Marie Paule Schneider1 (presenting) Michel Boissonnault1, Emmanuelle Huchet1, Stephane Lavoie1, Louise Charest1, Daniel Murphy1, Daouda Sissoko2, Benoit Trottier1 1 Community Pharmacy, Department of Ambulatory Care & Community Medicine, University of Lausanne, Lausanne, Vaud, 1 Clinique Médicale l’Actuel, Montréal, QC, Canada Switzerland 2 Université de Montréal, Montréal, QC, Canada 2 Infectious Disease Service, CHUV, University of Lausanne, Switzerland, Lausanne, Vaud, Switzerland Background: Post sexual-exposure prophylaxis (sPEP) poses sever- 3 Community Pharmacy, Department of Ambulatory Care & al challenges both to clinicians and patients. Better understanding of Community Medicine, University of Lausanne, Switzerland, sPEP use and outcomes is crucial in adapting our current practices. Lausanne, Vaud, Syria ABSTRACTS In the context of increasing use and scaling up of this strategy, and knowing that drug intolerance has often led to suboptimal adher- Background: Medication adherence is crucial in HIV. Among many ence, we aimed at assessing the impact of PEP regimen in enhanc- other factors, social support may influence adherence. This study ing adherence to prophylaxis. aims at investigating disclosure, social reaction to disclosure, participants’ living situation, and correlations with cART adherence. Methods: From 2000 to 2012, we prospectively enrolled patients consulting for sPEP in a single site cohort study. Our primary outcome Methods: Participants were eligible if they were older than 18, fluent was adherence to 28 days of treatment (PEP adherence). Factors in French, and had been taking part for more than 1 month in a phar- associated with adherence to PEP treatment were identified using macy medication enhancing program combining electronic meas- backward stepwise logistic regression analyses by SPSS 17.0. urement of medication adherence, and brief and repeated motiva-

POSTER tional interviews. The questionnaire administered in this cross-sec- Results: A total of 2,287 sPEP consultations were included. Patients tional design measured the number of people the HIV status has consulting for sPEP were mostly MSM (80%), with median age 32y, been disclosed to, social attitude after disclosure, and percentage of coming after receptive anal intercourse in 38% of cases. Source per- disclosed-to people living in the same home. Electronic medication son was known HIV+ in 26% of cases and risk of HIV acquisition was adherence data were retrieved over the last 3 months. moderate or high in 81% of cases. In 1,750 (78%) of our sPEP consultations treatment was advised, mainly by CBV/LPV (N = 243), Results: Over the 159 eligible patients, 31 refused the study and 25 TVD/LPV (N = 1,248) or TVD/RAL (N = 51). Global adherence to treat- postponed participation till the end of the recruitment period; 103 ment was 1206/1750 (69%, ITT). The adherence to treatment was participants took part in the study. They were in majority white (63% greater among patients receiving TVD with either LPV (73%) or RAL vs. 35%, p = 0.049), high educated (70% vs. 30%, p <0.001) and homo- (77%) regimen than among those receiving CBV.LPV (57%; p = 0.010). sexual (33% vs. 9%, p = 0.019) compared to those who refused. 38% of patients reported adverse events during PEP. The most com- Adherence was high (median 99%, IC 96,100), 12 (12%) participants mon adverse event reported was GI disturbance (32%) followed by did not disclose at all, 28 (28%) to 1 person and 63 (63%) to more than fatigue (11%). Only 6% of patients discontinued treatment premature- 1 person. Social reaction to disclosure was perceived as positive by ly. More patients in the CBV/LPV group had adverse events resulting 73 (75%) participants. Among the 62 (62%) participants living with in discontinuation of the prophylaxis (10%) compared to 5% for other people, 33 (52%) disclosed to all living-with people. No correla- TVD/LPV and 2% for TVD.RAL (p <0.001). We had 9 cases of serocon- tion with adherence was statistically significant. version among patients who had 1-month prophylaxis, but since there was a history of re exposition to HIV for 5/9 of them, we cannot Conclusion: HIV disclosure is a sensitive issue in Europe, possibly conclude to a PEP failure (seroconversion rate = 0.17%). explaining the high refusal rate and possible selection bias. Indeed, we cannot exclude that participants who did not disclose refused to DETERMINANTS OF A GOOD ADHERENCE TO PROPHYLAXIS ANALYZED BY participate. This question should therefore be further explored to MULTIPLE LOGISTICAL REGRESSION ANALYSES. determine its real impact on individual adherence and the way to approach it within a medication adherence clinic. Crude OR Adjusted OR Independent variables (95% CI) (95% CI) Regimen: CBV/LPV vs. TVD/LPV 0.46 (0.33 - 0.63) 0.48 (0.36 - 0.66) TVD/RAL vs. TVD/LPV 0.96 (0.48 - 1.95) 0.90 (0.45 - 1.82) Other vs. TVD/LPV 0.60 (0.42 - 0.86) 0.64 (0.45 - 0.90) Gender (Women vs. Men) 0.52 (0.36 - 0.76) 0.54 (0.37 - 0.78) Age 1.03 (1.01 - 1.04) 1.03 (1.01 - 1.04) Risk evaluation by physician: Moderate vs. Low 0.75 (0.36 - 1.53) -- High vs. Low 0.63 (0.31 - 1.24) -- Number of episode (first vs. subsequent) 1.22 (0.92 - 1.61) --

88 9th International Conference on HIV Treatment and Prevention Adherence adverse effectsrelatedtoARV. Interestingly, veryfewpatientsabandonedtheprophylaxisfor reduced adverseeventsconducivetodiscontinuationofPEP. to besuperiorCBV(withLPV)inincreasedadherencedrugsand sen forprophylaxisdoesmatter:TVD(witheitherLPVorRAL)proved high inourcohort.Thepresentstudysuggeststhattheregimencho- Conclusions: 9th InternationalConference on The adherenceratetoprophylaxisforsPEPwasquite HIV T reatment andPrevention Adherence 348 Caroline Watiri Methods: Background: 3 2 1 Michael Scanlon Children’s AdherencetoAntiretroviralTherapy Prospective CohortStudyMeasuring a Conclusions: time instudywerenotassociatedwithMEMSfailure. Results: resource-limited settingshasnotbeenadequatelyexplored. adherence toantiretroviraltherapy(ART)forpediatricpatientsin MEMS withliquidARTformulationsdeservefurtherinvestigation. among childreninthissetting.Potentialcomplicationsofusing MEMS notrecordingbottleopeningsordownloading). to investigatefactorsassociatedwithMEMSfailure(definedas Univariate analysesusingPearsonChi-squaretestswereperformed condition ofthebottle,andrecordedanyproblemswithMEMS. adherence data,discussedresultswithparticipants,evaluatedthe (EFV) pills.Atmonthlyvisits,studypersonneldownloadedMEMS bottles containingnevirapine(NVP)pills,NVPliquid,orefavirenz in alargeHIVcareprogramwesternKenya.MEMSwereusedon among HIV AARDEX, Inc.)bottlecapstoevaluateARTadherencefor6months dichotomized adherence( MEMS caps,usingwithNVPvs.EFVpills, associated withhavingaMEMSonNVPliquids(p=.07).Conditionof “very dirty.” Inunivariateanalyses,MEMSfailurewassignificantly majority ofMEMShad“usualwear”(85%)withonly2%classifiedas used atall(e.g.,medicationremovedfrombottleonce).The from bottlescontainingpillstoliquidsorviceversa,andnotbeing monly reportedproblemswithMEMSwere:leftathome,switched with NVPliquidbottles.MEMSfailureratewas2%.Additionalcom- was 96%.MostMEMSwereusedwithpillbottles,but41% evaluated at1,104studyencounters.MedianadherencebyMEMS percent werefemaleandmeanagewas8.2years.MEMSdata Eldoret, Kenya Academic ModelProvidingAccesstoHealthcare(AMPATH), Moi UniversitySchoolofMedicine,Eldoret,Kenya Indiana UniversitySchoolofMedicine,Indianapolis,IN,USA MEMS wereusedfor191HIV We usedMedicationEventMonitoringSystem(MEMS -infected childrenages0to14yearsreceivingcarewith Medication EventMonitoringin Feasibility andFailureRatesUsing The feasibilityofusingtechnologiesformonitoring MEMS isafeasibleARTadherence-monitoringtool 3 , RachelVreeman 1 (presenting) ³ 90% dosestakenor<90%)andlengthof , WinstoneNyandiko 1 -infected children.Fifty-five 2 , ® - , POSTER ABSTRACTS 89 Using Optimal Data Analysis to Testing a Mediated 349 Predict Antiretroviral Adherence 352 Biopsychosocial Model to Predict among HIV-Positive Youth Viral Load among People Living with HIV

Israel Gross1 (presenting), Sybil Hosek2, Maryse Richards3, Jennifer Attonito1 (presenting), Jessy Devieux1, Brenda Lerner1, Isabel Fernandez4, Heather Huszti5 Michelle Hospital1, Rhonda Rosenberg1

1 Loyola University Chicago/John H. Stroger Jr. Hospital of Cook 1 Florida International University, Miami, FL, USA County, Chicago, IL, USA 2 Stroger Hospital of Cook County, Chicago, IL, USA Background: Antiretroviral therapy (ART) adherence has been asso- 3 Loyola University Chicago, Chicago, IL, USA ciated with increased mortality, viral replication, and increased risk 4 Nova Southeastern University, Miami, FL, USA of transmitting the virus to others. Neurocognitive (NC) impairment, 5 Children’s Hospital of Orange County, Orange, CA, USA drug and alcohol abuse, and psychosocial factors such as stress all have been associated with poorer adherence, while social support

ABSTRACTS Background: Adherence to antiretroviral therapy is crucial for has predicted better adherence rates. This study sought to explore thwarting disease progression and reducing secondary transmis- the effects of several psychosocial variables upon viral load (VL) via sion, yet HIV+ youth struggle with adherence. The highest rates of mediation of ART adherence among HIV-seropositive adults. new HIV infections occur in young African American men (YAAM), thus understanding reasons for non-adherence in this group is criti- Methods: Multivariate linear regression was used to test a mediated cal. Reasons for non-adherence can be complex and multifactorial, biopsychosocial model predicting VL among a cross-sectional sam- and innovative methods of exploration are needed for advancing pre- ple of 246 HIV-positive adults who were on ART. Exogenous variables vention and treatment efforts. were tangible social support, barriers to ART adherence, and stress. Moderators were alcohol use, marijuana use and NC impairment. Methods: A sample of 387 HIV+ YAAM who reported currently taking Associations were analyzed using structural equation modeling HIV medications were selected from a cross-sectional assessment of (SEM) with Mplus. Full mediation through ART adherence was tested POSTER 2,226 HIV+ youth from sites within the Adolescent Trials Network for using Baron and Kenny’s 4-step approach. HIV/AIDS Interventions (ATN) from 2009-2012 (12-24 years-old, Median = 22.00, SD 2.06). Participants completed self-reported adher- Results: A small positive association between marijuana use and ence, demographic, health, and psychosocial measures. Seventy-two adherence approached significance (␤ = 0.15, p = .057) however, theoretically relevant predictors of adherence underwent Optimal “barriers to medication adherence” was the only independent vari- Data Analysis (ODA) to construct a classification tree which hierar- able significantly associated with both ART adherence and VL in chically maximizes the classification accuracy of 100% adherence. SEM, so was tested in a mediated relationship. Each additional barrier predicted a 10% decrease in adherence rates and a 0.42-unit Results: Sixty-two percent reported 100% adherence (no missing increase in log10 VL. No other factors were significantly associated doses) over the past 7 days. Frequency of cannabis use was the with either VL or adherence and no interaction effects between strongest predictor of adherence, yielding moderate effect strength exogenous variables and moderators were identified. The SEM sensitivity, ESS = 27.1, p <0.00. Among participants with infrequent explained 31% of variance in ART adherence and 18% of variance in cannabis use, 72% demonstrated full adherence, while only 45% of log10 VL. The relationship between barriers to adherence and VL participants who used cannabis (monthly or more) demonstrated full was found to be partially mediated by ART adherence. adherence. Classification tree analysis (CTA) correctly classified 82.35% of those who were adherent and 64.85% of those who were Conclusions: These findings provide modest support for the biopsy- non-adherent. The final CTA adherence model was strong (ESS = chosocial model in predicting virologic response to ART. Outcomes 49.12) identifying 4 pathways towards adherence and 5 pathways may provide a more complete picture of the complex factors affect- toward non-adherence. Participants most likely to be adherent were ing the health of HIV-positive adults. The effects of marijuana on ART those less likely to have substance abuse issues and reported low adherence require further exploration. levels of psychiatric distress (92.59% were adherent).

Conclusions: This research demonstrates the impact of substance use and mental health on adherence among YAAM. Moreover, this analysis identifies complex and multiple profiles of adherence among HIV+ YAAM and suggests that targeted interventions may be most prudent.

90 9th International Conference on HIV Treatment and Prevention Adherence 353 PrEP inclinicalpractice. implications forfutureresearchandtheimplementationofTVD analysis. and providerdemographics,laboratorydata)wasavailablefor de-identified patientdata(druginformation,medicalclaims, TVD forPrEPbetweenJanuary2011andSeptember2013.Detailed electronic datafrom~55%ofallUSretailpharmaciesthatdispensed tions byage,gender there aresignificantregionaldifferencesinTVDforPrEPprescrip- Conclusions: treated byanon-IDphysician(95%CI3.3-4.2). than 25yearsold(95%CI1.3-1.7)and3.7timesmorelikelytobe ly tobefemale(95%CI1.7-2.0);1.5timesmorelikelyyounger patients, individualsreceivingTVDforPrEPwere1.8timesmorelike- and PhysicianAssistants9%respectively. ComparedtoHIV-positive Internal Medicine16%;InfectiousDiseases12%;NursePractitioners with thehighestinitiationofPrEPprescriptions:FamilyPractice17%; 38.5 +12.2years,and11.8%ofindividuals<25yearsold.Specialties users onMedicaid(15.9%).Overall,themeanageofPrEPwas sons prescribedPrEP(16.2%);andthehighestproportionof PrEP users(34.7%).TheMidwesthadthesmallestproportionofper- largest proportionofindividualsprescribedPrEP(32.9%)andfemale of newHIV PrEP. Women accountedfor47.5%ofPrEPuserscomparedto32.04% Results: Methods: HIV infectionsin2011. istics ofTVDforPrEPuserscomparedtoCDCdemographicsnew HIV-1 inadultsathighrisk.We describethedemographiccharacter- exposure prophylaxis(PrEP)toreducetheriskofsexuallyacquired Background: 1 Staci Bush Robertino Mera1 9th InternationalConference on Gilead Sciences,FosterCity, CA,USA A totalof2,491uniqueindividualswereprescribedTVDfor A previouslydescribedalgorithmwasappliedtonational 1 -positives. AmongUSCensusregions,theSouthhad (presenting) Initial analysisofnationalpharmacydatasuggeststhat In July2012theFDAapprovedTruvada (TVD)forpre- Utilization ofTruvada forPrEP Regional Differencesinthe , andprescriberspecialty , KeithRawlings HIV T reatment andPrevention Adherence 1 , LeslieNg . Thesefindingshave 1 , 354 Methods: Background: 4 3 2 1 Tracy Glass Daniel Nyogea Tanzania: TheKIULARCOCohort Conclusions: (aOR: 1.32,95%CI:1.00-1.76). HIV statuswastheonlysignificantpredictorofworseningadherence longer timeonART(aOR:0.87,95%CI:0.78-0.98).Nondisclosureof aOR: 1.18,95%CI:0.87-1.62).Riskofnon-adherencedecreasedwith (aOR): 1.32,95%confidenceinterval(CI):1.00-1.75;last4months, (compared tomiddle4months:firstmonths,adjustedoddsratio ence initiallydecreasedandthenincreasedoverthestudyperiod visits-pairs with45.5%ofchangesbeingfortheworse.Non-adher- reason (46.1%).Changesinadherencewerereportedon16.4%of occurred on12.0%ofvisitswithdrugstockoutcitedastheprimary ting (48.0%)andrunningoutofmedication(19.4%).Drugswitching weeks, respectively and 15.6%ofindividualsmisseddosesinthepast3days1-4 60.7% wereonnevirapine-basedregimen.Onatleast1visit,20.0% HIV status.MediantimeonARTwas2.4years(IQR:1.2-3.5)and female, medianagewas39(IQR:32-47),and57.9%disclosedtheir tionnaires (median=5,interquartilerange(IQR):3-6),67.6%were Results: adherence questionnairedissipatedovertime. changed overtime;initialimprovementslikelyduetointroducingan ence toART, suchasdisclosure,inotherRLS.Adherencepatterns the KilomberoandUlangaAntiretroviralCohort(KIULARCO)study. (RLS). We conductedaprospectiveadherenceassessmentwithin ence toantiretroviraltherapy(ART)inresource-limitedsettings Lars Henning adherence. explore factorsassociatedwithnon-adherenceandchangesin doses. Repeatedmeasureslogisticregressionmodelswereusedto were included.Non-adherencewasdefinedasanyreportedmissed Chronic DiseasesClinicIfakarabetweenJune2011andMay2012 University HospitalZurich,Switzerland Basel, Basel-Stadt,Switzerland Swiss Tropical andPublicHealthInstitute, Ifakara HealthInstitute,DaresSalaam,Tanzania Ifakara HealthInstitute,Ifakara,Tanzania We included1,822patientscompletingatotalof8,379ques- All patientsinKIULARCOreceivingARTandattendingthe 3 4 (presenting) Limited informationisknownaboutpatternsofadher- We foundsimilarlevelsofandfactorsaffectingadher- , FabianFranzeck Patients inSouthernRural Adherence PatternsinHIV-Positive 1 , EvelineGeubbels . Mainreasonsfornon-adherencewereforget 3 , EmiliLetang 2 , SallyMtenga 3 , MarcelTanner 1 , IngridFelger 3 , 3 , - POSTER ABSTRACTS 91 Feasibility of an HIV Self-Test Hospitalized Out of Care Persons 355 Voucher Program to Raise 356 with HIV Infection Face Numerous Community-Level Serostatus Awareness, Barriers to Retention in Care Los Angeles Thomas Giordano1 (presenting), Christine Hartman2, Jessica Davila1, Jeffrey Cully1, Melinda Stanley1, Michael Kallen3, Robert Marlin1, Sean Young1, Emily Huang2 (presenting), K Rivet Amico4 Claire Bristow1, Greg Wilson3, Jeffrey Rodriguez4, Jose Ortiz2, Rhea Mathew2, Jeffrey Klausner1 1 Baylor College of Medicine, Houston, TX, USA 2 DeBakey VA Medical Center, Houston, TX, USA 1 David Geffen School of Medicine at University of California, 3 Northwestern University, Chicago, IL, USA Los Angeles, Los Angeles, CA, USA 4 University of Connecticut, Brighton, MI, USA 2 University of California, Los Angeles, CA, USA 3 Reach Los Angeles, Los Angeles, CA, USA Background: Finding persons with HIV who are out of care to re- 4 Los Angeles Gay & Lesbian Center, Los Angeles, CA, USA ABSTRACTS engage them is challenging. Hospitalizations represent such an opportunity. We are conducting a randomized study of an interven- Introduction: Up to half of all new HIV cases in Los Angeles may be tion to improve retention in care, enrolling at a public hospital. The caused by the 20-30% of men who have sex with men (MSM) with degree of barriers to retention in care for such a population is not unrecognized HIV infection. MSM are at high risk for being sero- well characterized. unaware and might benefit from increased access to novel testing methods, such as the recently FDA-approved OraQuick® In-Home Methods: We enrolled persons out of care or newly diagnosed with HIV Test. HIV infection in the trial and assessed participants’ self-reported barriers to outpatient HIV care. We assessed 7 barriers, based on prior Description: From July-November 2013, we examined the feasibility literature: a new diagnosis of HIV infection, no health insurance, at- of implementing a voucher program for free OraQuick® tests target- risk alcohol use (AUDIT-C score >4 for men and >3 for women); illicit

POSTER ing high-risk MSM in Los Angeles. We determined feasibility based substance use (any use in last 3 months, excluding marijuana use), on ability to: (1) establish a voucher redemption and third-party pay- depression (PHQ-8 score >10, which is high risk for moderate depres- ment system, (2) use community-based organizations (CBOs) to dis- sion), homelessness (living on street or in shelter, half-way house, or seminate vouchers, and (3) collect user demographics and test drug rehabilitation facility), and at least 1 unmet need out of 19 serv- results with an anonymous telephone survey. We defined high-risk ices, including medical and dental care, mental health and substance MSM as those with > 1 partner, untested for HIV, and with inconsis- use care, social and case management services, legal and immigra- tent condom use in the past 6 months. We calculated descriptive sta- tion services, nutrition services, and general health as well as HIV- tistics using Microsoft Excel® and STATA® 13. specific information, in the last 6 months. We calculated the percent with each barrier, as well as the proportion with any barrier. Lessons Learned: We partnered with Walgreens® to create a voucher and third-party reimbursement system for free OraQuick® tests. Of Results: We enrolled 460 participants between 2010 and 2013, and 641 vouchers supplied to CBOs and other distributors, 274 (42.7%) 456 completed the baseline survey: mean age was 42.1 years; 67% went to clients; 53 (19.3%) were redeemed. Fifty (94.3%) of 53 vouch- were black, 19% Hispanic, 14% white; 73% male, 26% female, 1% er-redeemers were surveyed: 39 (78%) reported being likely to repeat transgender; 80% unemployed. Barriers were frequent: 5% were voucher use, 44 (88%) reported reviewing pre-test information, 37 homeless, 11% were newly diagnosed, 27% reported recent illicit (74%) the post-test information, and 12 (24%) were high-risk MSM. drug use excluding marijuana (83% of which was cocaine); 32% Three (6%) of 50 respondents reported newly testing HIV-positive of reported at risk alcohol use; 46% were high risk for moderate depres- whom all (100%) reported seeking medical care. Two withheld their sion, 71% had no private or public health insurance; and 83% report- results, of whom, 1 sought medical care. ed at least 1 unmet need (median [IQR] number of unmet needs = 3 [1, 6]). Overall, 95% of participants had at least 1 barrier to retention in Recommendations: Developing a voucher system to facilitate HIV care (median number of barriers [IQR] = 2 [1, 3]), including the follow- self-testing with linkage to care was feasible. Our findings suggest ing distribution of participants with 0 to 7 barriers, respectively: 5%, the voucher program was associated with increasing the identifica- 21%, 31%, 26%, 13%, 4%, <1%, and 0%. tion of new cases of HIV infection with high rates of linkage to care. Expanded research and evaluation of voucher programs for HIV self- Conclusions: A patient population hospitalized with HIV infection in test kits among high-risk groups is warranted. our public hospital has multiple significant barriers to retention in care. Coordinated inpatient and outpatient social services and case management services that address barriers are likely critical. Other interventions should be developed for and deployed in similar settings.

92 9th International Conference on HIV Treatment and Prevention Adherence Conclusions: Results: Methods: tered HIV research tocomparedifferentmeasuresofadherenceamongregis- adherent behaviorrequiresadditionalmeasures.We undertookthis count andviralloadarecollectedregularly itoring oftheirhealthstatus.ClinicalmeasuressuchasCD4cells HIV-positive patientsshouldvisittheclinicsevery6monthsformon- sufficiently studied.AccordingtoRussianfederalregulations,all available inRussia,measuresofadherencetoHAARThavenotbeen Background: 4 3 2 Andrey Ustinov Alena Suvorova 1 adherence mightimprovetheagreementbetweenmeasures. adherence inself-reports.Possibly, consideringmorethan2levelsof priate consistencyandthiscanbeexplainedbyanoverestimationof fair agreement(kappa=0.28,p<0.001). between doctors’estimateandclinicalmeasures,whereweobtain measures) showedonlyagreementbychance(kappa<0.12),except sources (self-reports,pharmacyrefills,doctorsestimate,clinical adherence (kappa=0.71,p<0.001).Measuresobtainedfromdifferent p <0.001),substantialagreementbetween30-dayandvisualscale <0.001), between6-monthandvisualscaleadherence(kappa=0.52, agreement between30-dayand6-monthadherence(kappa=0.55,p reported adherenceingeneralagreedwitheachother:moderate age, maritalstatus,andincome.Thedifferentmeasuresofself- tistically significantdifferencesinallmeasuresofadherencebysex, of malewithameanageis35years(SD=7.5).We observednosta- We usedCohen’s Kappatocomparethescales.Thesamplewas57% reported adherenceasthecutoffbetweenlowandhighadherence. the clinicalmeasures. mate inmedicalcharts,measurementbasedonpharmacyrefills,and log scale.We use3additionalmeasuresofadherence:doctors’esti- and self-reported6-month30-dayadherenceusingavisualana- answered questionsabouttheirsocio-demographiccharacteristic St. Petersburg.Duringface-to-faceinterviewsparticipants 9th InternationalConference on Russian Federation St. PetersburgCityAIDSCenter, St.Petersburg, St. Petersburg,RussianFederation Bekhterev ResearchPsychoneurologicalInstitute, Yale University, NewHaven,CT, USA St. Petersburg,RussianFederation Automation oftheRussianAcademySciences), SPIIRAS (St.PetersburgInstituteforInformaticsand 357 Adherence wasconvertedtobinaryscaleusing95%self- -positive peopleinSt.Petersburg. We surveyed210HIV-positive patientsreceivingHAARTin Although highlyactiveantiretroviraltherapy(HAART)is Different measuresofadherencedonotshowappro 3 1 , AndreyBelyakov (presenting) HAART intheRussianContext Measuring Adherenceto , OlgaLevina HIV 4 T reatment andPrevention Adherence 2 , buttheunderstandingof , RobertHeimer 2 , - 360 Olga Levina Background: 4 3 2 1 Andrey Ustinov Preliminary Results among St.PetersburgHIV-Positive Patients: Conclusions: Results: high adherence. cators ofARTadherence,whichwasthencharacterizedaslowor viral loaddatafrommedicalchartsduringthelast6monthsasindi dictors ofinterest.We additionallyobtainedCD4cellcountandHIV 90-R) todeterminethelevelofsubstanceuseandmentalillnesspre- Severity Index(ASI)andtheSymptomCheckList90Revised(SCL- taking ARTnotlessthan6months.InterviewincludedtheAddiction Methods: illness, thatarecommonamongHIV that arelikelytoimpactARTadherence--substanceuseandmental studied inRussia.Forthisanalysiswefocusedonspecificfactors tors predictingARTadherencesuccesshavenotbeensufficiently (ART) hastransformedHIVtoatreatablecondition,butspecificfac cases thananyoftheotherregionsinRussia.Antiretroviraltherapy public healthconcern.ThecityofSt.PetersburgnowhasmoreHIV Anxiety Compulsive Symptoms,InterpersonalSensitivity, Depression, and SCL-90-Rsymptomconstructs:Somatization,Obsessive- includes ASI:Alcoholuse,DrugLegalstatus,Psychiatric However, thebestsetofpredictors(stepwiseregression,AIC=54) sample (n=55)bygender, alsonosignificantdifferenceswerefound. significant differenceswerefoundonlybygender. Afteradjustingthe (51.5%) withlowand31(48.5%)highadherence.Statistically ment atasinglesite. intake andintegratepsychiatric,substanceabuse,HIVtreat include screeningforsubstanceuseandmentalillnessatHIVclinic level ofARTadherencecanbeimprovedthroughprogramsthat predictors ofinterest.Furtherstudymayrevealthatachievinghigh results suggestedconnectionsbetweenthelevelofadherenceand AIDS Center, St.Petersburg,RussianFederation Yale University, NewHaven,CT, USA St. Petersburg,RussianFederation Institute forInformaticsandAutomationofRAS, St. Petersburg,RussianFederation Bekhterev ResearchPsychoneurologicalInstitute, , Phobic-Anxiety A vailable dataon64patientsallowedcharacterizationof33 We conducted80face-to-faceinterviewswithpatients 3 , AndreyBelyakov The currentRussianHIVepidemicrepresentsaserious Although theresultsarelimitedbysamplesize, as PredictorsofARTAdherence Substance UseandMentalIllness 1 (presenting) , ParanoidIdeation,Psychoticism. , AlenaSuvorova 4 -infected patientsinRussia. 2 , RobertHeimer 3 , - - - POSTER ABSTRACTS 93 Novel Techniques for Investigating The Science of Being a Study 361 Adherence Behavior Using Data 364 Participant: FEM-PrEP from Electronic Drug Monitor Participants’ Explanations for Over- Reporting Study Product Adherence and the Christopher Gill1, Mary Bachman DeSilva2 (presenting), Lora Sabin1 Whereabouts of the Unused Study Product 1 Boston University School of Public Health, Boston, MA, USA 2 Center for Global Health and Development, Boston University, Amy Corneli1 (presenting), Kevin McKenna1, Kawango Agot2, Boston, MA, USA Khatija Ahmed3, Brian Perry1, Fulufhelo Malamatsho3, Jacob Odhiambo2, Joseph Skhosana3, Lut Van Damme4 Background: Electronic drug monitors (EDM) have proven highly useful in the study of adherence. Typically, EDM data are expressed as 1 FHI 360, Durham, NC, USA the percentage of doses taken or doses taken on time, or 2 Impact Research and Development Organization, dichotomized above or below 95% adherence. Here, we describe a Kisumu, Kisumu, Kenya

ABSTRACTS series of investigations aimed at making fuller use of the full EDM data 3 Setshaba Research Centre, Pretoria, Pretoria, South Africa set from the ‘Adherence to Life’ (AFL) study conducted in Dali, China. 4 The Bill & Melinda Gates Foundation, Seattle, WA, USA

Methods: We created 2 methods for displaying individual level EDM Background: FEM-PrEP could not determine the effectiveness of data over time. Method 1 (Unimodal) express calendar date on the Y- emtricitabine/tenofovir disoproxil fumarate (FTC/TDF) for HIV preven- axis; the X-axis displays when the EDM bottle was opened on an tion among women in sub-Saharan Africa because of low adherence absolute scale, such that an event occurring on time falls at the ori- to the study product. Self-reported adherence was high and pill- gin, with greater deviations (early or late) as one moves right from the count data suggested good adherence. Yet, analyses of drug con- origin. Method 2 (Bimodal) reverses the X and Y-axes with calendar centrations revealed evidence of recent pill use in only a minority of date on the X-axis, and the Y-axis expressing openings relative to the participants. We conducted a follow-up study in 2 FEM-PrEP sites to scheduled dose time, and early/late openings deviating plus or minus understand factors that influenced over-reporting of adherence and POSTER from the origin. The resulting bimodal scatter plots were categorized to learn the whereabouts of unused study pills. into those with high (nearly all events occurring within 1 hour of dose time), moderate (occasional deviations from dose window) or poor Methods: We conducted qualitative, semi-structured interviews (SSIs) adherence (frequent events out of dose window). with 88 participants assigned FTC/TDF and quantitative audio computer-assisted self-interviews (ACASI) with 224 participants assigned Results: 68 AFL subjects contributed data. From the bimodal scatter FTC/TDF or placebo. We used thematic analysis and descriptive statis- plots, 29/68 (43.1%) were highly adherent, 18/68 (26.5%) were moder- tics to analyze the qualitative and quantitative data, respectively. ately adherent, and 20/68 (29.4%) were moderately adherent. Within the moderately/poorly adherent subjects, 5/38 (13.2%) were early Results: Among ACASI participants, 31% (n = 70) said they over- mistimers; 9/38 (23.7%) were symmetrical mistimers; and 24/38 reported adherence during the trial (69%, n = 154, said they did not). (63.2%) were late mistimers. From unimodal plots, mean slopes are Of the 70 participants, 69% (n = 48) believed reporting non-adherence calculated per subject and reported in the poster. Slopes ranged would terminate them from the trial, 63% (n = 44) did not want to dis- from nearly vertical (~90 Degrees, highly adherent) to horizontal (~45 appoint staff, 61% (n = 43) thought misreporting was easier than degrees, extremely non-adherent). reporting actual adherence, and 56% (n = 39) thought staff would scold them. Frequently-mentioned whereabouts of the unused pills Conclusions: Late mistiming was the dominant pattern of non-adher- were: returned to clinic (83%, n = 185) or discarded (35%, n = 78). Few ence. Further analyses will explore the relationship between these participants reported stockpiling pills (7%, n = 15) or giving them to metrics and biological measures of HIV disease control (VL, CD4). someone with HIV (2%, n = 5). Site differences exist. Many SSI participants said other participants counted and removed pills from their bottles to appear adherent.

Conclusions: Despite repeated messages that non-adherence would not upset staff, participants had difficulty reporting that they weren’t following study procedures, primarily because of perceived negative consequences. Uneasiness continued after the trial, as many participants in this study said they always reported adherence accurately during the trial. Efforts to improve self-reported measures should identify alternative methods for creating supportive environments that allow participants to feel comfortable reporting actual adherence.

94 9th International Conference on HIV Treatment and Prevention Adherence 365 ed individuals. in theworld,withBAAmenrepresentingmajorityofincarcerat- other races.IncarcerationratesintheUnitedStatesarehighest rates andworseHIV-related healthoutcomes,comparedtoMSMof Black orAfrican-American(BAA)MSMsufferfromhigherinfection who havesexwithmen(MSM),intheUnitedStates.Inaddition care. sion modelswereconstructedtodefinethecorrelatesofretentionin provider atbaselinepriortoincarceration,comparedwithothermen. MSM weresignificantlylesslikelytohaveaccessanHIVhealth compared toothermalereleasees.Additionally, theyoungBAA bility toberetainedinHIVcarethefirst3monthspost-jailrelease, found thatyoungHIV+BAAMSM(<30yearsold)hadalowerproba- Jail: ResultsfromaMulti-SiteStudy with MenbeforeandafterReleasefrom Conclusions: Results: in JailSettingInitiative Methods: Background: 1 Panagiotis Vagenas ventions forformerlyincarceratedyoungHIV+BAAMSM. Angela DiPaola 1270), HRSA’s prospective cohortstudyofHIV 9th InternationalConference on Yale University, NewHaven,CT, USA Among the22%ofBAAmenwhoself-identifiedasMSM,we This studyanalyzeddatafromthelargestmultisite There isanurgentneedforspecificallytargetedinter The HIV/AIDSepidemicisconcentratedamongmen Enhancing LinkagestoHIVPrimaryCareandServices Infected BlackMenWhoHaveSex Retention inCareamongHIV- 1 , FrederickAltice 1 (presenting) (EnhanceLink). Multivariatelogisticregres- -infected releasedjaildetainees(N= HIV , AlexeiZelenev 1 , SandraSpringer T reatment andPrevention Adherence 1 , 1 - Therapy at6EthiopianHIVClinics,2012-2013 tus topeopleintheirliveswhocanoffersupport. some individualsmayneedadditionalhelptodisclosetheirHIVsta and reporteddisclosureexperienceswerelargelypositive.However, sure to (males: 58%,females:66%,p=0.030).Inmultivariablemodels,disclo- 95%, females:89%,p=0.022),followedbyotherfamilymembers,63% disclosed sharedtheirstatuswithmainpartner, 92%(males: men: median1;women:2).Themajorityofparticipantswho tus confidedinamedianof2persons(interquartilerange[IQR]1-4; Denis Nash Background: 5 4 3 2 Tsigereda Gadisa Conclusions: and partnerviolence(61%). main concernsincludedunauthorizeddisclosure(72%),gossip(67%), HIV status.Amongthe136patientswhodidnotdisclosetoanyone, concerned thatsomeonetheyhadnotdisclosedtomightlearntheir treatment. Despitethesepositiveexperiences,52%ofpatientswere response: 84%werecomfortedand85%encouragedtoget 3.41). Upondisclosure,mostparticipantsexperiencedasupportive and havingeverhadchildren(aORref=no=2.22,95%CI:1.45- ated withnotlivingalone(aORref=live=5.40,95%CI:3.34-8.73) who discloseddidsowith professional priortoARTinitiation.Fifty-eightpercentofparticipants 0.006) disclosedtheirHIVstatustosomeoneotherthanahealthcare Results: Methods: used toexaminecorrelatesofHIVdisclosure. tionnaire. Chi-squaredtestsandadjustedoddsratios(aOR)were ics inOromia,Ethiopia,wereinterviewedusingastructuredques- 1 Maria Lahuerta diagnosis ofHIV-positive familymembersandsexpartners. HIV statustofacilitatetreatmentadherence,socialsupport,and Sarah GorrellKulkarni [CI]:1.54-2.84), whereasdisclosureto other HIV+person(aORref=0=2.09,95%confidenceinterval Oromia StateRegionalBureauofHealth,AddisAbaba,Ethiopia HIV CenterforClinicalandBehavioralStudies,NewYork, NY, USA New York, NY, USA Mailman SchoolofPublicHealth,ColumbiaUniversity, CUNY SchoolofPublicHealth,NewYork, NY, USA New York, NY, USA ICAP-Columbia University, MailmanSchoolofPublicHealth, 367 ³ Overall, 86%ofrespondents(males:82%,females:88%,p= 2 personsvs. Adults initiatingART(N=945)in2012and2013at6HIVclin 2 , ZenebeMelaku Ethiopia’s nationalguidelinesencouragedisclosureof HIV statusdisclosurepriortoARTinitiationwashigh 1 , MulunehYigzaw 1 , Olga Tymejczyk Persons InitiatingAntiretroviral HIV DisclosureHistoryamong 2 , SusieHoffman ² 1 wasassociatedwithknowingatleast ³ 2 persons.Thosewhodisclosedtheirsta- 1 1 , ShaloDaba 2 (presenting) ³ 1 personvs.nonewasassoci- 3 , RobertRemien 5 , BatyaElul , 4 , 1 , - - POSTER ABSTRACTS 95 Factors Contributing to Missed ART Towards the Elimination of HIV 368 Doses among HIV-Infected Patients 376 Mother-to-Child Transmission: The with Suboptimal Adherence Participating in Experience of a Buenos Aires Public Hospital an Adherence-Focused Care Management Fernando Murano1 (presenting) Randomized Controlled Intervention Trial Integrated into Routine Clinical Care 1 Hospital de San Fernando, Buenos Aires, Argentina Background: To analyze the clinical and epidemiology issues of preg- Rob Fredericksen1 (presenting), Tyler Brown1, Anna Church1, nant HIV-positive women (HPW) and the rate of mother to child Anna Harrington1, Paul Crane1, Paul Ciechanowski1, Kari Nasby2, transmission (MTCT). Jennifer Magnini2, Heidi Crane1 Methods: Prospective cohort of HPW from 1997-2010 was analyzed. 1 University of Washington, Seattle, WA, USA It compares Group 1 (G1): 1997-2003 with 63 Newborns (NB) and 2 Harborview Medical Center, Seattle, WA, USA ABSTRACTS Group 2 (G2): 2004-2010 (102 NB). The NB HIV diagnosis was performed by viral culture and p24 until 1999, and after that DNA PCR Background: We identified key factors contributing to suboptimal and ELISA at 18 months. medication adherence based on documented discussions between HIV-infected patients and their case managers in clinical care. Results: 165 children were born. HPW median age was 26 (15-42) G1: 24.5, G2: 28 (P = 0.001).HPW drug users (DU) 42% G1, 43% G2 (P = 0.3). Methods: Patients in routine care with suboptimal medication adher- Time of diagnosis: G1: Prior to pregnancy (PP) 33/62 (53.2 %), before ence were enrolled in a randomized controlled intervention trial inte- delivery (BD) 6 (9.7 %), G2: PP 74/103 (71.8 %), BD: 7 (7 %). Use of PI grated into care using a stepped-care approach consisting of up to 6 was 8.1 % for G 1 and 57.3 % for G2. (P = 0.001). NNRTI 35,5% in G1 sessions, depending on adherence level, at 6- to 8-week intervals. and 26,2% in G2 (P 0,001). Median CD4 was 350 in G1 and 366 in G2 (P Clinic case managers served the care manager role to demonstrate = 0.66). VL was <1000 in 57.1 % in G1 and 76.1 % in G2 (P = 0.11). No

POSTER feasibility in care by using regular clinic personnel. Structured ses- VL results BD: 55.5 % in G1 and 31.4% in G2 (P = 0.002) ARVs during sions targeted adherence barriers using a problem-solving approach. labor: 88.7% of G1 and 91.3% in G2 (P = 0.99). ARV in NB was 95.2% in Case managers documented “problems addressed” and “solutions G1 and 99% in G2 (P = 0.045). Cesarean section (CS) was 35.5% G1 to discussed” concerning missed doses. We used an open-coding sys- 67% G2 (P = 0.0001). Median weight of NB was 3250gr G1 and 3000gr tem to create categories of reasons for suboptimal adherence. G2 (P = 0.002). Infection was confirmed in 9/165 NB (5.45%) 5/63 (7.94 %) for G1 and 4/102 (3.92 %) for G2. Only 2/9 in G2 had CV results BD Results: Case managers documented adherence discussions for 153 and none in G1. sessions for 60 unique patients. In one-third of sessions, patients denied having current adherence issues. Half (50%) of patients dis- Conclusions: Reduction of MTCT in recent years was related to early cussed more than 1 problem affecting their adherence. Most com- diagnosis, increased use of HAART and cesarean section. VL results mon were depression in 21% of sessions, fatigue/sleep problems BD increased in G2. It is a priority to improve access and retention in (12%), substance use (11%), forgetting (11%), separation from med- care of mothers and children to meet the challenge of PAHO/WHO 2015. ications (8%), disruption in basic needs such as food, shelter, money, or transport (8%), being distracted/too busy (7%), irregular work schedule (5%), and issues with ingestion, such as poor appetite or difficulty swallowing (5%). Of these, interrelated barriers to adherence were common, such as irregular work schedule followed by sleep disruption and/or separation from medication.

Conclusions: Suboptimal adherence often derived from a multitude of contextual factors and their relationships were often complex and cyclical: i.e., financial problems leading to social isolation, followed by depression and substance abuse. It may be necessary to address multiple barriers to adherence simultaneously to have a substantial impact on complex adherence behavior. For patients facing complex, interrelated psychosocial and logistical problems, case managers are an integrated part of routine clinical care and can facilitate identifying and addressing adherence challenges.

96 9th International Conference on HIV Treatment and Prevention Adherence 378 in NewYork City. eration offutureusePrEPinmenwhohavesexwith(MSM) online surveytoassesscorrelatesofawareness,uptake,andconsid- Sarit Golub New York City Men UsingMobileDatingApplicationsin among High-RiskMenWhoHaveSexwith ing thelargegapbetweenpotentialandactualuse. efforts forPrEPneedtofocusonincreasingawarenessandaddress- High-risk MSMweremorelikelytoconsiderPrEPuse.Dissemination ple. InterestinPrEPwashigh,butratesofactualusewereverylow. Conclusions: Results: Methods: Background: 3 2 1 Viraj Patel race, HIVtestinghistory 1.55, 95%1.15-2.08),butnotassociatedwithage,income,education, past 6months(AOR1.48,CI1.13-1.96)andnotusingacondom sis, considerationofPrEPwasassociatedwithhaving>4partnersin would recommendPrEPtoafriendorpartner. Inmultivariateanaly- brief PrEPinformation,66.6%wouldconsidertakingand79% Overall, 42.7%wereawareofPrEPbut4.2%hadused had anHIVtestinthepast6months,76%werenotarelationship. 6 months.Over32%didnotusecondomsatlastintercourse,71.6% Hispanic, and21%Black.Themenhadamedianof5partnersinpast lege degreeorhigher, 44.4%madeover$50,000/year;and31%were future use.W PrEP awareness,currentorprioruse,andconsiderationof graphics, HIVstatus,condomuse,lasttest,numberofpartners, months andlivinginNewYork City. Participantsself-reporteddemo- mobile datingapplicationstomenhavinghadanalsexinpast12 analysis tothosehavingaHIVnegativeorunknownstatus(n=1,238). lates ofconsiderationPrEPuse(primaryoutcome)andlimitedthe 9th InternationalConference on Hunter College,CUNY, NewYork, NY, USA Harlem UnitedCommunityAIDSCenter, NewYork, NY, USA Bronx, NewYork, USA Montefiore MedicalCenter;AlbertEinsteinCollegeofMedicine, Participants wereonaverage32yearsold,59%hadacol- Cross-sectional surveyadministeredon2MSM-specific 1 3 (presenting) e usedmultivariatelogisticregressiontoidentifycorre To informPrEPdisseminationefforts,weconductedan A majorityofMSMwereunawarePrEPinthissam- of OralPre-ExposureProphylaxis High AcceptabilitybutLowUptake , orpriorknowledgeofPrEP , SherryEstabrook HIV T reatment andPrevention Adherence 2 , KimberleighSmith . . Having 2 , - 384 1 Background: 3 2 Modibo Keita Infection Goal Mali in2012:HopefortheZeroNewHIV Growing AreaoftheOfficeduNigerin Conclusions: Results: Methods: Ilo Dicko stakeholders. tion programsforthistargetgroupinareaissupportedbyallthe new infectionhopeisattainableifthestrengtheningofHIVpreven- ers intheOfficeduNigerricegrowingareaMali2012.Thezero Siaka Coulibaly method wasusedtoestimatetheoverallincidenceofHIVinfection. identify HIV-1 recentinfections.TheKassanjeeetal.(2012)adjustment incidence testwiththeBED-captureEIA(EnzymeImmunoAssay)to above volunteerforaninterview, adiagnostictest(HIV-1/2) andan growing areas.Atotalof503seasonalworkersaged15yearsand Mali in2012. among seasonalworkersintheOfficeduNigerricegrowingarea aim ofthisstudywastoestimatetheoverallHIVinfectionincidence for indirectmeasuresofHIVincidenceinsomespecificgroups.The zero HIV-related deathandzerodiscrimination)havepavedtheway recent HIVinfectiontestsandthe3zerosvision(zeronewinfection, ment oftheimpactcontrolinterventions.Therecentdiscovery ed overallHIV was arecentinfectionusingtheBED-EIAincidencetest.Theestimat- infection, 3/503(0.6%)wereseropositiveandonly1ofthese3cases of theseasonalworkers[48.5%(244/503)]weresingle.RelativetoHIV favor ofmaleswithamedianage22years(15-68years).Almosthalf Centre HospitalierUniversitaireduPointG,Bamako,Mali Malaria ResearchCentreandTraining, Bamako,Mali Bamako, Mali Centre Nationald’AppuiàlaLutteContreMaladie, 2 Among the503studyvolunteers,sexratiowas1.46in , BrahimaDicko A cross-sectionalsurveywasconductedin2ofthe6rice 1 HIV incidenceisanimportantelementintheassess- The overallincidencewaslowamongseasonalwork- -1 incidencewaslowwith[0.37%,95%CI(0to1.10)]. Seasonal Workers intheRice Low IncidenceofHIV-1 among (presenting) 2 , SoryFomba 3 , OumarMaïga , Yaya Coulibaly 2 , KounandjiDiarra 1 , SambaSow 1 , MamoudouKodio 1 , BoubacarDiallo 1 1 , 1 , POSTER ABSTRACTS 97 Monitoring Self-Reported Determinants of Quality of Life in 385 Adherence to Antiretroviral 388 Newly Diagnosed HIV-Infected Therapy in the Brazilian HIV Care Facilities: Patients in Kenya National Cross-Sectional Measures of Anik Patel1 (presenting), Mia van der Kop2, David Ojakaa3, Adherence and Associations with Site Carlo Marra1, Patrick Igunza3, Richard Lester1 Organizational Characteristics 1 University of British Columbia, Vancouver, BC, Canada 2 Karolinska Institutet, Vancouver, BC, Canada Maria Altenfelder Santos1 (presenting), 3 AMREF, Nairobi, Kenya Maria Ines Battistella Nemes2, Cáritas Relva Basso2, Ernani Tiaraju Santa-Helena3, Francisco Assis Acurcio4, Background: Determinants of quality of life (QoL) in HIV-infected Gustavo Machado Rocha5, Maria das Graças Ceccato4, patients include disease severity, social support and coping mecha- Palmira Bonolo6, Felipe Vale2, Ana Alves2, Wania Carvalho2, nisms. Common symptoms of HIV may determine QoL and health Paulo Menezes2, Mark Drew Crosland Guimaraes4 ABSTRACTS state utilities. The objective of this study was to examine the influence of symptoms associated with HIV disease on physical, mental 1 University of Sao Paulo/Yale University, New Haven, CT, USA and health state utility scores of HIV infected patients in Kenya. 2 University of Sao Paulo, Sao Paulo, Brazil 3 Regional University of Blumenau, Blumenau, Methods: Between April and October 2013, an adapted Kiswahili ver- Santa Catarina, Brazil sion of the Short Form 12 (SF12) was administered to newly diag- 4 Federal University of Minas Gerais, Belo Horizonte, nosed HIV-infected patients participating in a randomized controlled Minas Gerais, Brazil trial in Nairobi, Kenya. Patients were also asked if they were experi- 5 Federal University of Sao Joao Del-Rei, Divinopolis, encing 19 common symptoms of HIV/AIDS (e.g. fatigue, loss of Minas Gerais, Brazil appetite, depression or diarrhea) on a scale of no symptoms, mild, 6 Federal University of Ouro Preto, Ouro Preto, Minas Gerais, Brazil moderate and severe symptoms. SF-12 survey responses were POSTER scored to derive a physical component score (PCS), a mental compo- Background: Adherence to recommendations for taking antiretrovi- nent score (MCS), and an overall utility score (SF6D). Linear regres- ral therapy (ART) is critical for good outcomes. We sought to under- sion was used to examine the association between each symptom stand the extent to which adherence varies across care sites. and each score. Methods: In 2010 we conducted a survey of 2,424 patients treated in Results: 135 respondents were included in the analysis with 7 obser- a nationally representative sample of 55 Brazilian public HIV care vations removed due to missing data. Severe fatigue was associated facilities. We selected a stratified sample of facilities in 7 regions in with 15.0 point (p <0.01) reduction in PCS while severe lack of Brazil within 7 performance levels that were determined in a previ- appetite was associated with an 8.6 point (p <0.01) reduction in PCS ous national evaluation of HIV care sites. Participants answered a compared patients not experiencing those symptoms. Those experi- pre-validated web-based questionnaire that included 3 questions encing severe depression had 12.2 point (p <0.01) lower MCS and about deviations from recommended ART in the past 7 days: if those experiencing severe anxiety had 8.4 point (p <0.01) lower MCS patients took fewer than recommended number of pills (dose), took compared to those not experiencing those symptoms. Severe fatigue medicine at non-recommended times (timing), and did not take 1 of (-0.10), depression (-0.18), fever (-0.12), numbness (-0.11) and lack of their medications (medication). We created a 4 category scale of appetite (-0.06) were all associated with lower SF6D scores (p <0.05). adherence: perfect adherence, non-adherence in 1 dimension, 2 dimensions, and 3 dimensions (i.e., dose, timing, medication). Conclusions: This study is the first to investigate symptoms that are Associations between non-adherence and site characteristics were associated with QoL outcomes and health state utilities in a cohort of estimated in multinomial logistic regression models, using weighted HIV-infected patients in Kenya. The drivers of lower physical, mental data to account for the sampling design. and overall health scores may be valuable to inform targeted clinical management and program planning efforts in HIV treatment and care. Results: Approximately half of participants reported non-adherence with respect to dose timing. Perfect adherence was reported by 38.9% (CI 95% 36.3-41.5) of participants, and some degree of non- adherence was reported by 61.1% of participants (30.8% on 1 measure, 18.6% on 2 measures, and 11.7% on 3 measures). The extent of non-adherence varied across sites. Site size (number of patients receiving ART) and region independently predicted non-adherence, with sites with 101 to 500 patients showing lower odds of non-adherence. Contrary to our expectations, lower odds of non-adherence were obtained in lower performance levels.

Conclusions: Brazilian HIV care facilities should implement strategies to support adherence among patients who are having difficulty following treatment recommendations, specially focusing on adherence to recommended dose timing. Associations between patient adherence and care site characteristics should be studied further.

98 9th International Conference on HIV Treatment and Prevention Adherence 390 edge andcomfortregardingPrEPprovision. providers, particularlyprimarycarephysicians,toincreaseknowl PrEP, educationaleffortsshouldbedirectedtowardspotential PrEP providers. To helpachievesuccessfulimplementationofreal-world across groups;however, PrEPknowledgewaslowamongnon-HIV 1 Douglas Krakower Providers? Difference betweenHIVandNon-HIV Prophylaxis forHIVPrevention:Istherea Conclusions: Results: Methods: Background: 3 2 Jill Blumenthal about PrEPdeliveryinthereal-world. (FTC/TDF) forpre-exposureprophylaxis(PrEP)hasraisedquestions ence tofollow-up;30%identifiedriskcompensationasaconcern. were drugtoxicities,developmentofresistanceandpatientadher- 0.49). ThemostcommonconcernsaboutPrEP(>40%ofproviders) prescription betweenHIVandnon-HIVproviders(70%vs.64%,P= there wasnodifferenceinthelikelihoodofendorsingfuturePrEP providers (31%vs.10%;p=0.001).Among150potentialprescribers, PrEP prescriptionwashigheramongHIVprovidersthannon-HIV icantly associatedwithhigherknowledgescores.Therateofprior <0.001). CDCguidanceandPrEPclinicaltrialsfamiliarityweresignif- = 0.002),andthosethatpreviouslyprescribedPEP(3.2vs.2.3,p <0.001), age>41(mean2.8vs.2.3;p=0.017),Whiterace(2.82.2; (max 5)weresignificantlyhigherfor:HIVproviders(2.9vs.2.2;p cians and55%wereHIVproviders.MeanPrEPknowledgescores predictor variables. assess associationsbetweenpastorfuturePrEPprescriptionand provider groupsandotherfactors.Fisher’s exacttestwasused to one-way-ANOV answers from5PrEPknowledgequestions.Univariatet-testand of HIVexperience.Knowledgescoreswerethesumcorrect defined asprimaryHIVprovidersand/orhavinggreaterthan5years barriers andmotivatorstofutureprovision.HIVproviderswereself- non-HIV providerknowledge/experiencewithPrEPandperceived related conferencesinCaliforniaandNewYork toassessHIVand 9th InternationalConference on Harvard University, Boston,MA,USA Beth IsraelDeaconessMedicalCenter, Boston,MA,USA University ofCalifornia,SanDiego,CA,USA Of 171respondents,themeanagewas41,58%werephysi- We gaveiPAD-based questionnairestoprovidersatHIV- Positive attitudestofuturePrEPprescriptionwerehigh The 2012FDAapprovalofemtricitabineandtenofovir Attitudes RegardingPre-Exposure Knowledge, Experienceand A wereusedtocompareknowledgescoresamong 1 (presenting) 2 , KennethMayer , SoniaJain HIV 3 T , RichardHaubrich reatment andPrevention Adherence 1 , XiaoyingSun 1 1 , - 391 1 ing 2012-2013at6clinicsinOromia,Ethiopia. various dimensionsofHIVbeliefsamongpersonsinitiatingARTdur- positive healthbehaviors,includingadherencetoART. We examined Background: 4 3 2 Olga Tymejczyk Antiretroviral TherapyinOromia,Ethiopia Conclusions: Results: relates ofaccuratebeliefs(averagescores>3)ineachdomain. strongly agree).Logisticregressionmodelswereusedtoassesscor- with responseson4-pointLikertscales(1=stronglydisagree-4 and thecurativepotentialofholywater(apopularbeliefinEthiopia), tor analysis:ARTandhealthoutcomes,transmissionrisk, initiation abouttheirbeliefsacross3domainsidentifiedthroughfac- Methods: Robert Remien Sarah GorrellKulkarni of accuratebeliefs. Christians =0.23,95%CI:0.12-0.46)wasassociatedwithlowerodds 95% CI:0.17-0.61)orEthiopianOrthodox(aORref=non-Orthodox beliefs. ForHolyW (aOR4-8 vs.2-3=0.43,95%CI:0.27-0.68)hadloweroddsofaccurate (aORref=primary =0.64,95%CI:0.51-0.81)andmoreclinicvisits 1.62, 95%CI:1.06-2.47),whereaspatientswithsecondaryeducation associated withhigheroddsofaccuratebeliefs(aORref=primary= accurate beliefs.ForARTandtransmissionrisk,noeducationwas 95% CI:0.43-0.74)weresignificantlyassociatedwithloweroddsof 0.70) andMuslimreligion(aORref=non-OrthodoxChristians=0.56, health outcomes,noeducation(aORref=primary=0.63,95%CI:0.57- had accuratebeliefsonARTandtransmissionrisk.For ART andhealthoutcomes(90.0%)holywater(66.8%),only13.4% Denis Nash not receivedin-depthHIVeducationsinceenrollmentincare. have missedtherecentemphasisontreatmentaspreventionor ly thosewithlonghistoriesofandpossiblegapsincare,whomay forced amongpatientsofalleducationalbackgrounds,andespecial risk offuturenon-adherence,HIVeducationshouldberoutinelyrein- mission andmanybelievedHolyWater couldcureHIV. To reducethe ART washigh,butfewpatientsknewthatitpreventssexualtrans- HIV CenterforClinicalandBehavioralStudies,NewYork, NY, USA New York, NY, USA ICAP-Columbia University, MailmanSchoolofPublicHealth, New York, NY, USA Mailman SchoolofPublicHealth,ColumbiaUniversity, CUNY SchoolofPublicHealth,NewYork, NY, USA Whereas mostrespondentsheldaccuratebeliefsregarding Adults (n=1,177)wereinterviewedwithin2weeksofART 1 At thetimeofARTinitiation,beliefinhealthbenefits Patient beliefsarecriticaltosuccessfuladoptionof among PatientsInitiating HIV CareandTreatment Beliefs 4 , BatyaElul 1 (presenting) ater 1 , beingapregnantwoman(aORref=men=0.32, , Tsigereda Gadisa 3 , Wafaa El-Sadr , SusieHoffman 3 , MariaLahuerta 3 , ZenebeMelaku 2 , 3 , 3 , - POSTER ABSTRACTS 99 Knowledge of and Experience with Reading between the Numbers: 392 HIV Pre-Exposure Prophylaxis 393 Identifying Disparities in the Virally among New York State Healthcare Providers Unsuppressed in an HIV Clinic in North Carolina Ethan Cowan1, Yvette Calderon1 (presenting), Kelly Chacon1, Sara Rahman1, Siddhi Mankame1, Puthiery Va1, Sophie Sohval1, Katherine Schafer1 (presenting), Jennifer Keller1, Jason Leider1 Erica Sickelbaugh1, Mary Beth Cox2, Aimee Wilkin1 1 Jacobi Medical Center, Bronx, NY, USA 1 Wake Forest University School of Medicine, Winston Salem, NC, USA Background: The safety and efficacy of HIV pre-exposure prophylax- 2 Communicable Disease Branch, North Carolina Department is (PrEP) has been established in several high-risk populations, of Health and Human Services, Raleigh, NC, USA underscoring the importance of its integration into clinical practice. However, the implementation of PrEP as a population-level preven- Background: Of the 1.1 million Americans living with HIV, only 25% ABSTRACTS tion strategy has not been demonstrated. Several questions remain, are virally suppressed. A more comprehensive description of unsup- particularly the local healthcare community’s knowledge, experi- pressed patients could help identify specific risk factors for targeted ence, and attitudes towards PrEP. This study assessed knowledge interventions to improve virologic control. and willingness to prescribe PrEP among New York State (NYS) healthcare providers. Methods: Wake Forest’s HIV clinic conducted a retrospective review of patients with 1 medical visit in the period 8/1/2012-7/31/2013 using Methods: A prospective, cross-sectional study was conducted on a ³ CAREWare. Data were stratified by demographic, clinical, and convenience sample of NYS providers. Eligible participants complet- socioeconomic variables of interest. The threshold for “unde- ed anonymous written or web-based surveys about experience, tectable” viral load (VL) was <200 copies/mL (last VL in period). knowledge, and willingness to prescribe PrEP. Results: Out of 1,909 patients, 408 (21%) were unsuppressed. The POSTER Results: The study population (n = 147) was 53.1% male, 23.1% White, unsuppressed were 60% male, 77% Black/African American (AA), 22.4% Asian, 14.3% Hispanic, and 4.8% Black. Practice settings 70% aged <50, and 34% uninsured. 74% of unsuppressed patients included primary care (42.9%), HIV clinic (17.7%), and emergency attended clinic within the past 6 months. medicine (12.2%). Levels of clinical practice included attending physician (46.9%), intern/resident (29.9%), and nurse AA patients were 2.1 times more likely to be unsuppressed compared practitioner/physician assistant (4.7%). Most participants correctly to non-AA patients (95% CI 1.67-2.76, p <0.001) and AA women were 1.7 defined PrEP (82.3%) and identified eligible patients. Additionally, times more likely than non-AA women to have detectable VL (95% CI most believe PrEP is effective (64.6%) and that their patients would 1.07-2.60, p = 0.02). AA patients composed 64% of the general clinic be interested (54.4%). However, only 43.5% feel comfortable starting population but 77% of the unsuppressed (p <0.0001). Women of child- their high-risk patients on PrEP, and only 12.2% had actually pre- bearing age (ages 20-39) were 1.5 times more likely than women of scribed PrEP. A majority of respondents expressed concerns about other ages to have viremia (95% CI 1.01-2.22, p = 0.047). Young men non-adherence (61.3%) and not having enough time during an (ages 20-29) were 2.6 times more likely to be unsuppressed than men of appointment session to discuss PrEP (51.0%). Most participants other ages (95% CI 1.77-3.81, p <0.001), and were overrepresented in the believe patients could be initiated and followed on PrEP in primary unsuppressed vs. general clinic population (20% vs. 11%, p <0.0001). care clinics (92.5%), HIV clinics (79.6%), STI/sexual health clinics (87.1%), and women’s health clinics (78.9%), but not in mental health Conclusions: African Americans, women of childbearing age, and clinics (70.7%) or the emergency room (70.7%). young men were disproportionately unsuppressed compared to the general clinic population. This review identified populations for which Conclusions: Despite knowledge of PrEP and recognition of its effi- further research and the development of targeted interventions are cacy, many NYS healthcare providers would not feel comfortable needed. While a more comprehensive description of the unsup- prescribing PrEP. Perceived barriers included non-adherence and pressed is necessary for better understanding ongoing viremia, the inadequate time in an appointment to discuss PrEP. Programs fact that 74% were seen within the past 6 months suggests that clinic addressing these barriers will be needed to facilitate the widespread attendance may not fully explain lack of virologic control. implementation of PrEP in the clinic.

100 9th International Conference on HIV Treatment and Prevention Adherence 394 Kathleen MacQueen Recommendations: Lessons Learned: Description: ence scale. tive approach:thedevelopmentofavalidatedpsychometricadher- ACASI) andmemoryaids(e.g.,diarycards).We proposeanalterna- measures havelargelyfocusedonmodeofdatacollection(e.g., Efforts toimprovevalidityandreliabilityofself-reportedadherence inexpensive, non-invasive,andallowsforimmediatefeedback. adherence remainsastandardcomponentoftrialsbecauseitis Introduction: 3 2 Elizabeth Tolley contractive ormultipurposepreventionvaginalringresearch. statistical analysesfordeterminingringefficacyinHIVprevention, use fordiscussioninadherencesupportsessions;itcouldalsoadvise could provideimmediatefeedbackforcorrectuseorpotentialnon- Ultimately and potentialitems,aswellnextstepsforscalevalidation. participation morebroadly. influenced byknowledge,attitudesandmotivationsspecifictotrial trial isnotmotivatedbyassuranceofhealthpromotion,butshouldbe have beenused,ringadherencewithinaplacebo-controlledclinical ence. Incontrasttoothercontextswithinwhichadherencescales items thatcouldtheoreticallyassessARV ability andadherencedata,weidentifyamultidimensionalsetof ment error es,” enablingminimizationofbothsystematicandrandommeasure- consistent repliesthatare“lesspronetosociopsychologicalbias precise thansingleitemonesbecausetheyelicitasetofinternally Validated multi-itemmeasurestendtobemorestable,reliableand more complexconstructthatmaynotbedirectlyobservable. the formofquestionsorstatementsthat,whencombined,measurea 1 in situationswheresuchreportsareintentionallyavoided. non-adherence ratherthanrelyingondirectreportof fying non-adherencebyfocusingonthemultiplefactors“driving” 9th InternationalConference on University ofConnecticut,Brighton,MI,USA Miriam Hospital/BrownMedicalSchool,Providence,RI,USA FHI 360,Durham,NC,USA , avalidandreliablemeasureofvaginalringadherence . Assuch,theymayprovideasuperiorstrategyforidenti Psychometric scalesarecomposedofmultipleitemsin Despite shortcomings,theassessmentofself-reported Psychometric AdherenceScale Development ofaVaginal Ring 1 (presenting) Drawing onbehavioraltheoryandexistingaccept- W 1 e describeourconceptualframework,domains , KathleenMorrow HIV T reatment andPrevention Adherence -based vaginalringadher 2 , KRivetAmico 3 , - - - 398 support forHIVcarelinkage. marginalized OOCHIV and directrecruitmentisfeasibleforfindinghard-to-reach,severely Amy Rock-Wohl Description: Introduction: Rhodri Dierst-Davies Medical CareinLosAngelesCounty HIV-Infected PersonsintoConsistentHIV Recommendations: Lessons Learned: out ofcare(OOC)HIV-infected personsandlinkthemtoHIVcare. are presented. for linkagetocareandhelpednegotiatebarriers.Preliminaryresults verified foraltersusingLACHIVsurveillancedata.Staffadvocated and recentlyreleasedfromjailnoHIVprovider copies/mL; newly-diagnosedandnotlinkedtocarewithin3months; months; noHIVcarevisitsin7-12monthsandmostrecentVL>200 ed anHIVmedicalvisit.OOCwasdefinedasnocarevisitsin>12 view andseedswerereimbursed$40whenareferredaltercomplet ers. Seeds/alterswereincentivized$40tocompleteabaselineinter social networkreferralsfromseeds,directlybystafforthroughfly- clinics/agencies. OOCpersons(alters)wererecruitedeitherthrough ment wasused.Recruiters(seeds)wererecruitedfromHIV 1 retained incareafter6months.Programacceptabilitywashigh. 16/22(76%) completed1medicalvisit,andofthose88%(n=8)were included 6hoursofstafftimeonaverage.Amongpatientsenrolled, copies/mL). Ittookanaverageof11daystolinkaclientcarethat months) andanelevatedVLpriortoenrollment(mean=55,536 Participants reportedanaverageof8sexualpartners(previous6 had previouslyattended3HIVclinics;andahistoryofART(64%). months (27%).Onaverage,participantshadbeenOOCfor9months; individuals whohavetradedsexfordrugsormoneyintheprevious6 (64%), thoserecentlyincarceratedhomelesspersons(73%)and ipants include:AfricanAmericans(41%),MSM(99%),theuninsured network referrals(n=20)ordirectrecruitment2).Enrolledpartic CA, USA Los AngelesCountyDepartmentofPublicHealth,Angeles, A combinationofsnowballsamplinganddirectrecruit- Project Engage(PE)wasdesignedtofindhard-to-reach Technique forFindingandLinking Project Engage:AnInnovative 1 To date,22OOCpersonswereenrolledviasocial A combinedmethodologyofsnowballsampling 1 -infected personsinordertoprovideneeded (presenting) , SalonikiJames 1 . OOCstatuswas , - - -

101 POSTER ABSTRACTS When Need Plus Supply Does Not Symptom Bothersomeness and 399 Equal Demand: Challenges in 401 Symptom Attribution in Adults on Uptake of Depression Treatment in HIV HIV Medications Clinical Care Holly Swan1 (presenting), Sarah McDannold1, Keith McInnes1, Mark Glickman2, Joel Reisman3, Allen Gifford1 Brian Pence1 (presenting), Evelyn Quinlivan1, Amy Heine1, Malaika Edwards2, Riddhi Modi3, Nathan Thielman2, 1 Veterans Health Administration, Bedford, MA, USA Bradley Gaynes2 2 Boston University and Department of Veteran Affairs, Bedford, MA, USA 1 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA 3 Department of Veteran Affairs, Bedford, MA, USA 2 Duke University, Durham, NC, USA 3 University of Alabama at Birmingham, Birmingham, AL, USA Background: Patients who attribute their symptoms to HIV medications - rather than disease - may be prone to switching antiretrovirals Background: Depression is highly prevalent among patients in HIV care ABSTRACTS (ARVs), and may experience poor retention/adherence to care. Certain and is associated with worse health behaviors and outcomes. Effective symptoms (e.g., nausea/vomiting) are more often experienced as a treatments for depression are readily available, yet depression remains side effect, but little is known about whether symptoms attributed to widely underdiagnosed and undertreated in HIV clinical care. medications are more bothersome. Symptom attribution relationships may be important in guiding symptom and medication management. Methods: As part of a multisite randomized trial of depression treatment integrated into HIV clinical care (the SLAM DUNC Study), we Methods: Adults on ARVs in 3 urban US outpatient clinics (N = 260) examined the proportion of positive depression screens that resulted completed survey responses about 20 symptoms common in HIV. For in study enrollment and reasons for non-enrollment. each symptom, participants rated symptom bothersomeness on a 1 “it doesn’t bother me” to 4 “it bothers me a lot” scale. Participants were Results: Over 33 months, patients presenting for care at 2 sites com- also asked whether they believed each symptom was a side effect of

POSTER pleted 9,765 PHQ-9 depression screens; 1,852 (19%) screens were their ARVs. Bivariate tests examined bothersomeness differences by positive for depression (PHQ-9 score 10) and 1,628 (88%) positive ³ symptom between those with and without attribution to ARVs. screens were assessed for study eligibility. Of positive screens that were assessed, 186 (11%) resulted in study enrollment. Of screens Results: Average symptom bothersomeness when attributed to ARVs not resulting in enrollment, in 649 cases (40%) the HIV provider did ranged from 2.51 (dizzy/lightheaded) to 3.23 (cough/shortness of not recommend enrollment (patient’s mental health managed by out- breath); when not attributed to ARVs, average bothersomeness side provider = 14%; HIV provider preferred to manage independent- ranged from 2 (hair loss/changes) to 2.98 (sleep difficulty). Adjusting ly = 7%; HIV provider thought mental health picture too complicated for the multiplicity of tests, 5 symptoms were significantly more both- = 5%; HIV provider thought patient was not depressed = 5%); in 431 ersome between those attributing them to ARV side effects and those cases (26%) the patient declined to enroll (not interested in treatment attributing them to other causes (all p <.01): shortness of = 8%; wanted more time to consider = 6%; did not want medications breath/cough [mean = 3.23 (n = 22) vs. mean = 2.63 (n = 62)], weight = 5%); and in 360 cases (22%) the patient did not meet study inclusion loss/wasting [3.21 (n = 24) vs. 2.46 (n=28)], hair loss/changes [3.17 (n = criteria (current/past bipolar = 7% or psychotic disorder = 4%; not 18) vs. 2.00 (n = 17)], skin problems/rash [3.09 (n = 46) vs. 2.59 (n = 58)], taking antiretrovirals = 4%). While these results reflect multiple and weight gain/fat deposits [3.06 (n = 51) vs. 2.42 (n = 33)]. No symp- screens per patient, analyses restricted to narrower time windows toms were significantly less bothersome when attributed to ARVs. which largely reflecting unique patients yielded similar results.

Conclusions: Despite a high burden of untreated or undertreated Conclusions: Five symptoms were significantly more bothersome in depression, only 1 in 9 positive depression screens resulted in enroll- people who attributed the symptoms to medications rather than ment in the depression treatment study. While some of the reasons for something else (e.g., disease). These findings may have implications low enrollment were study-specific, others speak to potentially modi- for ARV medication adherence management and outcomes. fiable patient- and provider-level barriers to depression treatment Incorporating patient beliefs about causes of symptoms and side engagement. Addressing such barriers will be critical to maximize the effects may contribute to improved symptom and medication man- reach of depression treatment services for HIV-positive patients. agement, and to better ARV medication adherence.

102 9th International Conference on HIV Treatment and Prevention Adherence 410 their perceptionsof“incare”donotalignwithnationalstandards. well. PatientsmaynotberetainedinHIVcareatleastpartbecause seen morefrequentlythanguidelinesrecommend,evenwhenfeeing VL suppression.Morethanhalfofpatientsbelievedtheyneededtobe being inregularcare,andalmostthree-quartersofthemdidnothave Hospitalized HIV-Positive Patients Conclusions: Results: copies/mL wasconsideredVLsuppression. HIV doctorwhentheyarefeelingwell.”ConcurrentVL<200 asking howoftenpersonswithHIVinfectionshould“beseenbytheir HIV careonaregularbasis”duringthepast12-months,and1item survey included1itemaskingwhetherthepatient“gotoutpatient tient visitin<2ofthe4quarterspriortohospitalization.Thebaseline vention trialwereused.Outofcarewasdefinedashavinganoutpa- patients whowereoutofcareandenrolledintoarandomizedinter- Methods: VL suppression. care andevaluatedtheassociationbetweentheseperceptionswith are unknown.We examinedpatientperceptionsaboutbeinginHIV about howoftensuchpatientsfeeltheyshouldseetheirHIVprovider patients whoareoutofcarehasnotbeenexplored.Further, beliefs Introduction: 3 2 1 Jessica Davila Melinda Stanley 10% >6months;therewasnodifferencebyVLstatus(p=0.06). should beseenmonthlyormorefrequently 52% ofrespondentsthoughtthatHIV-infected personsfeelingwell patients whoreportedreceivingregularcarehadVL<200(p<0.01). did notreportbeinginregularcarehadVL<200copies/mL,27%of outpatient HIVcare“onaregularbasis.”While7%ofpersonswho of carebasedonquartersincare,however, 47.8%reportedreceiving 9th InternationalConference on University ofConnecticut,Brighton,MI,USA DeBakey VA MedicalCenter, Houston,TX,USA Baylor CollegeofMedicine,Houston,TX,USA 407 patientscompletedthesurvey Baseline datafromacohortofhospitalizedHIV-infected About 50%ofpatientswhoareoutHIVcarereport Perceived engagementinoutpatientHIVcareamong in OutpatientCareamong Patient PerceptionsofEngagement 1 , ChristineHartman 1 , KRivetAmico 3 HIV , Thomas Giordano 2 , JeffreyCully T reatment andPrevention Adherence , 38%every3months,and . 98.5%(n=401)wereout 1 , 3 (presenting) 411 3. Community Clinic Intervention PilotedinanInner-City Single-Session Retention-Promotion tion inHIVcare. develop andrigorouslytestinterventionsforpromotingpatientreten may offerguidancetoresearchersandpractitionersseeking investigation, thismanualized60-minutetheory-basedintervention 2. 1. unique sections: Information-Motivation-Behavioral Skillsmodel,comprisedof4 Intervention contentwasinformedbyasituatedapplicationofthe care interventionoratime-and-attentioncontrolcondition. a historyofpoorretentionwererandomized1:1tothein ical staffinacommunityclinic.AsmallsampleofPLWH (N=16)with based, 60-minuteintervention,developedtobedeliveredbynon-clin Introduction: 4 3 2 Laramie Smith Recommendations: Lessons Learned: Description: 1 ue tobemonitoredthroughApril2014. (M=3.64,SD=1.19). Changesinretentioncarebehaviorswillcontin- (M=4.00, SD=.093),andlikelytobeusefulinthenext6-months 5=Strongly agree),patientsperceiveditaspersonallyrelevant community clinic.Usinga5-pointscale(1=Stronglydisagree, oped ortestedinterventionstoimproveretentionincarebehaviors. factors associatedwithpoorretention,fewstudieshaveactuallydevel- ber ofstudieshavebeenconductedtodescriberetentionratesand (PLWH) arepoorlyretainedinroutineHIVmedicalcare.Whileanum- Thomas Patterson 4. practiced insession. ence behaviors.Skillsformanagingaffect-relatedbarriersare about livingwithHIVrelatesto health concernsmostimportanttothepatient. Bronx, NY, USA Montefiore MedicalCenter/AlbertEinsteinCollegeofMedicine, University ofConnecticut,Storrs,CT, USA University ofConnecticut,Brighton,MI,USA University ofCalifornia,SanDiego,CA,USA Future well-being HIV Emotional health Physical health lar barrierstoimprovefuturevisitattendance. poor retentionandproactivelystrategizeswaystoaddresssimi- the next6-months. in sections1-3toimplementpatient-identifiedhealthgoalsover -care history In theBronx,NY, wepilotedaproof-of-concept,theory- In theUSA,approximately50%ofpersonslivingwithHIV Care Interventions?Discussionofa How AreWe BuildingRetentionin 1 (presenting) 1 - reviewsandprioritizesHIVnon-HIVrelated The interventionwasfeasiblyimplementedina , ChinazoCunningham - identifiesissuesrelatedtorecentperiodsof - exploreshowpositiveandnegativeaffect While theefficacyofthissmallpilotisunder - developsaplanbasedoninformationelicited , KRivetAmico -care andtreatmentadher 4 2 , JeffreyFisher 3 , - - -

103 POSTER ABSTRACTS Framework for a Coherent, “Trials and Tribulations of 412 Validated, and Feasible Adherence 415 Finding that Friend”: Measurement and Support Package for Use Developing a Smartphone Application to in ARV-Based Vaginal Ring Trials Communicate Peer-Generated Strategies for Building Social Support Kathleen MacQueen1 (presenting), Elizabeth Tolley1, Derek Owen1, K Rivet Amico2, Kathleen Morrow3, Thomas Moench4 Erin Plews-Ogan1 (presenting), Tabor Flickinger1, Erin Wispelwey1, Colleen Laurence1, Karen Ingersoll1, Rebecca Dillingham1 1 FHI 360, Durham, NC, USA 2 University of Connecticut, Brighton, MI, USA 1 University of Virginia, Charlottesville, VA, USA 3 Miriam Hospital/Brown Medical School, Providence, RI, USA 4 ReProtect, Baltimore, MD, USA Introduction: The value of social support in coping with HIV and maintaining adherence is firmly recognized, as is the potential posi- Introduction: Long-acting ARV-based vaginal rings for HIV preven- ABSTRACTS tive role of HIV status disclosure in recruiting this support. However, tion may simplify use instructions and require less user action, there- among a patient population in rural western Virginia, social stigma by facilitating adherence. However, participant motivations, counsel- and limited knowledge about HIV sharpen the risk of rejection as a ing support, and measurement challenges during ring trials must still result of disclosure. To address these challenges, we are developing be addressed. We propose a framework for methodically accelerat- and piloting video and community message board features within the ing work in this area. Positive Links Smartphone application to communicate peer-generated stories and strategies. Description: To-date, adherence support and assessment procedures in microbicide, PrEP and ring trials have not benefited from rigorous Description: Qualitative interviews with 17 HIV-positive participants evaluation identifying effective approaches. As such, the current evi- explored the role of social support during the year following their dence base is challenged by poor understanding of motivations for diagnosis and how a smartphone application might support positive

POSTER adherence within placebo-controlled clinical trials and, relatedly, coping strategies and medication adherence. We then collected strategies to promote product adherence specifically within clinical insights and stories expressed through art, voice and video record- trial settings. Nonetheless, process-models exist to guide the devel- ings to communicate key messages about building support through opment of behavioral interventions and measurement strategies. An the voice of peers. Using the methodologies of Photovoice and important first step is the articulation of a basic framework where crit- Appreciative Inquiry, we conducted a focus group and in-depth indi- ical areas (targets) are identified and relevant evidence, best prac- vidual interviews to gather participants’ artwork depicting social sup- tices, and innovation within each are defined. Drawing from available port and their personal experiences and advice. We sought to capture evidence, experiences and cross-cutting fields of inquiry, we devel- strategies for recruiting support, especially considerations regarding oped a framework for vaginal ring adherence identifying key areas status disclosure, and provide a vehicle for these experiences and and targeted research needed to produce effective and generalizable insights through video and community message board features. adherence promotion and assessment strategies. Lessons Learned: Initially drawing on a few trusted relationships Lessons Learned: Three critical areas need to be addressed. First, strengthened participants’ self-confidence and emboldened them to adherence support requires sequential efforts to define motivators of risk disclosing to others who might react negatively. Participants study-product adherence; develop, test, and refine adherence sup- emphasized that disclosure is not the only pathway to support and port messages and; evaluate these in a rigorous fashion. Second, the must not be understood as a mandate, but instead requires nuanced limitations of single-item self-report adherence measurement high- considerations about the specific relationship and each party’s lights the need for improved psychometric approaches. These multi- readiness to cope with the disclosure. item scales, once validated, are easily administered and may capture vaginal ring use with improved predictive ability at screening, base- Recommendations: The impact of this video and the community mes- line and follow-up. Third, real-time adherence monitoring and cumu- sage board will be qualitatively assessed through interim usability lative measurement to correlate adherence with overall product and final evaluation interviews, as well as analysis of app metadata effectiveness requires innovative designs, models, and prototypes capturing exchanges of social support through the community mes- using “smart” and biometric technologies to detect ring insertion sage board. and/or removal.

Recommendations: Interdisciplinary collaboration is needed to create a coherent, validated, and feasible measurement and support framework for future vaginal ring trials.

104 9th International Conference on HIV Treatment and Prevention Adherence 417 HIV andLowHealthLiteracy Antiretrovirals amongPeopleLivingwith adherence inpeoplelivingwithHIVandlowliteracy(PLWH-LL). the utilityofsingleitemvisualanalogscale(VAS) forassessingART forming mentalcalculations.Thepurposeofthisstudywastoexamine particularly challengingduetothecognitivedemandsofrecallandper- Conclusions: Results: Methods: Background: Jennifer Pellowski hol useamongpersonswithlowliteracy although itsutilitymaybecompromisedbylimitedincomeandalco- lates ofadherence.HIVRNAviralloadswerechart-abstracted. monthly unannouncedpillcounts(UPC)and(4)measuresofcorre a computer-delivered VAS, (2)3monthlyphone-deliveredVAS, (3)3 ed theTest ofFunctionalHealthLiteracyforAdults(TOFHLA)and(1) 1 the VAS andUPC(p=0.024,p0.008,respectively). alcohol useweresignificantpredictorsofthediscrepancybetween mon correlatesofadherence.Lowerincomeandhigherrates the VAS andUPC,multivariateregressionwasconductedwithcom- 0.366, p<0.001).To determinepredictorsofthediscrepancybetween RNA viralload(phoneV Finally, theVAS demonstratedgoodcriterion-relatedvaliditywithHIV comparable validitywhenadministeredbyphoneandcomputer. = -5.23,p<0.001).Additionally, theVAS demonstratedfavorableand V significantly differentbyliteracylevel,suchthattheUPCandPhone current validitywiththeUPCs(r=0.66,p<0.001).However retest reliability(r=0.52).Thephone-deliveredV 9th InternationalConference on AS correlatedlesswellforthosewiththelowestlevelsofliteracy(z University ofConnecticut,Storrs,CT, USA The phone-deliveredVAS demonstratedgood1-monthtest- This studywasconductedwith471PLWH-LL whocomplet- Assessing adherenceinlow-literacypopulationsmaybe Assess Adherenceto Using theVisual AnalogScaleto The VAS wasreliableandvalidamongPLWH-LL, 1 (presenting) AS: r=-0.250,p<0.001;computerV HIV , SethKalichman T reatment andPrevention Adherence . AS hadgoodcon 1 , DavidFinitsis , thiswas AS: r=- 1 - - 418 Study inCalifornia Methods: Background: Shannon Fuller engagement inthesetypesofsettings. shared decision-makingmaytakeamorevariableroleinpatient shared decision-making,ourpreliminaryfindingssuggestthat Conclusions: Results: Shane Collins health careamongthesepatients. clinics andtoidentifyfactorsthatleadmoreorlessengagementin We soughttoassesspatientengagementinthecontextofHIVcare engagement havetendedtofocusonpatientsinprimarycareclinics. better healthoutcomesinavarietyofconditions.Previousstudieson decision-making betweenpatientsandproviders,isassociatedwith 1 home activitiesinthe5HIVclinics. gation waspartofalargerstudyevaluatingpatient-centeredmedical grounded theoryapproachtoidentifyemergingthemes.Thisinvesti in eachofthe5clinics(n=68).We analyzedtranscriptsusinga speaking patientsage18orolder, and2)providersorotherclinicians interviews at5HIVclinicsinCaliforniawith:1)HIV-positive, English- spectrum ofactivitiesandbehaviorsthatsignifyengagementincare. and overallclinic.We presenttypologiesandcasestudiestoshowa future-oriented thinkingandatrustingrelationshipwiththeprovider engagement includehealthliteracy, computerandInternetskills, provider outsideofregularclinichoursarecommon.Facilitatorsto consultant. Openlinesofcommunicationbetweenthepatientand responsibility fortheircareandimagineproviderasaguideor Some monitortrendsintheirlabvalues.Patientsoftenassumefull research online,adheringtotreatmentandkeepingappointments. friends andprovidersquestionsabouthealthconcerns,doing narratives; however, patientsseemengagedincarebyaskingboth UCSF CenterforAIDSPreventionStudies,SanFrancisco,CA,USA Instances ofshareddecision-makingappearedrarelyinthe Our teamofqualitativeresearchersconductedin-depth Patient engagement,typicallycharacterizedasshared 1 in PublicHIVClinics:AQualitative Understanding PatientEngagement , PatrickHazelton Although patientengagementisoftendefinedas 1 (presenting) , KimberlyKoester 1 , StarleyShade 1 , RyanGuinness 1 , Wayne Steward 1 , 1 -

105 POSTER ABSTRACTS Developing a Conceptual Model for An Information-Motivation- 419 ART Adherence among Kenyan 420 Behavior Skills (IMB) Measure of Men Who Have Sex with Men “Retention in HIV Care” for Use in the Mentor Approach for Promoting Patient Self- Susan Graham1 (presenting), Murugi Micheni2, Elise van der Elst3, Bernadette Kombo3, Don Operario4, Eduard Sanders3, Jane Simoni1 Care (MAPPS) Study of Out-of-Care Persons Living with HIV 1 University of Washington, Seattle, WA, USA 2 Kenya Medical Research Institute, Mombasa, Kenya Michael Kallen1, K Rivet Amico2 (presenting), Melinda Stanley3, 3 Kenya Medical Research Institute, Kilifi, Kenya Jeffrey Cully3, Jessica Davila3, Thomas Giordano3 4 Brown University, Providence, RI, USA 1 Northwestern University, Chicago, IL, USA Background: Men who have sex with men (MSM) are highly stigma- 2 University of Connecticut, Brighton, MI, USA tized and often criminalized in sub-Saharan African, and may have 3 Baylor College of Medicine, Houston, TX, USA ABSTRACTS difficulty accessing quality HIV care. To achieve an understanding of antiretroviral therapy (ART) adherence in this group, a sociocultural- Background: The IMB Model is a well-known health behavior model ly valid conceptual model is needed. that has recently been applied to retention in HIV care. Operationalizing and measuring retention-related information, moti- Methods: Our proposed conceptual model was adapted from an vation, and skills is critical to documenting the influence of IMB- access-information-motivation-behavioral skills (access-IMB) model model based interventions, yet there are few validated measures proposed by Simoni et al. We posited that stigma’s effects would be presently available to do so. We developed and evaluated an IMB- moderated by psychosocial support, when available. We explored this model based “retention in HIV care” measure to address this gap. model of ART adherence by analyzing qualitative data from 28 interviews with HIV-positive Kenyan MSM, of whom 16 were taking ART. To Methods: English-speaking, out-of-care, HIV-positive individuals evaluate our model, we used the heuristic schema of Ware et al.,

POSTER completed an IMB-retention in care measure consisting of 12 infor- which asks 4 analytic questions about model validity and relevance: (1) mation, 23 motivation and 17 behavior skills items. Information index Are the model’s basic concepts relevant in the setting? (2) Are all basic item responses were scored as correct/incorrect and summed. concepts important to the setting represented? (3) Are the meanings of Motivation and Behavior Skills scale item responses, ranging from 1 the model’s basic concepts accurate in the setting? and (4) Does the (“strongly agree”) to 6 (“strongly disagree”), were fit to unidimen- model capture the complexity of adherence in this setting? sional confirmatory factor analysis models; items with low factor loadings (<0.50) were excluded. Responses to retained scale items Results: The basic concepts of the access-IMB model were relevant were assessed for internal consistency reliability, and average to this population, in which stigma related to HIV infection and to scores calculated. Spearman correlations were computed among male-male sex negatively impacts men’s adherence. While psy- the 3 IMB components. chosocial support from providers and others was important, we found that the concept of resiliency, described by Herrick et al as Results: Three hundred seventy-nine persons completed the survey, coping with negative experiences and avoiding negative trajectories, which was administered by audio-computer assisted self-interview was a key moderator of men’s adherence. Meanings of the model’s (A-CASI). 6 Information items were excluded, due to low IMB-model concepts were accurate in this setting, but resiliency-related con- component association (r < 0.10). The resultant 6-item Information cepts such as self-monitoring, adaptability, and optimism emerged as index had a median score of 6 (min = 2, max = 6); 60.8% got all items critical factors affecting adherence. correct. Four Motivation items and 2 Behavior Skills items were excluded, due to low internal validity. The resultant 19-item Motivation Conclusions: Our data support the utility of an access-IMB model as scale (alpha = 0.92) had a mean score of 4.45 (SD = 0.89); the resultant a potential framework for understanding ART adherence in this pop- 15-item Behavior Skills scale (alpha = 0.87) averaged 4.44 (SD = 0.80). ulation. However, data indicate a need to consider the pervasive stig- Intra-IMB scale correlations were as follows: Information-Motivation, ma and discrimination Kenyan MSM experience, as well as their rho = 0.18 (p = 0.001); Information-Behavior Skills, rho = 0.17 (p = 0.002); innate resiliencies. Motivation-Behavior Skills, rho = 0.75 (p = 0.000).

Conclusions: Consistent with IMB model hypotheses, Information, Motivation, and Behavior Skills scores were significantly inter-correlated. The IMB Retention in HIV Care measure is currently being used to evaluate the impact of a re-engagement and retention in care intervention targeting the IMB constructs.

106 9th International Conference on HIV Treatment and Prevention Adherence 421 patients atincreasedrisk. engagement andadherence,enhancingactivemonitoringof delivery includeincorporatingvisualaids,gaugingfacilitatorsof of providerpractices.Potentialopportunitiestoenhanceservice clinical caresitesdemonstratedbothconsistencyandheterogeneity in theUnitedStates. SOC surveywasofferedtoprovidersat4largeHIVoutpatientclinics established, andantiretroviraltherapy(ART)initiatingpatients.The rent strategiesusedtopromoteengagementinHIVcareamongnew, 1 Anne Zinski Patients inHIVCare Adherence StandardofCareSupportfor Conclusions: Results: Methodology: Background: 6 5 4 3 2 Laramie Smith K RivetAmico ence orretentionproblemswiththeirpatientswasreportedby<30%. Feedback fromclinicdatabasesalertingproviderstopotentialadher- ed notaskingaboutadherenceunlessthereisasuspectedproblem. ence barriers,whileonly28%askaboutfacilitators,and11%report- patients toestimatetheiradherence,66%askaboutspecificadher- about safer substance usewithinaharmreductionframeworkand51%inquire some (34%)alsousingvisualaids(pictures,videos).56%discuss Information waslargelydeliveredviaverbaldiscussion(100%),with tions, and56%askedaboutbarrierstocomingintocarevisits. although nearlyallincorporateviralload,CD4count,andmedica- ered byotherstaff.Educationalcontentwasheterogeneous, tion tonew-to-carepatients;while8%reportedmaterialbeingcov- utes). Mostproviders(85%)reporteddeliveringbasicpatienteduca guidelines, todocumentpre-interventionSOC. delivered survey mote engagementamongHIVcareinitiators,wedevelopedaweb- ration foralargerandomizedcontrolledtrialofaninterventiontopro- gies, services,andprogramsinagivenclinicareneeded.Inprepa- not wellcharacterized,andmeasuresthatdocumenttheSOCstrate ment inHIV Katya Roytburd 9th InternationalConference on University ofWashington, Seattle,WA, USA Johns HopkinsUniversity, SchoolofMedicine,Baltimore,MD,USA University ofCaliforniaSanDiego,CA,USA University ofNorthCarolinaatChapelHill,NC,USA University ofAlabamaatBirmingham,AL,USA University ofConnecticut,Brighton,MI,USA Sixty-two providerscompletedthesurvey(median15min -sex practicesateachvisit.ForpatientsonART -care (retentionandadherence)intheUnitedStatesare 2 , MichaelMugavero Current standardsofcare(SOC)foroptimizingengage- Evaluation ofSOCengagementpracticesacross4HIV 1 4 Characterize Retentionand A BriefProviderSurveyto A diverseexpertpanelconstructedameasureofcur- , JeanneKeruly 3 (presenting) , SolongeMontue , guidedbyinternationalevidenceinformedpractice , RiddhiModi 5 , EvelynQuinlivan HIV 2 5 , RobFredericksen T reatment andPrevention Adherence 2 , CarolGolin 3 , HeidiCrane 6 3 , , , 87%ask 7 , - - - 422 =.00) thantheRDScommunitysamples. a previousHIV-test (p=.04)andtoknowtheywereseropositive accessing primary-careserviceswerebothmorelikelytohavehad =.00); neithergroupwasmorelikelytoknowtheirstatus.Participants likely tohavepreviouslytestedforHIVcomparednon-PWIDs(p previous HIVtestthannon-MSM(p=.01),whereasPWIDswereless knew theywereHIV+(p=.02).MSMmorelikelytohavehada treatment. Observeddifferencessuggestwomenlessfrequently meaning only3.7%ofallidentifiedHIV+personswereonactive linked tocare,33.3%reportedtheywerecurrentlyreceivingART, reported initiatingHIV who hadevertested,24.2%knewtheirHIV+status,95.5%ofwhich 47.6% reportedeverhavingpreviouslybeentestedforHIV. Ofthose (RDS). AmongHIV+personsinthesecommunitysamples(N=191), among MSMwhowereidentifiedviarespondentdrivensampling community clinicto4.6%amongPWID,5.0%FSWand17.4% among indigent/high-riskpersonsseekingprimarycareservicesata Background: 3 2 Laramie Smith Treatment CascadeinTijuana, Mexico Levels ofEngagementinCareacrosstheHIV Conclusions: Results: Methods: risk populations. known aboutthestateofengagementinHIV ers (FSW),andmenwhohavesexwith(MSM).However centrated amongpersonswhoinjectdrugs(PWID),femalesexwork- prevalence thatis2.6timeshigherthanthegeneralpopulation,con- tion, deportation,druguseandsextourismcontributetoalocalHIV 1 Carlos Magis-Rodriguez methods mayincreaseroutinetestingandsubsequentARTuptake. tinuum inTijuana. Integrating test-and-treatinterventionswithinsuch ods mayhelptoidentifyHIV+personsnotretainedintheHIVcarecon- Sarah Rojas tion ofindividualsretainedacrosstheHIV-care continuum. nity-based studies(N=3,668,191wereHIV+)toexaminethepropor San DiegoStateUniversity, SanDiego,CA,USA SIDA, MexicoCity, FederalDistrict,Mexico Centro NacionalparalaPrevenciónyelControldelVIH University ofCalifornia,SanDiego,CA,USA Overall, HIVprevalenceaveraged5.2%,rangingfrom3.0% W 1 e pooledavailableepidemiologicaldatafrom6commu , MaríaLuisaZúñiga Tijuana isamajorMexico-USbordercitywheremigra- Findings suggestcommunity-basedsurveillancemeth- Surveillance DatatoExplore Utilizing Community-Based 1 (presenting) -care post-diagnosis(i.e.,linked).Ofthose 2 , Victoria Ojeda , ThomasPatterson 3 , SteffanieStrathdee 1 , JoseLuisBurgos -care amongthesehigh- 1 , 1 1 , , littleis - -

107 POSTER ABSTRACTS The Depression Treatment Gap Attributes of Effective Patient- 423 among HIV-Infected Patients in 424 Reported Medication Adherence Routine HIV Care in a US Clinical Cohort Measures across Diseases for Use with Patients with Chronic Co-Morbidities Rushina Cholera1 (presenting), Brian Pence2, Heidi Crane3, Riddhi Modi4, Rob Fredericksen3, Michael Mugavero4 Rob Fredericksen1 (presenting), Anna Harrington1, Matthew Jackson1, Anna Church1, Susan Yount2, Elizabeth Hahn2, 1 University of North Carolina Gillings School of Global Public Sofia Garcia2, Mimiaga Matthew3, Honghu Liu4, Jane Simoni5, Health, Chapel Hill, NC, USA Kenneth Mayer6, Mari Kitahata1, Donald Patrick1, Heidi Crane1 2 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA 3 University of Washington, Seattle, WA, USA 1 University of Washington, Seattle, WA, USA 4 University of Alabama at Birmingham, Birmingham, AL, USA 2 Northwestern University, Chicago, IL, USA 3 Harvard University, Cambridge, MA, USA Background: Depression affects 20-30% of HIV-infected patients, 4 UCLA School of Dentistry, Los Angeles, CA, USA ABSTRACTS and has been associated with reduced antiretroviral adherence and 5 University of Washington, Seattle, WA, USA poor health outcomes. Although effective depression treatment is 6 Fenway Community Health/Brown University, Boston, MA, USA available, up to 95% of HIV-infected depressed patients nationally are either untreated or undertreated (i.e., fail to achieve remission). Introduction: Patients are increasingly accruing multiple co-morbidities We quantified the depression treatment gap among HIV-infected for which medication adherence is key to averting illness and/or death. patients in the nationally distributed US CNICS observational HIV A practical cross-disease approach to measuring patient-reported clinical cohort. adherence is needed. Adherence measures in routine clinical care should be clinically relevant, brief, valid, and easy to understand. Methods: The CNICS cohort combines detailed clinical data with regular, self-reported depressive severity assessments using Patient Description: In order to advance a cross-disease approach and learn Health Questionnaire-9 (PHQ-9). HIV-infected patients who newly

POSTER from adherence work done across fields, we reviewed the most entered HIV medical care at a CNICS site from 2004-2013 and had a widely cited adherence items across disease categories, and identi- PHQ-9 assessment within 6 months of entering care were included. fied attributes of measures likely to be useful in clinical care for aging patients with multiple co-morbidities. Results: 1,884 patients at 6 study sites were eligible (mean age = 45, 88% male, 48% White, 28% Black, 19% Hispanic). 703 participants Lessons Learned: Measures of self-reported medication adherence (37%) either started anti-depressants (ADs) prior to PHQ-9 screening varied widely by disease with varying feasibility for use in clinical care. (n = 238) or scored ³10 (indicating likely major depression) on the PHQ-9 (n = 465). Of 465 depressed AD-naïve participants, 17% (n = 79) • Many measures were developed initially for research, not clini- initiated ADs within 1 month and 29% (n = 133) within 6 months. cal care. Among 133 participants who started ADs within 6 months and had a • Many measures included several dimensions of adherence, repeat PHQ-9 assessment within 12 months (n = 86), 14 (16%) including self-efficacy, side effects, reasons for stopping, and achieved depression remission (PHQ-9 score 0-4) and 33 (38%) had a attitudes about need for medication. Brevity was rare, limiting score <10. Of 320 participants who had a high PHQ-9 score but did practicality for clinical care. not initiate ADs and had a repeat PHQ-9 screening within 12 months • Negative focus was emphasized in content of many widely-used (n = 179), 39 (22%) achieved remission and 88 (49%) scored <10. instruments, i.e., references to missed doses or assessing own Conclusions: Over one-third of patients entering the CNICS cohort ‘carelessness.’ had an indication for depression treatment, and about one-third of • “Disease-specific” adherence measures rarely differed from those received treatment. Remission rates were similar among general adherence measures beyond cosmetic differences, i.e., depressed patients who did and did not initiate ADs, suggesting cli- insertion of name of medication class. Diabetes was the excep- nicians may be distinguishing patients who might benefit most from tion, often with blood glucose level items. ADs. Regardless of AD treatment, nearly 80% of all depressed • Single-item and self-rating measures were most commonly used patients did not achieve remission within 12 months. in HIV. • Recall periods varied widely. • Scoring systems of multi-item measures were too general/difficult to interpret, e.g., rendering results like “adherent” when not fully adherent, or “low adherence,” the meaning of which may differ across medications classes.

Conclusions: Across many diseases, the lack of clinically relevant adherence measures lends support for developing generalized, brief adherence measures for use with patients with multiple chronic comorbidities, for which a patient may be able to answer for each medication class.

108 9th International Conference on HIV Treatment and Prevention Adherence 425 for HIVAdherenceInterventions A Mixed-MethodsStudyandItsImplications Robert Savage Conclusions: Results: Methods: Background: 6 5 4 3 2 1 Mirjam-Colette Kempf Erin Stevens ence anddepressionwillbediscussed. of thesefindingsforevidenced-basedtreatmentHIVcareadher- reports difficultiesinengagingdepressiontreatment.Implications strategies forHIV-infected AAwomen,apopulationthathistorically which maybeimportantinthedevelopmentofdepressiontreatment (e.g., prayer primary copingstrategy, resultsindicatethatbeliefsandpractices tractions. Additionally sion werespirituality(i.e.,faithandprayer),socialsupport,dis strategies usedbyparticipantsfordealingwiththeirHIVanddepres- ses wereconductedusingQRSNVivo 10 mixed methodsstudy coping strategies(BriefCOPE)andspiritualbeliefs(SBI-15R).Inthis telemedicine andcompletedself-reportquestionnairesassessing services inruralAlabamacounties.Participantswereinterviewedvia ry ofdepressionwererecruitedfromoutpatientclinicsproviding utilized byAfrican-American(AA)womenintheruralSoutheast. current studywastoexplorepositiveandnegativecopingstrategies sion amongHIV-infected womenisnecessary. Theobjectiveofthe adherence, abetterunderstandingofcopingstrategieswithdepres HIV care,withprevalenceofdepressionashigh70%.T associated withnon-adherencetoantiretroviraltherapy(ART)and were BehavioralDisengagement,Humor Reframing, andPlanning.Theleastutilizedcopingstrategiesreported disease anddepression:Religion,Distraction,Acceptance,Positive are mostlikelytoutilizethefollowingstrategiescopewiththeirHIV statistics supportedqualitativefindings,andsuggestedparticipants from theirrespectivereligiouscommunities.Quantitativedescriptive sion amongAAwomenintheSouththansocialsupportderived 9th InternationalConference on Birmingham, AL,USA Nursing andHealthBehavior, UniversityofAlabamaat Massachusetts GeneralHospital,Boston,MA,USA Birmingham, AL,USA Department ofMedicine,UniversityAlabamaatBirmingham, Harvard MedicalSchool,Boston,MA,USA Birmingham, AL,USA School ofNursing,UniversityAlabamaatBirmingham, Birmingham, AL,USA Department ofPsychiatry, UniversityofAlabamaatBirmingham, Qualitative analysesrevealedthatthemostutilizedcoping Ten HIV-infected AAwomen(M=37yearsold)withahisto- , faith)maybemorecentraltocopingwithHIVanddepres 1 , CorilynOtt Positive Women withDepression: Coping StrategiesamongHIV- Among HIV-infected women,depressionisstrongly Findings fromthecurrentstudyhighlightstrategies 1 , NicholasVan Wagoner , usingaconvergentparallelstudydesign,analy- , whileparticipantsdescribedspiritualityasthe 6 (presenting) 2 , ChristinaPsaros HIV T reatment andPrevention Adherence , andSubstanceUse. 4 © , SteveSafren and SPSSv.21 software. 3 , KarenCropsey 5 , o improve 1 , - - - utable toHIV-risk concerns. were alsoaskedtoindicatethepercentoftheirdailylifestressattrib and againstcondomuse,interestintakingPrEP. Participants ual riskdataandreportedtheirperceivedHIVrisk,motivationsfor past 30days)MSM(38%Black,17%Latino)completeddetailedsex acceptability. 107highrisk( Methods: sex withmen(MSM)whocontractHIVfromtheirmainpartner. ed recentfindingsregardingthehighpercentageofmenwhohave with condomlesssexandmultiplepartners,butfewhaveincorporat- vating factor. HIVpreventionmessagesfocusonrisksassociated change focusontheimportanceofriskperceptionasacriticalmoti- Background: 1 Sarit Golub PrEP, TasP, andOtherPreventionStrategies Relationships andIntimacy:Implicationsfor Conclusions: Results: 1.06, 2.52,p<0.05),butnotforthosewithoutmainpartners. PrEP acceptabilityforthosewithmainpartners(aOR=1.63,95%CI: model, intimacymotivationsforcondomlesssexsignificantpredicted themselves atsignificantlylowerHIVrisk(p<.01).Inamultivariate than twiceaslikelytoreportlowHIVconcern(p<.001),andperceived was relationshipstatus,suchthatthosewithmainpartnersweremore of partners)didnotpredictHIVconcern;theonlysignificantpredictor their dailystress.Riskbehavior(i.e.,numberofUAsexactsor ever over40%reportedthatHIVconcernrepresentedlessthan5%of lives ofMSM. relationship andintimacyfactorsascriticalelementsinthesexual with thesepartners.Effectivepreventionprogramsmustrecognize may seekPrEPinanefforttomaximizetheirpotentialforintimacy MSM mayunderestimateHIVriskassociatedwithmainpartners,but ception onrelationshipfactorsratherthanobjectiveriskbehavior Hunter College,CUNY, NewYork, NY, USA 426 81% ofparticipantsreportedstressrelatedtoHIV-risk; how- Data weredrawnfromalargerNIH-fundedstudyofPrEP 1 (presenting) High-risk MSMappeartobasepersonalizedriskper Almost alltheoriesofhealthbehaviorand MSM isDrivenby HIV RiskPerceptionamong , KristiGamarel ³ 1 unprotectedanal(UA)sexactinthe 1 , InnaSaboshchuk 1 - - - .

109 POSTER ABSTRACTS Barriers to ART Adherence in Dealing with Diagnosis: 432 Minority Groups in the United 433 Incorporating Positive Coping States - A Systematic Review of Intervention Strategies into a Smartphone Application for Research HIV Care

Akesha Edwards1 (presenting), Silvia Rabionet1 Colleen Laurence1 (presenting), Rebecca Dillingham2, Erin Wispelwey2, Erin Plews-Ogan2, George Reynolds3, 1 Nova Southeastern University, Fort Lauderdale, FL, USA Karen Ingersoll2, Tabor Flickinger2

Background: For benefit to be derived from antiretroviral therapy 1 University of Virginia School of Medicine, Charlottesville, VA, USA (ART) there must be strict adherence to the medication. Interventions 2 University of Virginia, Charlottesville, VA, USA for ART adherence can range from behavioral, clinical, directly 3 Health Decision Technologies, Sheboygan, WI, USA administered or even self-administered. HIV-positive patients belonging to minority groups of Caribbean descent generally have a Introduction: The global, psychic, and familial distress that individuals ABSTRACTS greater issue of non-adherence to ART than those of non-minority may experience on learning they are HIV positive has repercussions groups. Efforts to evaluate the effect of interventions in this popula- not only for their HIV care, but for their health and well-being. tion are now increasing. Individuals may cope with this distress in positive and negatives ways. To address this common experience, we included several features that Methods: Electronic databases were searched during the period promote positive coping on our smartphone application (app) for rural 8/24/13 to 12/05/13. Combinations of search terms included patients. These features include a community message board, goal- “HIV/AIDS, ART, HAART, adherence, minorities, ethnicity and inter- setting feature, and the ability to track stress and mood. vention.” Inclusion criteria were English language, human subjects, any form of peer reviewed/scholarly articles published between Description: Audiotaped, semi-structured individual interviews were 1/1/2000 and 11/30/2013 conducted in the United States. conducted with 17 PLWH in rural southwest Virginia from June to

POSTER August 2013. Participants were asked about their process of coping Results: A total of 72 abstracts from articles were reviewed after ini- with and initiating HIV care after diagnosis as well as how a smart- tial article extraction. Of these 45 articles were excluded because of phone application might support individuals to link to and engage in inappropriate research question or no reporting of adherence results. care. These interviews were transcribed and coded, and key mes- In the remaining studies, interventions were either behavioral, clinical sages from them incorporated into app design. or peer based. Methods of reporting varied widely and there were many cases of incomplete data being presented. Means of measuring Lessons Learned: HIV-positive patients relied on diverse coping adherence were pill counts, visual scales, MEMS caps, refill rates, strategies that were historically situated and shaped by place and self-reporting. All interventions showed improved adherence rates existing sources of support on learning of their diagnosis. during the study but there was an inability to sustain this. Adherence Participants diagnosed in the pre- and early HAART era reported self-efficacy had a direct effect on adherence to ART in minority seeking out information and acceptance in support groups. patients. Recurring barriers to adherence in this population were Participants diagnosed more recently did not note engaging in sup- alcohol use, low income, low education literacy, depression, racial port groups and instead sought support from friends, family mem- discrimination, lack of trust in the medical system or providers, lack of bers, and providers. Participants noted that these support systems social support, being male, complicated regimens and older age. reminded them that they were not defined by their HIV diagnosis, which provided some emotional comfort. Speaking with people living Conclusions: Minority HIV/AIDS patients of Caribbean descent are with HIV also provided concrete information and a visual reminder highly vulnerable to the mal-effects of non-adherence. There is no that they can live successfully with HIV. best intervention to amend this problem, however, a combination of strategies tailored to the patient will give more reliable results. Recommendations: The impact of this app in enabling newly-diagnosed patients to engage in positive coping strategies will be assessed through a 3-week usability interview and a final interview as well as app metadata. Participants also complete the Brief COPE scale at baseline and endline.

110 9th International Conference on HIV Treatment and Prevention Adherence 434 Sarit Golub Mobile Dating‘Apps’inNewYork City in MenWhoHaveSexwithUsing Conclusions: Results: Methods: Background: 3 2 1 Viraj Patel tors mayfurtherenhancePrEPadoption. groups byprioritizingtheirspecificsubgroupconcernsandmotiva tion interventions.Tailoring interventionstohigh-riskMSMsub- motivators fortakingPrEPmayserveastargetsdissemina- vator itemsdidemergeinstratifiedsubgroupanalyses. age, andriskbehaviors,butdifferencesinspecificconcernmoti- exploratory analysis,overallresultsdidnotsignificantlyvarybyrace, sexual partner(77%),andincreasedpleasure(75%).In in talkingtoapartneraboutHIVrisk(79%),increasedintimacywith health (91%),decreasedanxietyaboutHIV(89%),increasedcomfort risk ofcontractingHIV(96%),increasedsensecontroloversexual All potentialmotivatorstousePrEPwerehighlyendorsed:decreased PrEP (44%),andhowtotalkapartneraboutbeingon(44%). from aconvenientdoctor(65%),howtotalkyourabout side effects(82%),notbeingeffectiveenough(73%),gettingPrEP Most highlyendorsedconcernsaboutusingPrEPwerecost(84%), an HIVtestinthepast6months,and75.3%werenotarelationship. 6 months,34.9%didnotusecondomsatlastintercourse,73.3%had Hispanic and21%Black.Themenhadamedianof5partnersinpast lege degreeorhigher, 54%madeunder$50,000/year;31%were who havesexwithmen(MSM)inNewYork City. assess concernsaboutandpotentialreasonsfortakingPrEPinmen ventions, weconductedasurveyonmobiledatingapplicationsto after dichotomizingLikertscaleresponses. PrEP (identifiedfrompriorresearch)andcalculatedfrequencies rate eachof10concernsaboutand6potentialmotivatorsforusing use, lastHIVtest,andnumberofpartners.Likertscaleswereusedto PrEP (n=693).Participantsself-reporteddemographics,condom sex inpast12months,livingNewYork City, andconsideringusing mobile datingapplicationstoHIV 9th InternationalConference on Hunter College,CUNY, NewYork, NY, USA Harlem UnitedCommunityAIDSCenter, NewYork, NY, USA Bronx, NY, USA Montefiore MedicalCenter;AlbertEinsteinCollegeofMedicine, Participants wereonaverage32yearsold,59%hadacol- Cross-sectional surveyadministeredon2MSM-specific 1 3 (presenting) To informpre-exposureprophylaxis(PrEP)uptakeinter- Addressing concernsaboutandhighlightingpotential Adopting Pre-ExposureProphylaxis Motivators andConcernsfor , SherryEstabrook HIV -uninfected menhavinghadanal T reatment andPrevention Adherence 2 , KimberleighSmith 2 , - 435 Patients would useitinaclinicalstudy the devicelong-termfordiseasemanagement,92%indicatedthey Acceptance: Althoughonly25%reportedtheywouldcontinueusing lation intodeviceandinsertingthestrawwasdifficult.System use. 50-89%reportedthat15minwastoolongfromingestiontoexha- easy: touse,hold,lift,readLCDoutput,unwrapstraw, rememberto ed. EaseofUsage:90-100%reportedthedevicetobeveryeasyor white, 5AfricanAmerican,1multiracial)aged48.6±7.1wererecruit- 1 take thisinformationintoaccount. ence theiracceptabilityofthesystem.Futuredesigniterationswill strengths/weaknesses ofthedesign,andidentifiedfactorsthatinflu positive patients.Participantsprovidedvaluablefeedbackonthe tions, andfollow-upelaboration. rated thedeviceonvariousdimensions,asetofopen-endedques- interview oftheirexperiencesusingmultiple-choicequestionsthat butanol taggedcapsules,and3)UsabilityAssessment:Participant ply ofdisposablestraws(insertedforexhalationintomGC)and2- home dailyusefor10dayswiththeSMARTsystem10-daysup investigator-monitored testtoensureability, 2)ResidentialUse:in- with writteninstructions,set-uppracticeattheirresidence,andan nents: 1)Training: educatedtousetheSMARTadherencesystem Gainesville AreaAIDSProjectandcompleted3sequentialcompo Background: 2 Donn Dennis Conclusions: Results: Methods: Richard Melker test acceptanceofthesystem. adherence. We conductedapilotstudyinHIV-positive patientsto uploaded HIPAA-compliant adherence datacanbeusedtoimprove food additive,2-butanol(metabolizedto2-butanone).Cellphone- butanone) inhumanbreathafteringestingHAARTdrugslinkedtothe graph (mGC)capableofmeasuringvolatilemetabolites(e.g.,2- ed anoveladherencesystemconsistingofmini-gaschromato- antiretroviral therapy(HAART)ischallenging.We designedandtest- University ofFlorida,Gainesville,FL,USA Xhale, Incorporated,Gainesville,FL,USA Enrollment: Afterinformedconsent,12adults(4F/8M;6 After IRBapproval,subjectswererecruitedfromthe 1 Measuring andmonitoringadherencetohighlyactive SMART isaviableadherencesystemforuseinHIV- Adherence SysteminHIV-Positive Usability EvaluationofSMART (presenting) 1 , LiseAbrams . 2 , Timothy Morey 2 , - - -

111 POSTER ABSTRACTS Structural Factors of Adherence to Drug Therapy Problems and 438 HAART in St. Petersburg, Russia 439 Adherence amongst HIV-Positive Patients Attending a Treatment Site in Andrey Belyakov1 (presenting), Olga Levina2, Vadim Rassokhin3, Robert Heimer2, Alena Suvorova4, Andrey Ustinov5 Middle Belt Nigeria

1 St. Petersburg City AIDS Center, St. Petersburg, Russian Rosaline Aderemi-Williams1 (presenting), Eric Akoji2, Patricia Ogbo1 Federation 2 Yale University, New Haven, CT, USA 1 University of Lagos, Lagos, Nigeria 3 St. Petersburg AIDS Center, St. Petersburg, Russian Federation 2 General Hospital, Ankpa, Nigeria 4 SPIIRAS (St. Petersburg Institute for Informatics and Automation of the Russian Academy of Sciences), St. Petersburg, Russian Background: With the expansion in the accessibility of antiretroviral Federation drugs (ARVs) in the management of people living with HIV and AIDS 5 Bekhterev Research Psychoneurological Institute, St. in resource limited environment, drug therapy problems (DTPs) can Petersburg, Russian Federation be a possible challenge. Hence the need to identify DTPs and the ABSTRACTS possible causes amongst HIV-positive patients on highly active anti- Background: In the Russian Federation, medical care for citizens liv- retroviral Therapy (HAART). ing with HIV is financed by the state and provided at regional AIDS Centers. Despite free highly active antiretroviral therapy (HAART), Methods: This was a 6-month prospective study using a form designed adherence is often low. St. Petersburg AIDS Center employed 2 mod- to extract the relevant data that met the study objectives. The data col- els of HIV health care delivery: in addition to a central treatment facil- lated were analyzed using descriptive statistical methods. ity, care in some districts of the city were organized through district infection disease departments (DIDD) beginning in 2012. DIDD pro- Results: Most (76%) of the patients affected by DTPs were females vide basic HIV care with, theoretically, easier access and less and aged more than 15 years old (73%). A total of 18 DTPs were iden- stigmatization. The associations between structure of health care tified. The major 5 were: unavailable prescribed medicines (30;

POSTER delivery and adherence to HAART have been insufficiently studied. 13.95%), undesirable effects (24; 11.16%), use of non-prescribed We conducted exploratory research to compare the 2 models of care sedatives (23; 10.70%), use of non-prescribed antimalarial drugs (21; delivery, focusing on patients’ outcomes. 9.77%) and use of herbal medicines (20; 9.30%). The major causes of the identified DTPs were: unnecessary drug therapy (59; 27.44%), Methods: We surveyed 347 HIV-positive patients receiving HAART in non-adherence (47; 21.86%), patient needed additional therapy (36; St. Petersburg (DIDD n = 143, AIDS Center n = 204). During face-to- 15.74%), and adverse drug reactions (30; 13.95%). face interviews participants answered questions about their socio- demographic characteristic, special aspects of medical care in the Conclusions: DTPs exists in this treatment center and the causes AIDS Center or DIDD, and self-reported adherence. We used 3 addi- include non-adherence. There is need for continual pharmacist mon- tional measures of adherence: doctors’ estimate in medical charts, itoring and intervention with other healthcare providers as part of the pharmacy refills, and virological parameters. In this exploratory pharmaceutical care plan when patients come for drug refill. study, we relaxed the significance level to p <0.1.

Results: Regression model with structural (time in lines, number of services used, time to get to the clinic) and psychosocial factors showed the influence on adherence (doctors’ estimate) of such factors as clinic (negative association for AIDS Center compared with DIDD, p = 0.02), time since HAART prescribed (positive association, p = 0.05), searching for additional information about HIV (positive association, p <0.1), and living separately from the partner (negative association, p <0.1).

Conclusions: Sampling at AIDS centers and DIDDs revealed differences in structural parameters and in adherence (doctors’ estimate). Choosing between AIDS Center and DIDD is an important predictor of adherence, but it can be associated with features of physicians’ method of evaluation of adherence, because differences in self- reported adherence were not detected.

112 9th International Conference on HIV Treatment and Prevention Adherence 442 Organization inLima,Peru Based SourcesataCommunity-Based Using Internet-Basedvs.Non-Internet- using othersources(friends,partners,orprintmaterial)(N=1,101). via theinternetoronlinesourcesofinformation(N=408)andthose to compareHIVprevalenceamongMSMwhofoundoutaboutEpicentro time betweenApril2012andOctober2013.Chi-squaredtestswereused ing servicesatEpicentroSalud(Epicentro)inLima,Peru,forthefirst Connie Haley testing messages. Internet-users athighriskofHIV between the2recruitmentgroups,toidentifywhatplacesMSM analyses willattempttoexplorewhysexualriskfactorsdidnotdiffer important firststepintheHIVcareandtreatmentprocess.Ongoing limited settings,thismaybeaneffectivestrategyfortackling fully identifyMSMathighriskofHIV. ForCBOsworkinginresource- care. Internet-basedpromotionandrecruitmentappearstosuccess Conclusions: unprotected sexatlastsexualencounter(43.9%vs.41.1%,p=0.642). did notexhibitstatisticallysignificanthighersexualriskbehavior, i.e., recruited throughothersources,MSMviaonlinesources ment sources(p=0.004).However Epicentro fromonlinesourcescomparedto17.2%forotherrecruit- Results: Methods: Background: 1 Colby Passaro to usetheirlimitedfundingmaximizetheimpactofservicedelivery. which recruitmentstrategiesreachthehighestriskpersonsinorder organizations (CBOs)workingtoreducethisdisparityunderstand lation (12.4%vs.0.4%in2011).Itiscriticalforcommunity-based men (MSM)inPeruis30timeshigherthanthegeneraladultpopu- 9th InternationalConference on Vanderbilt InstituteforGlobalHealth,Nashville,TN,USA HIV prevalencewas23.5%amongMSMwhofoundoutabout This isacross-sectionalstudyofadultMSMseekingHIVtest- 1 The prevalenceofHIVamongmenwhohavesexwith HIV diagnosisisthefirststepincontinuumof among MSMRecruitedforTesting Differences inHIVPrevalence 1 (presenting) , AaronKipp , andtobettertargetprevention HIV , whencomparedtothoseMSM T reatment andPrevention Adherence 1 , StenVermund 1 , - 445 Implications forFutureInterventions HIV Care:EarlyLessonsLearnedand ing surveysforsome. schedule andtransportationissues.Poorliteracyhinderedcomplet- Low enrollment(5/11eligible)occurredbecauseofself-reported reported highsatisfactionwiththecourse,materials,andduration. ized importanceofpostpartumretention.Onehundredpercent = 84(±48),MOS-SSS66(±39)].Allparticipatedactivelyandverbal- 6.6 (±1.3)].We alsomeasuredself-efficacyandsocialsupport[ASES symptoms [PHQ-9=5(±4.1)]andadequateHIVknowledge[BEHKA were employed.Baselinemeasuressuggestedmilddepressive Description: HIV+ pregnantwomenareunderstudied. ence poorpostpartumretention.Interventionsfocusedonretaining Northwestern University’s PerinatalHIVProgram(PHP),38%experi- Introduction: Ramona Bhatia Recommendations: Lessons Learned: ity ofcommonlyusedinstruments. existing appointmentsorbetterincentivized.Poorliteracylimitsutil barriers, suggestingthatgroupsmayneedtobescheduledwithin minority HIV+pregnantwomen.Enrollmentislimitedbystructural patients at discussions werepatient-driven.We enrolledEnglish-speakingPHP by leaders(physicians,nursepractitioners,trainedHIV+mothers),but received CP-typenotebookswithmaterials.Groupswerefacilitated ing viralload,anddescribingHIV retention inHIV+women(i.e.,describingthecarecontinuum,chart- cation, assessment,support)andmaterialsintoactivitiestoimprove and CP leaders. We adapted CPandWiLLOWprinciples(healthedu- met withPHPobstetricians,socialworkers,andnursepractitioners (CenteringPregnancy, CP)andHIV+(WiLLOW)women.Initially, we cohort retentionandpositivehealthbehaviorsinpregnant,non-HIV on establishedgroup-learningprotocolsthathavedemonstratedhigh 1 Sarah Sutton Mean agewas33y( HIV care(2visitsinthefirstpostpartumyear). were administeredatbaseline.Theprimaryoutcomeisretentionin 5-7 women.V Northwestern University, Chicago,IL,USA ³ 20 weeksofgestation.Theyattendedone2-hourgroup Our objectivewastodevelopagroupinterventionbased 1 alidated instruments(PHQ-9,ASES,MOS-SSS,BEHKA) Improve PostpartumRetentionin Developing aGroupInterventionto Retention inHIVcareimprovesoutcomes.In 1 (presenting) W Group interventionsappearacceptabletoethnic e held3monthlygroupswith5totalparticipants. ± 5.9), allwereBlack(80%immigrant),and40% , DonnaMcGregor -related socialsupport).W 1 , PhyllisChong omen 1 , -

113 POSTER ABSTRACTS Do Adherence Barriers Differ by Implementation of an Emergency 446 Gender? Results of Pooled Data 447 Department Opt-Out HIV Testing from MACH14 and Linkage to Care Program: Lessons Learned and Recommendations Carol Golin1 (presenting), Yan Wang2, Jie Shen2, Ira Wilson3, Judith Erlen4, Jane Simoni5, Marc Rosen6, Robert Remien7, Kelly Ross-Davis1 (presenting) Donna Evon8, Honghu Liu9 1 University of Alabama at Birmingham 1917 Clinic, Birmingham, 1 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA AL, USA 2 University of California, Los Angeles, CA, USA 3 Brown University, Providence, Rhode Island, USA Introduction: Due to the high rate of undiagnosed HIV infection, par- 4 University of Pittsburgh, Pittsburgh, PA, USA ticularly among disproportionately affected populations, a standard 5 University of Washington, Seattle, WA, USA of care, routine opt-out HIV testing strategy was implemented in our 6 Yale University School of Medicine, New Haven, CT, USA academically affiliated hospital emergency department (ED) in ABSTRACTS 7 HIV Center for Clinical and Behavioral Studies, New York, NY, USA accordance with CDC recommendations. Emphasis has been placed 8 University of North Carolina at Chapel Hill School of Medicine, on linkage to care (LTC) based on the necessity for enhancing health Chapel Hill, NC, USA outcomes among newly diagnosed persons. 9 University of California, Los Angeles, School of Dentistry, Los Angeles, CA, USA Description: Prior to implementing the opt-out HIV testing program in the ED, we conducted a pilot for a 72-hour period to assess the need, Background: Several studies, including analyses using data from the potential barriers, and individual acceptance of HIV testing in the ED MACH14 collaboration, have demonstrated lower adherence among setting. In August 2011, after considerable planning and engagement women than men. Few studies, however, have sought to determine of stakeholders, the ED implemented 24-hour, opt-out routine HIV test- whether factors influencing adherence differ for men and women. ing to all ambulatory patients ages 18-64. Concurrently, linkage coordi-

POSTER nators were integrated in the testing program to provide assistance for Methods: MACH14 has pooled data from 16 diverse longitudinal stud- HIV-diagnosed persons endeavoring to enter medical care. The link- ies with electronically measured (MEMS) adherence. We used age coordinator responds to each reported case within 1-2 days of pooled data from 10 MACH14 studies that collected, in addition to diagnosis. Preliminary contact includes counseling to support mental MEMS data, assessments of depression (cognitive, vegetative, and and emotional concerns, and assistance with navigating the health- overall), substance use, and alcohol abuse. 4,213 observations from care system, scheduling a linkage visit, and new patient orientation. 1,762 patients were available for longitudinal analysis. We defined adherence as the percent of prescribed doses taken. We fitted Lessons Learned: The national LTC goal is to ensure that 85% of Random Intercept (RI) models, adjusted for alcohol use, to check for patients attend an initial medical provider visit within 90 days of diag- interactions between gender and depression, and between gender nosis. Initial face-to-face contact with the linkage coordinator after and use of each of 3 substances (heroin, cocaine, other stimulants) new HIV diagnosis at the ED allows for a more active role in patient known to be associated with worse adherence. We then examined care and personal connection to follow-up resources. Improved the association between gender and adherence interaction over communication with ED personnel, collaboration with community time, controlling for race, age, employment, education, CD4 count, resources, ARTAS training, and an added linkage coordinator were and alcohol use. essential to serving a diverse and complex population seeking medical care but not anticipating an HIV diagnosis. Linkage to care in 90 Results: Thirty-four percent of the analytic sample participants were days increased from 45% in 2011-2012 to 71% in 2013. female; 49% African American, 27% Caucasian; 16% Latino, 10% other; 23% had less than a high school (HS) education, 64% had HS Recommendations: We successfully implemented a routine, opt-out degrees and only 13% higher than a HS degree; mean age was 41.6 ED-based HIV testing program coupled with linkage to care activi- (8.2); 23% were part-time or full-time employed; median CD4 count ties. Extensive planning, piloting and collaboration of a range of part- was 306 cells/mm3 (range 0-3029). Mean Adherence among women ners was essential for our program’s success in offering routine test- was 67% of prescribed doses vs. 73% for men. Multivariate mixed ing and optimizing linkage. Future strategies will include identifying effects models showed that only the interaction between gender and and empowering newly diagnosed patients ambivalent to HIV care cocaine use was significant, such that cocaine was more strongly and connecting to a broader range of community resources. associated with worse adherence among women than men, controlling for demographic factors.

Conclusions: Most factors associated with worse adherence that we assessed were the same for men and women, although cocaine use may be a more important barrier for women than men.

114 9th International Conference on HIV Treatment and Prevention Adherence 450 for ourongoing“ recruitment campaigntoreachSpanish-speakingLatinogaycouples to successfullydesignandimplementaneffectivesocialmedia Spanish-speaking LatinogaycoupleslivinginNewYork City. Inorder vention interventioninitiallycreatedforBlackgaycouples, target populationor“receiver attention toeachmarketing“channel,”targeted“messages,”and matrix (1985).Theguidedourevaluation,payingparticular ers andresearchteamusedMcGuire’s communication/persuasion the projectandallsocialmediacomponentsbuiltintoit. range ofcommunity-basedorganizationsinNewYork Citytopromote and Jack’d)helpedrecruitLatinogaycouples.We alsoreliedonawide Craigslist, andvarioussmartphoneapplicationssuchasGrindr, Scruff, ferent socialmediaplatforms(includingFacebook,theprojectwebsite, retention. Inaddition,thevarietyof“messages”usedacrossmanydif- process-from pilotprojectwritinganddevelopmenttorecruitment base thatbecameinvolvedearlyoninallaspectsoftheresearch social mediarecruitmentcampaignwastobuildastrongstakeholder pants throughdifferentchannels.Oneofthemajorsuccessesour 14 couples,N=28participants)andreachedmorethan3,000partici- ment. Asaresult,injust1month,werecruitedallofoursubjects(N= our researchstaffandstakeholderstoconductsocialmediarecruit- Brian Dodge Intervention Study Latino GayCouplesforanHIVPrevention Conclusions: Results: Methods: lations engagewithsocialmediaplatforms. among primarilySpanish-speakingpopulationsandhowthesepopu- However, littleisknownabouttheuseofonlinesocialnetworking recognized asaplatformforhealthcommunicationandeducation. promote healthwellness.Inparticular could beusedtoreach,retain,andmaintainindividualsincare (MSM), makingitimportanttounderstandhowthesetechnologies among African-AmericanandLatinomenwhohavesexwith Background: 8 7 6 5 4 3 2 1 Omar Martinez Alex Carballo-Dieguez Juan DelaCruz as itisoftencharacterizedwithinpublic healthliterature. MSM showsthatthispopulationisnot particularly“hard-to-reach,” social mediarecruitmentstrategy with Spanish-speakingLatino participants inHIVinterventions.Particularly, thesuccessofour communication technologiesintorecruitment andengagementof 9th InternationalConference on Community Activist,NewYork, NY, USA Lehman College,NewY University ofTexas, ElPaso,TX,USA Latino CommissiononAIDS,NewYork, NY, USA AID forAIDSInternational,NewYork, NY, USA Bloomington, IN,USA Indiana UniversityBloomingtonSchoolofPublicHealth, Columbia UniversitySchoolofSocialWork, NewYork, NY, USA NY, USA State PsychiatricInstituteandColumbiaUniversity, NewYork, HIV CenterforClinicalandBehavioralStudiesattheNewYork We developedasocialmediarecruitmentprotocolandtrained Our projectaimedtoadapt 3 , ZachShultz Our findingshighlighttheimportance ofincorporating Online socialnetworkingusehasincreasedrapidly Using SocialMediatoRecruit Still aHard-to-ReachPopulation? 7 1 , AntonioPorras (presenting) Latinos enPareja 1 , SilviaChavez-Baray 1 ork, NY , JonathanCapote , ElwinWu .” 8 HIV , USA ” study, ourcommunitystakehold- T “Connect n’Unite,” reatment andPrevention Adherence , theInternetisincreasingly 2 , TheoSandfort 6 , EvaMoya 4 , HugoOvejero 1 6 , an HIVpre , 5 , - 452 Methods: Background: 3 2 Sharada Wasti People LivingwithHIVinNepal Conclusions: Results: 3.18, p=0.009),traveltimetohospital>1hour(OR2.84,0.035). effects (OR=6.04,p0.025),ARTstarted<24monthsduration female (OR=6.91,p=0.001),beingilliterate(OR4.58,0.015),side- (OR =17.99,p0.014);alcoholuse12.89,<0.001),being factors influencingadherencewere:non-disclosureofHIVstatus ence, i.e.nomisseddosesinthe4-weekspriortointerview promote adherenceinNepal. ART whilepolicy-makersshoulddevelopappropriatesocialpolicyto ers shouldaddresssomeofthepracticalandculturalissuesaround regimens tailoredtoindividualpatients’lifestyles.Healthcarework- ment; accessibletreatment;clearinstructionaboutregimens;and tors associatedwithadherencetoART. ing. Amultivariatelogisticregressionmodelwasusedtoidentifyfac- in early2010.Adherencewasassessedthroughsurveyself-report- understand thefactorsinfluencingonadherencetoARTinNepal. ART isessentialforHIVinfectionmanagement.Thisstudyaimedto patients treatedforHIV/AIDSandmaintainingoptimaladherenceto 1 Bournemouth University, Bournemouth,UnitedKingdom University ofSheffield,UnitedKingdom FHI 360Nepal,Kathmandu,Nepal A totalof282(85.5%)respondentsreportedcompleteadher- A quantitativestudysurveying330ART-prescribed patients Antiretroviral therapy(ART)isalifesaverforindividual Improving adherencerequiresasupportiveenviron to AntiretroviralTreatment for Factors InfluencingonAdherence 1 (presenting) , PadamSimkhada 2 , EdwinTeijlingen . Major 3 -

115 POSTER ABSTRACTS A Qualitative Study of Maximal HIV Adherence/ 453 Antiretroviral Treatment 454 Engagement Given Minimal Mental Adherence amongst Adolescents and Young Health Resources: Lessons Learned from China Adults in Soweto, South Africa Jane Simoni1 (presenting), Wei-ti Chen2, Cheenghee Shiu1, Hongxin Zhao3, Hongzhou Lu4 Fatima Laher1, Stefanie Hornschuh1, Matamela Makongoza1 (presenting), Janan Dietrich1 1 University of Washington, Seattle, WA, USA 2 Yale University, New Haven, CT, USA 1 Perinatal HIV Research Unit, Soweto, Gauteng, South Africa 3 Ditan Hospital, Beijing, China 4 SPHCC & Fudan, Beijing, China Background: HIV management of adolescents and young adults (AYAs) is pertinent to sub-Saharan Africa, which has the highest Introduction: PLWHA experience disproportionate levels of psychi- pediatric HIV incidence. Antiretroviral treatment (ART) adherence is atric illness and distress, with 1 meta-analysis indicating major a challenge for AYAs. This qualitative study explored factors of ART ABSTRACTS depression was twice as prevalent in PLWHA as the rest of the pop- adherence among AYAs attending treatment at the Perinatal HIV ulation. Psychiatric distress causes serious barriers to optimal HIV Research Unit, Soweto, South Africa. medication adherence and engagement in ongoing care. Given the dearth of specially trained mental health professionals in low- and Methods: Four focus group discussions were conducted with HIV-pos- middle-income countries (e.g., in China there are just 17,000 licensed itive 15 to 25-year old ART-recipients who provided consent/assent. psychiatrists, yielding a per capita rate of approximately 1 psychia- Transcripts were coded using thematic analysis. trist for every 80,000 people), alternative means of providing comprehensive care to optimize clinical outcomes are urgently needed. Results: Participants (n = 18) were aged median 18.5 years, 72% female and 67% virally suppressed <400copies/mL. Three main Description: We will overview our experience during the last decade themes emerged: (i) correct knowledge and perceptions about adher- working in urban areas in China. Multiple NIH-funded surveys and

POSTER ence, (ii) social, personal and medication-related reasons for subop- randomized trials will be described, citing the need for assistance timal adherence and (iii) reminder, concealment and motivational and the potential relevance of cognitive behavioral strategies. strategies to optimize adherence. (i) Most participants stated that Preliminary (unpublished) data collected on pilot projects will be counsellors and doctors taught them why and how medication should highlighted. Specific projects include an R34-funded RCT of nurse- be taken. Strict dosing time and interval observance was emphasized. facilitated adherence counseling in Beijing, an R21-funded parental Adherence was perceived with health, a healthy lifestyle and longevi- HIV disclosure intervention in Shanghai, and a K24-funded project on ty. Many described viral resistance as the consequence of non- computer-based counseling. adherence. One participant experienced non-adherence resulting hospitalization. (ii) Nearly all participants denied desire for or actual Lessons Learned: Two themes based on our findings and reflected in non-adherence but easily speculated on perceived adherence barri- our current research program aims are (1) the need for task ers, including personal (forgetting while being busy, not understand- shifting/sharing away from highly trained mental health personnel to ing the rationale of ART, self-pity and denial), social (being embar- more readily available and receptive personnel such as nurses and (2) rassed to take medication publicly) and medication-related (tedium of the potential for digital technology in the form of computer-based based daily dosing, large pill size, unpleasant smell and taste. (iii) Alarms, programs to supplement mental health services in primary HIV care. family members and friends were mentioned as reminder methods. Many family members and friends were aware of participants HIV Recommendations: In conjunction with public health officials and pri- status. Strategies to conceal medication were wrapping pills in paper, mary HIV care providers, forging alliances with nurses, social workers, using excuses at dosing times like buying cell phone airtime, do as well as paraprofessionals is crucial to expanding access to mental homework or use the bathroom. Some participants developed motiva- health care for PLWHA. Developing credentialed training programs - in tional messages for adherence, namely the desire to grow up and person or computer-based - constitutes an important first step. have a future. Children were also their motivation.

Conclusions: HIV-positive AYAs in our study understood and aimed for ART adherence motivated by the desire for longevity.

116 9th International Conference on HIV Treatment and Prevention Adherence 455 Inject Drugs adherence aswellretentionincare. vention withpositives,includingARTinitiation,continuation,and study helpedcreatePWID-specificservices.PCMcansupportpre HIV-positive PWIDtoARTatparticipatingHIVclinics,wherethe form ofgoods,whenadherentforatleast6months. clinic studylogstohaveinitiatedART PCM foreveryeligibleparticipantwhoisconfirmedbyconfidential needed referrals.Asmallconditionalcashtransferisgiventothe appropriate treatment,helpaddressbarriers,andfacilitateany initiate ART. PCMfollow-upwitheligibleparticipantstoensure study participantstoconnectastudy-participatingHIVclinicand positive andonART. PCMsupportHIV-positive, clinicallyeligible trust throughworkingwithotheractivePWID.SomePCMareHIV- selected fromapoolofrecoveredPWIDwhoshowedleadershipand Kenya, toHIVcareandretainthem.Peercasemanagers(PCM)were 1 Charles Cleland Recommendations: ties andnotimeforART. were incarcerated,and2died.Mostwhostoppedhadotherpriori ed ART, 9successfullycontinuedonART, 6stoppedtakingART, 2 gible toinitiateARTandwereassignedaPCM.Ofthose,19initiat PWID. Atthisfirstinterventionsite,21participantswereclinicallyeli care significantlyeasierandhavereduceddiscriminationtowards the HIV-positive PWID,andtheHIVclinicstaffhavemadelinkingto ART Lessons Learned: Description: Introduction: 2 John Lizcano (CD4 <350cells/ infected (CD4<350cells/ who injectdrugs(PWID)supportspecificallyinsub-SaharanAfrica. low-income settings,thoughithasnotbeenwidelyappliedtopeople 9th InternationalConference on Ministry ofHealthKenya,Nairobi,Kenya New York University, NewYork, NY, USA , hasbeenveryhelpful.TherelationshipsbuiltamongthePCM, 1 We usedpeersupporttolinkclinicallyeligibleHIV- Peer supportiseffectiveforHIVtreatmentaccessin Retain inHIVCarePeopleWho Use ofPeerSupporttoLinkand (presenting) ␮ 1 , PeterCherutich L) PWIDinNairobiandCoastalMombasa,Kenyato Use ofPCMtolinkclinicallyeligibleHIV-infected PCM canfacilitatelinkageofclinicallyeligible ␮ L) PWIDinNairobiandCoastalMombasa, , EvaMuluve HIV 2 T reatment andPrevention Adherence , andtoeveryparticipant,in 2 , AnnKurth 1 , - - - - in CaretheSwissCohort 1 Background: 4 3 2 JJ Parienti Marie PauleSchneider Conclusions: controls, p=0.63). RNA<50 cp/mLintheinterventiongroup(vs.179/227(78.8%)among Lost6 outcome.Attheendoffollow-up,160/198(80.8%)hadaHIV after adjustingforage,sex,raceandCD4cellsatbaselinethe group (p<0.001).Thisdifferenceremainedstatisticallysignificant group 72/200(37.5%)wereLost6vs.120/230(62.5%)inthecontrol Lost12 (7/64(10.9%)vs.19/352(5.4%),p=0.10).Intheintervention trol group(p=0.37).HCVco-infectedpatientsweremorelikelytobe vention group9/200(4.5%)wereLost12vs.15/230(6.5%)inthecon- age, 43(11%);African33.3%;baselineCD4cells291(198).Intheinter Results: Methods: enhancing programisassociatedwithabetterretentionincare. predicts survival.Ourobjectivewastoinvestigateifanadherence- antiretroviral toxicity gaps inscheduledvisitsandpotentiallypatients’safetyregarding an antiretroviraladherenceprogramhavethepotentialtolimitLost6 Lost12 orHIVRNA.Inthiscontext,effortsandresourcesinvestedin Delphine Maestralli visit, respectively. (Lost12) and>6-month(Lost6)intervalwithoutanyregisteredSHCS primary andsecondaryendpointweredefinedasa>12-month missed appointmentwerecalledback(2phonecalls,1letter).The action withpharmaceuticalandmedicalteams.Subjectswho combining motivationalinterview, electronicpillcontainer, andinter- adherence enhancingprogramwasaninterdisciplinary attended theadherenceenhancingprogramfor>6months.The infected patientsreceivingantiretroviraltherapy),200subjects ter oftheSwissHIVCohortStudy(SHCS).Withinthiscohort(430HIV- Normandy, France Centre HospitalierUniversitairedeCaen, Lausanne, Switzerland Infectious DiseaseService,CHUV, UniversityofLausanne, Lausanne, Switzerland Lausanne, Switzerland Care &CommunityMedicine,UniversityofLausanne, Community Pharmacy, DepartmentofAmbulatory 457 The mean(SD)follow-upwas3.9years(0.12);male63.4%; W 4 e conductedaretrospectivecohortstudyinsinglecen Retention inHIVcarepredictssubsequentretentionand In theSHCS,ourinterventionwasnotassociatedwith 2 , OlivierBugnon Program andRetention Adherence-Enhancing . 1 (presenting) 1 , MatthiasCavassini , AurelieRotzinger 1 , 3 , - -

117 POSTER ABSTRACTS Explaining Pre-Exposure Association between 460 Prophylaxis Efficacy: A Qualitative 461 Engagement in Substance Study of Message Comprehension and Abuse and Mental Health Services among Messaging Preferences among Men Who Recently Incarcerated HIV-Positive Have Sex with Men and Male Sex Workers Individuals and Medication Adherence/Viral in the United States Load: Early Findings from imPACT (Individuals Motivating to Participate in Kristen Underhill1 (presenting), Kathleen Morrow2, Don Operario3, Kenneth Mayer4 Adherence/Care/Tx)

1 Yale University, New Haven, CT, USA Carol Golin1 (presenting), Jessica Carda-Auten2, Jennifer Groves3, 2 Miriam Hospital/Brown Medical School, Providence, RI, USA Michelle Gould4, Kevin Knight4, Patrick Flynn4, David Wohl2 3 Brown University, Providence, RI, USA ABSTRACTS 4 Harvard, Boston, MA, USA 1 University of North Carolina at Chapel Hill, NC, USA 2 University of North Carolina at Chapel Hill, School of Medicine, Background: Oral pre-exposure prophylaxis (PrEP) can be an Chapel Hill, NC, USA acceptable HIV prevention strategy for US men who have sex with 3 University of North Carolina, Sheps Center for Health Services men (MSM) and male sex workers (MSWs). Clinicians prescribing Research, Chapel Hill, NC, USA PrEP must educate users about drug efficacy, but little is known 4 Texas Christian University, Dallas, TX, USA about how men may understand and apply efficacy information. We used focus groups and individual interviews to explore message Background: HIV-infected inmates re-entering communities face comprehension and message framing preferences for communicat- many challenges to antiretroviral adherence, often compounded by ing about PrEP efficacy with MSM and MSWs. mental health (MH) and/or substance abuse (SA) problems. We

POSTER sought to assess, among recently released HIV-infected prisoners, Methods: We conducted 8 focus groups (n = 38) and 54 individual the prevalence of SA/MH problems and SA/MH services utilization interviews with MSM and MSWs in Providence, RI; 47 of the 92 par- (SU), and the association of MH/SA SU and adherence and viral load. ticipants were MSWs. Participants were English-speaking male adults who self-reported negative/unknown HIV status and recent Methods: The impact (Individuals Motivating to Participate in unprotected anal sex with a potentially serodiscordant male partner. Adherence/Care/Tx) trial evaluates a new model for improving HIV- Participants received basic information about PrEP. Groups dis- infected prison releasees’ adherence to HIV care. 154 participants cussed comprehension, credibility, and acceptability of efficacy have completed week 14 of the study to date. We collected survey messages, including success-based framing, failure-based framing, data at 2, 6, 14 weeks post-release to assess, among other topics, comparisons with efficacy of other prevention methods, and para- MH (Kessler’s Psychological Distress Scale (K10) at weeks 2), SA phrasing efficacy (e.g., “PrEP is usually/highly effective”). (modified AUDIT and standardized drug use scales), service utilization (SU) (time-line follow-back), self-reported antiretroviral adher- Results: Findings indicated a range of comprehension and operational ence (visual analogue scale), and blood for viral load(VL). We understandings of efficacy messages. Across MSM and MSWs, men dichotomized the drug use scores to indicate presence or absence of tended to prefer percentage-based messages, reporting that para- SA problems corresponding to DSM drug dependence diagnosis or phrased messages may lack credibility or reduce willingness to use injection drug use. We dichotomized MH scores as with or without PrEP. Numerical translations of paraphrased information varied, with “high or very high” psychological distress. We compared partici- men guessing efficacy estimates of 50-90% for a “usually effective” pants with MH/SA needs with and without SU regarding their 14 drug, 90% or above for a “highly effective” drug, and failure rates of week adherence and VL. 10-60% for a drug that “sometimes fails.” Preferences varied for communicating about efficacy using a single percentage vs. a percentage Results: Among the 154 participants completing week 14 of imPACT range. Men reported uncertainty about how to interpret numerical [mean age 42; 21% Female; 69% African American, 23% Caucasian, estimates, and concern about whether estimates would predict per- 8% other; 5% Latino; 42% with < HS education], 25% reported an SA sonal effectiveness. Men also preferred success-based (or dual suc- problem, 18% reported “high/very high” psychological distress, and cess- and failure-based) messages to failure-based messaging alone, 38% report either a SA problem and/or “high/very high” psychologi- particularly when doctors were messengers. cal distress (SA/MH). Of 58 participants reporting SA/MH problem, 26% reported SU. Mean adherence and VL at 24 weeks for those with Conclusions: Providers involved in PrEP education, outreach, and and without SU, respectively, were (92% vs. 83%; p = 0.15; and 8,663 prescription may face challenges in communicating with potential vs. 21,640 copies/mL; p = 0.37). users about drug efficacy. Findings suggest that efforts to educate MSM and MSWs about PrEP should incorporate percentage-based Conclusions: HIV infected former prisoners with SA and/or MH information, but that message framing may influence credibility and needs who access SA/MH services may trend toward higher adher- overall PrEP acceptability. ence compared with those who do not access such services.

118 9th International Conference on HIV Treatment and Prevention Adherence 462 ence amongUSwomenwhousePrEP strategies capableofmeetingwomen’s needsandpromotingadher- gation. SuchresearchisessentialtothedesignofPrEPintroduction informal surveyprocesshighlightedissuesthatneedfurtherinvesti- qualitative researchspecificallyexploringwomen’ ple, well-designeddemonstrationprojectsenrollingwomenand inform PrEPintroductionamongwomenisneeded,includingmulti- provider andcommunityinvolvementtobesuccessful.Research tool mustincludebothextensiveprovidereducationandactive United Statesdemonstratesthatintroductionofanewprevention pragmatic issuesthatPrEPmightraiseintheirownlives. their levelsofinterestandawareness,thoughtsaboutadherence vey highlightedcommonthemesamongwomenatrisk-including and/or usersofillegaldrugs.Whilenotscientificallyrigorous,thissur- women in6UScitieswhoself-identifiedastransgender Group conductedCommunityDialoguesandinformallysurveyed85 and Women UsingIllegalDrugs among Transgender Women, SexWorkers, Concerns aboutPre-ExposureProphylaxis Effectively? ExploringAttitudesand der womenand/orrecreationaldrugswerealsoraised. drug interactionswithcontraceptives,hormonestakenbytransgen- tomers intentonavoidingcondomuse.Questionsaboutpotential vided, andtheriskofbeingforcedtotakePrEPbypartnersorcus areas includedtheabilitytobeadherent,levelofprotectionpro about itfocusedprimarilyoncost,access,andsideeffects.Other option. Afterlearningaboutit,65%saidtheymighttryPrEP. Concerns highest riskofHIV attention tothoseaffectingitsaccessibilityandutilitywomenat (PrEP) introductionandeffectiveusedomestically, withparticular tion addressingissuesthatshapepre-exposureprophylaxis’ US Women andPrEPWorking Groupisanational advocacycoali- results intermsofwomen’s adherencetodailyPrEPregimens.The Recommendations: Lessons Learned: Description: Introduction: 2 1 Anna Forbes 9th InternationalConference on Sister Love,Atlanta,GA,USA US Women andPrEPWorking Group,Kensington,MD,USA 1 In theabsenceofformalresearchonthis,Working , Pre-Exposure Prophylaxis Will High-RiskWomen Use Clinical trialstodatehaveproducedwidelyvarying Dazon DixonDiallo Most wereunawareofPrEPasanHIVprevention . The historyoffemalecondomuptakeinthe HIV 2 (presenting) T reatment andPrevention Adherence . s needs.This , sexworkers, - - 463 James Griffin Sex withMen in Young BlackHIV-Positive MenwhoHave tion adherenceratesinHIV+BMSM. nuanced understandingofthemechanismsunderlyinglowmedica- tus) andhighinfectionrates,thereisanurgentneedforamore With multiplestigmatizedidentities(race,sexualorientation,HIVsta- lenge forHIV-positive Blackmenwhohavesexwith(BMSM). Background: 3 2 1 Matthew Riccio Conclusions: Results: Methods: behind atoothbrushcontainer bottles behindaregularlyusedcoffeemuginthekitchencabinet, breath-mint containersfordiscretecarrying,andkeepingmedication concealing medicationintake,includingstoringin ticipants alsonotedcreativestrategiestoensureadherencewhile provider rapportincreasedtheriskfornon-adherence.Severalpar- tionship withtheirproviderfacilitatedadherence,whilepoor where mendidnotdiscloseHIVstatussocially, havingagoodrela- with aproviderasreason(s)forhigheradherencerates.Incases ence. Allparticipantsmentionedsocialsupportand/orstrongrapport the needtotakemedicationinprivateasapotentialbarrieradher- family, friends,andcolleagues,while93%ofparticipantsmentioned issues relatedtoadherence. and coded.Datawereanalyzedwithinacrosscasestoexplore and adherencestrategies.Interviewswererecorded,transcribed, adherence behaviors,medicationstoragemethods,supportsources, New Y highly activeantiretroviraltherapy(HAART)medicationin ducted withfifteenHIV-positive BMSMaged18-30yearsoldon medication complianceamongHIV-positive BMSM. first partofamulti-phaseinterventionstudyaimedatincreasing for medicationadherence.Ultimately oped creativestrategiesforprivatemedicationtakingasfacilitators noted social/familialsupport,strongproviderrapport,andself-devel- ments, amplifyingtheriskfornon-adherence.Successfuladherers medication concealment,bothathomeandinpublic/socialenviron Fordham University, Bronx,NY, USA New York, NY, USA Columbia UniversityMailmanSchoolofPublicHealth, Columbia University, NewYork, NY, USA ork City Less than10%ofparticipantsreportedfull-disclosureto In-depth, semi-structuredqualitativeinterviewswerecon- 3 Medication adherencepresentsaconsiderablechal- Taking andAdherenceStrategies Concealment ofMedication Participants hadlowHIVdisclosureratesandhigh , EmilyCherenack . Participantswereaskedquestionsregardingdaily 1 , Gertraud Stadler , andinabedpillow 1 , EllenHada 2 (presenting) , thisresearchrepresentsthe 1 , PatrickWilson , EfeKakpovbia . 1 1 , -

119 POSTER ABSTRACTS Peer Patient Navigation within a Factors Associated with Treatment 464 HIV Primary Care Setting: 465 Utilization, Medication Adherence, Outcomes and Implementation and Communication with Providers in Newly HIV-Diagnosed Men Who Have Sex with Susan Ryerson Espino1, Allison Precht2, Marisol Gonzalez1 (presenting) Men in New York City

1 The CORE Center, Chicago, IL, USA Jill Pace1, Patrick Wilson1 (presenting), Nathan Hansen2, 2 The Ruth M. Rothstein CORE Center, Chicago, IL, USA Arlene Kochman1, Kathleen Sikkema3

Background: Research examining the efficacy, design and imple- 1 Columbia University Mailman School of Public Health, mentation of peer-based HIV linkage and retention components is New York, NY, USA growing. Reviews suggest that peer-based programs across a vari- 2 University of Georgia, Athens, GA, USA ety of substantive fields can have significant effect on attitudes, 3 Duke University, Durham, NC, USA

ABSTRACTS knowledge and intentions but have been found to have mixed or limited impact on health-related behavior and clinical outcomes Background: In the USA, approximately 50% of known HIV-positive (Medley et al 2009, Tolli 2009; Simoni et al 2011; Webel et al 2010). individuals are not engaged in HIV care. Approximately 55% are ade- Further, additional studies are needed that examine the nuances of quately adherent to their ART regimen. A study of men who have sex intervention development and implementation. This paper con- with men (MSM) found 12% of participants not currently engaged in tributes to this growing literature by focusing on both clinical out- any care. The primary aim of this analysis is to examine the factors comes and factors that facilitated and challenged the development associated with treatment utilization, medication adherence, and of a HIV peer-based intervention fostering linkage and retention in a communication with providers in newly HIV-diagnosed MSM. HIV primary care setting. Methods: Data comes from the 6-month assessment of a randomized Methods: Over 300 women of color were enrolled in a HRSA SPNS controlled trial of newly-diagnosed HIV-positive MSM in New York POSTER initiative for at least 12 months aimed at increasing linkage and City from 2009-2013. Treatment utilization was measured by the retention through Peer Patient Navigation. To be eligible for the inter- aggregate number of medical appointments participants attended in vention women were either lost to care, sporadically engaged or the prior 3 months, including appointments for medical care, sexual newly diagnosed at baseline. The local evaluation included survey health, mental health, and substance abuse. Adherence was meas- interviews at baseline, 6 and 12 months. ured by combined number of ART doses missed within a 7-day period. Patient-provider communication was measured using the Patient Results: Overall, 75% were retained in HIV primary care and 84% of Self-Advocacy Scale. Stepwise multiple regression was used to newly diagnosed women were linked to care within 60 days. model the factors associated with treatment utilization, medication Baseline hopelessness and medication self-efficacy were found to adherence, and patient-provider communication. relate to baseline medication barriers. Women reporting more hopelessness and less medication efficacy reported more barriers at Results: The sample size for this analysis was 124. Total number of baseline. After 12 months of program exposure, local evaluation par- medical appointments attended in the prior 3 months was significant- ticipants reported significantly lower hopelessness, increased med- ly associated with active coping mechanisms (p = .002) and negative- ication efficacy, and fewer medication barriers when compared to ly correlated with alcohol and drug abuse (p = .009 and p = .019, baseline reports. Further, reports of fewer medication barriers at 12 respectively). Adherence to ART was negatively correlated with months was found to be related to 12-month viral load suppression. alcohol use (p = .011) and drug abuse (p = .002). Patient-provider communication was significantly associated with sensation-seeking Conclusions: We will discuss outcomes and factors impacting the scores and the treatment-related reduced HIV concerns scale development and implementation of Peer Patient Navigation within (p = .003 and p = .004, respectively). an HIV primary care setting and how such navigation impacts patient outcomes. Conclusions: Participants who took a more active coping strategy, were less sensation-seeking, and whose HIV concerns were reduced by treatment advances were more likely to attend more medical appointments and to have better patient-provider communication. Participants who experienced more substance use and abuse were less engaged in care and more likely to be non-adherent to medication. Engagement in care and adherence to medication is critical for HIV care and treatment, as well for the prevention of HIV transmission. Understanding the predicators of these treatment outcomes can help to better target and personalize HIV care and treatment, and increase engagement in care, adherence to medication, and reduce HIV transmission.

120 9th International Conference on HIV Treatment and Prevention Adherence 468 Africa: AQualitativeStudy Serodiscordant CouplesinSub-Saharan Prophylaxis forHIVPreventionamong support inmultiplesocialdomains. Relevant contentwasgroupedintocategoriesdescribingadherence content yieldinginsightintosocialresourcesforadherencesupport. for HIVpreventioninEastAfrica.Interviewdatawereexamined multi-site, double-blind,randomized,phase-IIIefficacytrialofPrEP were takingPrEPorplaceboaspartofthePartnersStudy, a Ugandan serodiscordantcouplesconstitutethedata.Interviewees Connie Celum ly usethem. health caresettingsandsystemsthatcoupleswillreadi- resources forsupportofPrEPadherenceareplentifuloutsideformal bers, andthelargercommunity. Thequalitativeresultssuggestsocial infected spousesthemselves,andbychildren,extendedfamilymem This analysischaracterizessupportforadherenceprovidedbyHIV- adherence amongHIV-uninfected individualstakingPrEPorplacebo. serve therelationshipwithaspouseasanimportantsocialmotivefor 1 Conclusions: Results: Methods: Background: 4 3 2 Norma Ware delivery strategiesandprogramplanningefforts. incorporation oflaysocialresourcesforadherencesupportinto experience, implementerscanmovetomeetthischallengethrough es tochallengedeliveryofPrEP. Buildingonthetreatment scale-up ed scale-upofantiretroviraltreatmentforHIV/AIDSinAfricapromis- tion scale-up. resources intoadherencesupportinitiativesaspartofimplementa- couples inUganda,layingthegroundworkforincorporationofthese resources forsupportofadherencetoPrEPbyHIVserodiscordant Using qualitativedata,thispresentationdescribeslaysocial adjunct toprofessionallytrainedpersonnel,whoareinshortsupply. nity memberstoantiretroviraltreatmentadherenceareanimportant contributions offamily, peers,layhealthworkersandothercommu- is (PrEP)forHIVpreventionwillrequirehighadherence.InAfrica, Monique Wyatt 9th InternationalConference on University ofWashington, Seattle,WA, USA Kabwohe ClinicalResearchCentre,Kabwohe,Uganda Massachusetts GeneralHospital,Boston,MA,USA Harvard MedicalSchool,MA,Boston,USA Previous analysisofthesedataidentifiedthedesiretopre- In-depth interviewswith88HIV-uninfected partnersin 1 Effectiveness ofantiretroviralpre-exposureprophylax- The sameshortageofhealthprofessionalsthatimped- , 4 of AdherencetoAntiretroviral Lay SocialResourcesforSupport , DavidBangsberg Emily Pisarski 1 , EliodaTumwesigye 1 (presenting) HIV 2 3 T , JaredBaeten reatment andPrevention Adherence , JessicaHaberer 4 , 2 , - Salvador, ElSalvador among HIV-Positive SubstanceUsersinSan adhering toART. using PLHfaceinenteringandstayingmedicaltreatment, Salvador providers and29crackcocaine-alcohol-usingPLHinSan Methods: Background: Andrew Petroll Conclusions: Results: 1 among drugusers. into medicalcareandpoormedicationadherence,particularly continues tobehighduelatediagnosisofHIVinfection,entry (ART) topersonslivingwithHIV. However, AIDS-relatedmortality introduced andexpandedtheavailabilityofantiretroviraltherapy to helpsubstanceusingPLHadhereARTareurgentlyneeded. ommendations forHIVtreatmentwithsubstanceusers.Interventions reduce stigmatizationofdrugusersandeducateprovidersaboutrec abandonment. InterventionswithHIVmedicalprovidersareneededto from initiatingoradheringtoARTandmaycontributetreatment effects ofmixingARTwithdrugsandalcoholdiscouragedrugusers given fewstrategiesorresourcestoquitusingdrugs. following anegativeinteractionwiththeirprovider. Participantswere PLH reportedavoidingdiscussingtheirsubstanceusewithproviders oped mechanismstomaintainhighlevelsofadherenceART Others ignoredproviders’warningsregardinginteractionsanddevel- ing orabandoningARTwhentheyrelapsedtodrugandalcoholuse. Because ofmedicalproviders’warnings,mostPLHreportedsuspend result, manyexperienceddepressionandincreasedtheirdruguse. their diagnosisandbelievedtheyhadonlyafewmonthstolive.As result inearlierdeath.MostPLHreportednoknowledgeofARTbefore hol beforebeginningART than halfoftheproviderscounseledPLHtoabstainfromdrugsandalco- quent barrierstoretentionincareandadherencemedication.More Medical CollegeofWisconsin,Milwaukee,WI,USA 469 Providers identifieddrugandalcoholabuseasthemostfre , ElSalvador We conductedin-depthinterviewswith13HIVmedical Since themid-1990s,manydevelopingcountrieshave Messages frommedicalprovidersabouttheharmful 1 (presenting) Care andARTAdherence Barriers toEntryintoMedical , toexplorethebarriersandfacilitatorsdrug , warningthemthatcombiningthe2would , JuliaDicksonGomez 1 . Some - - -

121 POSTER ABSTRACTS Healthy Divas: A Culturally Preliminary Qualitative 470 Relevant Intervention to Improve 474 Evaluation of an Intensive Engagement in Care among Transgender Case Management Intervention to Increase Women Living with HIV Linkage and Retention in Care: Successes in Social Support and Housing Jae Sevelius1 (presenting), Mallory Johnson1, Parya Saberi1 Michelle Broaddus1 (presenting) 1 University of California, San Francisco, CA, USA 1 Center for AIDS Intervention Research, Medical College of Introduction: Transgender women living with HIV (TWH) experience Wisconsin, Milwaukee, WI, USA disproportionate rates of AIDS-related mortality and uncontrolled viral load. They are less likely to receive ART and those on ART Background: The Wisconsin Department of Health Services’ demonstrate worse adherence, report less confidence in their abili- AIDS/HIV Program has instituted an intervention aimed at increasing ties to integrate treatment regimens into their daily lives, and signifi- ABSTRACTS linkage and retention in medical care among people living with HIV cantly less positive interaction with healthcare providers than non- (PLH). The intervention involves hiring full-time Linkage to Care transgender people. Specialists (LTCSs) within clinical agencies and community-based organizations to provide intensive, short-term case management Description: After formative interviews with patients, providers, and services for small caseloads of PLH who are newly diagnosed, focus groups, Healthy Divas was developed to address culturally recently released from incarceration, or not securely engaged in unique barriers to engagement in HIV care. Healthy Divas consisted medical care. of 3 peer-led individual sessions and 1 peer-led group workshop attended by 2 healthcare providers (1 HIV care expert and 1 in trans- Methods: As part of the qualitative evaluation of the intervention, in- gender health) to provide information, support, and skills building to depth interviews were conducted with 26 current or former clients of identify and accomplish individualized healthcare goals related to LTCSs. Interviews focused on the services LTCSs provided, the pro-

POSTER both gender transition and HIV care and treatment. gram’s successes, and experiences with barriers to medical care. Data were coded for key themes using MaxQDA software. Lessons Learned: Of 21 participants, 19 (90.5%) were retained through follow-up. Among 17 with analyzable Medication Event Results: Preliminary results indicate that the LTC intervention is suc- Monitoring System (MEMS) data, 12 (71%) increased ART adher- cessful in providing basic social support and securing housing. Many ence, 3 (18%) maintained perfect adherence, and 2 (12%) declined in participants had not disclosed their HIV status to family and friends adherence. We found modest increase in self-reported adherence, or indicated few sources of social support outside of their LTCS. with substantial decrease in ART coverage gaps (number of times Some of the LTCS’ housing successes included an assisted living when ART was not taken for at least 4 consecutive days). All partici- facility for an older man hospitalized after a stroke, permanent hous- pants (100%) reported being extremely satisfied with their overall ing for a homeless woman with a 4-year old daughter, residential experience of the intervention. drug treatment for a woman who had just given birth, and transitional housing for a man released from incarceration without family sup- Recommendations: In response to feedback, we will increase the port. Participants indicated that the success their LTCSs had in number of sessions from 3 to 6 and extend the length of the group securing stable housing, as well as direct social support they workshop to permit more discussion of transition-related health care received, motivated them to take care of their own health by keeping content. While MEMS use was not consistent during the trial, use of regular medical appointments. All participants described above indi- MEMS as an adherence feedback tool was well received. The accu- cated undetectable viral loads. racy of self-reported appointment adherence decreased as recall time increased, leading us to refine our methodology for acquiring Conclusions: Preliminary results indicate that short-term, intensive timeline follow-back data. Healthy Divas represents a promising case management interventions such as Wisconsin’s LTC Program approach to improving health outcomes among TWH. Efforts to eval- can contribute to successes in at-risk PLH’s engagement in medical uate its efficacy are in development. care through provision of social support and securing housing. Tenuous family relationships may contribute to housing instability, but can be tempered by direct social support in combination with housing security. The small caseloads carried by LTCSs may allow for greater provision of social support and contribute to greater success rates in locating housing than other types of case management models, and should be balanced against costs and availability concerns.

122 9th International Conference on HIV Treatment and Prevention Adherence 477 pared to15%ofinsuredrespondents(p<0.0001). sured individualshadnotvisitedaproviderinthepastyear also didnotvisitahealthcareprovider. Approximately40%ofunin- sex. Onethirdofmenand25%womenwhoreportedhighrisksex quent bingedrinking(OR=3.6)remainedassociatedwithhighrisk healthcare provider(OR=3.3),crackcocaineuse2.7),andfre- year uninsured and22.9%hadnotseenahealthcareproviderinthepast 51.7% hadahighschooldiploma/GED.32.7%ofparticipantswere poverty (83.0%)ordeep(57.9%),nevermarried(71.6%),and System National HIVBehavioralSurveillance are LivinginPhiladelphia:Resultsfromthe Conclusions: Results: Methods: risk sexfromNHBSamongheterosexualslivinginPhiladelphia. We evaluatedassociationsbetweenhealthcareaccessandhigh- into healthcaresettingsmaynotreachthoseathighestriskofHIV. the UnitedStates.RecommendationstointegrateroutineHIVtesting Background: 1 Kathleen Brady Catherine Mezzacappa other variables(significancelevelp<0.10). race, andmaritalstatusascovariatesbackwardselectionforall partner. Weighted logisticregressionwasperformedincludingage, in exchangefordrugs/money of 3high-risksexualbehaviorsinthepastyear:5+sexpartners, were analyzedtoidentifyfactorsassociatedwithreportingatleast1 increased riskforHIVinfection(HET).Datafromthe2010HETcycle employs respondent-drivensamplingtosurveyheterosexualsat HIV testinginhealthcaresettingsthroughimprovedaccesstocare. ance coverageinthispopulationmayhelprealizethegoalofroutine barrier toaccessingcareinthisat-riskpopulation.Expansionofinsur- abuse counselinginhealthcaresettings.Uninsuredstatuspresentsa HIV infectionarenotreachedbyroutinetestingandsubstance 9th InternationalConference on Philadelphia DepartmentofPublicHealth,Philadelphia,PA, USA . Inmultivariateanalysis,malegender(OR=3.3),notseeinga Respondents werepredominantlyblack(84.0%),livingin The NationalHIVBehavioralSurveillanceSystem(NHBS) A significantnumberofheterosexualsatgreatestrisk Heterosexuals accountfor25%ofnewHIVinfectionsin Heterosexuals atRiskofHIVWho Healthcare Accessamong 1 (presenting) 1 , JenniferShinefeld , andunprotectedsexwithanon-main HIV T reatment andPrevention Adherence 1 , , com - 478 Background: 3 2 Michelle Teti Medication UseamongPeoplewithHIV behavior amongthephotographersandotherswhoseepictures. strategy graphing adherencemotivatorsisapotentialintervention HIV. Ourfindingssuggestthatdirectingtheprocesstowardsphoto- participants thattheycouldbehealthyandincontroloftheirlifewith ence barrierslikethestigmaoftakingmedicines,andreiteratedto on whytheywerecommittedtoadherence,howmanageadher- mount adherencechallenges.Theprocesshelpedparticipantsreflect PLH talkaboutandidentifyreasonstobeadherenthowsur Conclusions: Results: narrative analysisstrategies. and individualsessionsover100photographsusingtheme conducted individualinterviewsandanalyzedtranscriptsofgroup HIV clinicwheretheprojecttookplace.T photo-sharing anddiscussionsessions,aphotodisplayinthe groups. Projectcomponentsincludedaphotographytutorial,3group Methods: ication routines,supportsystems,pets,self-portraits,andnature. take control over one’s health.Imageswerevariedandincludedmed- ness aboutbeinghealthy and remainadherent-importantpeopleorthings,spirituality,hopeful- United States.Keyprojectthemesrevolvedaroundmotivationstobe Participants werelow-incomeresidentsofacityintheMidwest, strategy forPLH. successes andchallenges-asapotentialadherenceintervention using picturestoidentify, discuss,andsharepersonaladherence ed apilotstudytoexplorethefeasibilityanduseofphoto-stories- struggle tousemedicationscorrectlyandconsistently. We conduct- improves healthandcanreduceHIVtransmission-yetmanyPLH to antiviraltreatmentamongpeoplewithHIV(PLH).Adherence 1 Truman MedicalCenter, KansasCity, MO,USA Kansas City, MO,USA The UniversityofMissouri,KansasCitySchoolMedicine, The UniversityofMissouri,Columbia,MO,USA . Photosandstoriescanbeusedtosupportadherence 16 PLH,9womenand6men,tookpartintheproject. Participants tookpartinthephoto-storyprojectsmall 1 Photo-stories areafeasibleandappropriatetooltohelp Novel interventionsareneededtoimproveadherence , Stories asaTool toPromote Snapshots ofAdherence:Photo- Mary Gerkovich , andanalogiesformedicationasawayto 2 (presenting) o assessprojectimpact,we , RoseFarnan 3 -

123 POSTER ABSTRACTS The HIV-HL-2: A Revised Low Health Literacy is Associated 480 Instrument to Assess the Health 481with Poor Attitudes toward HIV Literacy of Persons with HIV Care Providers Raymond Ownby1 (presenting), Drenna Waldrop-Valverde2, Drenna Waldrop-Valverde1 (presenting), Raymond Ownby2, Marcia Holstad2, Albert Anderson2 Albert Anderson1, Marcia Holstad1 1 Nova Southeastern University, Fort Lauderdale, FL, USA 1 2 Emory University, Atlanta, GA, USA Emory University, Atlanta, GA, USA 2 Nova Southeastern University, Fort Lauderdale, FL, USA Background: In a previous project, we developed and provided initial data on a brief measure of health literacy in persons with HIV infection, Background: Relationship with one’s HIV provider is an important the HIV-HL (AIDS and Behavior, 2013). The measure includes an assess- element for successful HIV care. Causal pathways between patient ment of patients’ listening comprehension for health information as well health literacy and health outcomes that rely on the health system of knowledge about the viral life cycle, treatment, and laboratory tests. have been proposed, suggesting that patient satisfaction with

ABSTRACTS The initial version of the HIV-HL was related to patients’ knowledge of provider encounters is essential. Therefore, the relation of health lit- HIV infection and its treatment as well as medication adherence, but it eracy to patients’ attitudes to their healthcare team was evaluated. showed ceiling effects and several of the original 20 items were elimi- nated due to poor psychometric performance. The measure was there- Methods: 323 persons treated for HIV infection participated in an fore expanded to include 25 items with a greater range of difficulties and ongoing study of racial disparities on medication use. Participants is being tested in another study. The purpose of this presentation will be completed questionnaires assessing demographic variables, the to present information on the psychometric characteristics of the Attitudes toward HIV Health Care Providers Scale, and health litera- revised and updated version of the measure, the HIV-HL-2. cy (the HIV-HL-2). Using a validated cut-off score of 16 on the HIV-HL- 2, the attitudes of participants with lower and higher levels of health Methods: As part of a larger study, 323 participants completed the literacy were assessed in a MANCOVA model that included the HIV-HL-2 via computer administration. A subgroup of 180 completed effects of age, education, race, and gender.

POSTER the measure on a second occasion 6 months later. In addition to providing demographic and educational history information, participants Results: Sixty-seven percent of participants were African American also completed measures of self-reported health literacy (e.g., “How (AA), 68% were men, their average age was 49.3 and averaged 12.8 often do you have difficulty understanding written health informa- years of education. Their mean score on the HIV-HL-2 was 16.7 (of a tion?”) and the LifeWindows IMB questionnaire assessing informa- possible total score of 23) with AAs having significantly lower health tion, motivation, and behavioral skills related to HIV treatment. literacy than whites. Skewness and kurtosis were assessed in order to evaluate the effectiveness of a wider range of test item difficulties. Test reliability was Level of health literacy was associated with an overall significant assessed via Cronbach’s alpha and test-retest correlations. Validity multivariate effect on patient attitudes toward their healthcare was assessed via correlations with other measures and receiver providers (F [19, 298] = 2.34, p < 0.001). Patients with lower levels of operating curve (ROC) analyses were used to identify cut-off values health literacy rated their medical team lower in knowledge about for the HIV-HL-2 associated with low levels of health literacy. HIV/AIDS, level of effort put into treatment and caring and were more Results: Item analyses showed that 2 items had low or negative item- likely to state that they believed their medical team saw the patient total correlations and were dropped from the scale. The scale’s mean as a bad person or “stupid” (all ps < 0.05). Interaction effects showed was 16.7 (of 23 items) while it median was 17, with skewness (-0.80) and that whites with low health literacy agreed more that providers saw kurtosis (0.64) values within acceptable ranges. Its internal consisten- them as “stupid.” cy was 0.75 (Cronbach’s alpha) and its test retest correlation over 6 months was 0.72. The scale’s total score was significantly correlated Conclusions: Low health literacy was associated with poorer atti- with the 3 LifeWindows scale scores and the self-report health literacy tudes toward one’s health care providers and may be a mechanism questions (e.g., r with LifeWindows Information = 0.30 and r with confi- by which health literacy affects health status and outcomes. dence in filling out medical forms = -0.40; for all correlations, ps <0.05). ROC analyses using the R statistical package OptimalCutpoints targeting maximum sensitivity and specificity showed that a cut-off score of 16 (of possible total of 23) had a sensitivity of 0.73 and a specificity of 0.67 in detecting individuals with difficulties in understanding written information; a value of 14 provided a sensitivity of 0.88 while a value of 22 yielded a specificity of 1.0. The area under the curve (AUC) for identifying self-reported difficulty in understanding health information was 0.70 (CI 0.62-0.78) with a positive predictive value (PPV) of 0.87, while that for identifying persons with less than 12 years of education was 0.77 (CI 0.70-0.83) with a positive predictive value of 0.91.

Conclusions: This additional development has been effective in improving some psychometric characteristics of the HIV-HL while preserving its relations to other measures. The HIV-HL-2 continues to be a promising measure for the assessment of HIV-related health literacy skills. This new version is now available on-line for use by qualified users.

124 9th International Conference on HIV Treatment and Prevention Adherence 484 Initiative: ImprovedRetentioninCare closest totheendofmeasurementperiod)werecompared. urement period)andviralsuppression(HIV-1 RNA<200copies/mL urement period),receiptofART(prescriptionduringthemeas- (evidence of2ormedicalvisitsatleast90daysapartina1-yearmeas March 31,2013.Pre-andpost-interventionratesofretentionincare ment andenrolledinBHCservicesbetweenJanuary1,2012 mined tohavebehavioralhealthneedsbasedonapreliminaryassess analysis includespersonswithHIV/AIDS(PLWHA), whoweredeter- housed in6ofPhiladelphia’ 1 Conclusions: Results: Methods: Background: 2 Susan Lehrman health andHIV-related healthdisparities. care, increasedratesofviralsuppression,andreducedbehavioral mental healthcare.Programobjectivesincludeimprovedretentionin primary caresettings,themodelisnotdesignedtoreplacespecialty Focused onimprovingthetreatmentofbehavioralhealthproblemsin time, site-basedmembersofHIVmedicalcarespecialtyteams. Consultants (BHCs),typicallyLCSWsorMSWs,areintegratedasfull- Primary CareBehavioralHealthmodel,inwhich Behavioral HealthInitiative(PIBHI).”Theprojectisbasedonthe implemented theSAMHSA-funded“PhiladelphiaIntegrative Coleman Terrell port BHCservicesmovingforward. ed, asisthedevelopmentofsustainablefundingmechanismstosup- retention incareandARTuse.Furtherstudyofthismodeliswarrant- individuals receivedARTand80.0%werevirallysuppressed(p=.06). recipients wereretainedincare(P<0.05),93.0%<0.05)ofthese virally suppressed.Inthefollowupperiod,90.9%ofBHCservice 80.7% wereretainedincare,87.0%onARTand74.6% the studyperiod.Baselinedataonthoseenrolledindicatedthat 9th InternationalConference on Philadelphia DepartmentofPublicHealth,Philadelphia,PA, USA Philadelphia University, Philadelphia,PA, USA Overall, 471PLWHA wereenrolledinBHCservicesduring 6 BHCswerefullyembeddedin9clinicalsiteswhichare In 2012,thePhiladelphiaDepartmentofPublicHealth BHC servicesinPL Integrative BehavioralHealth Early ResultsfromthePhiladelphia 1 2 , Emerson Evans s largestHIVtreatmentprograms.This HIV 2 WHA resultedinimprovements (presenting) T reatment andPrevention Adherence , KathleenBrady 2 , - - 485 1 Methods: Background: 2 Aimalohi Ahonkhai Early afterARTInitiationinNigeria Conclusions: Results: Ifeyinwa Onwuatuelo through 12/2012.UCIwasdefinedas ic andstartedARTbetween1/200912/2011.Follow-upwas affiliated HIVclinic.Thecohortincludedadultswhoenrolledinclin- (ART) inNigeria. consequence ofUCIinthefirst6monthsonantiretroviraltherapy been wellstudied.Ourobjectivewastodeterminetheimmunologic common inresource-limitedsettings,butitsclinicalimpacthasnot Kenneth Freedberg 4/month [95%CI-11,3]. [95%CI 6,12];thosewith2UCIhadameanCD4countdecreaseof first 6monthsonARTwas18/month[95%CI17,20]and8/month [IQR 213,562].Forthosewith0and1UCI,meanCD4increaseinthe and 2UCIwere190/uL[IQR92,285],211/uL95,374],403/ul 1, and8%had2UCI.MedianbaselineCD4countsforthosewith0, the immunologicbenefitsofART. Interventions tomaintainconsistentHIVcarearecriticalmaximize counts, negatingthepotentialbenefitofearlypresentation. highest CD4values,patientswith2UCIhadnoincreaseintheir CD4 increaseofthosewithnoUCI.DespiteinitiatingARTthe In thefirst6monthsonART months) afterARTinitiationineachgroup. model thechangeinCD4countearly(0-6months)andlater(>6 multivariate linearregressionadjustingforrepeatedmeasuresto patients into3groups:0,1,or2UCIinthefirstyearonART. We used oratory AIDS PreventionInitiativeinNigeria,Abuja,Nigeria Massachusetts GeneralHospital,Boston,MA,USA , orpharmacyvisits,butlaterreturntocare.W Among the2,027patientsinourcohort,54%had0,37% We conductedaretrospectivecohortstudyatuniversity- Unplanned careinterruption(UCI)fromHIVtreatmentis UCI wascommoninacohortinitiatingARTNigeria. Interruptions onCD4Response Impact ofUnplannedCare 1 1 , ProsperOkonkwo (presenting) 2 , IngridBassett , patientswith1UCIhadhalftherateof , JulietAdeola ³ 1 , SusanRegan 90 dayswithnoclinician,lab 2 , ElenaLosina 2 , BolanleBanigbe 1 , 1 e categorized 2 , -

125 POSTER ABSTRACTS Differences in Outcomes of Utilization of an Interactive Online 487 Engagement in the HIV Care 488 Tool to Visualize Geographic Continuum by Relationship Status and Patterns in the HIV Treatment Cascade in Partner Serostatus among Young Black Philadelphia HIV-Positive MSM Kathleen Brady1 (presenting), Michael Eberhart1, Alexandra Ricca2, Patrick Sullivan2, Travis Sanchez2 Judy Tan1 (presenting), Lance Pollack1, Greg Rebchook1, Susan Kegeles1 1 Philadelphia Department of Public Health, Philadelphia, PA, USA 2 Emory University, Atlanta, GA, USA 1 Center for AIDS Prevention Studies, University of California San Francisco, San Francisco, CA, USA Background: The HIV care continuum has been used to evaluate performance at the national, state, and local level on linkage to care, reten- Background: There is a dearth of research that examines engage- tion in care and viral suppression. Little attention has been paid, how- ABSTRACTS ment in the HIV care continuum (CC) among couples, and even less ever, to how these outcomes vary by geography. Online tools are robust among couples of young black men who have sex with men for mapping illness data, and allow for insights into geographic patterns (YBMSM), a population with some of the worst outcomes along the of illness and care that are not possible with aggregate data. CC relative to other racial/ethnic MSM groups in the United States. Despite the potential of antiretroviral therapy (ART) to lower trans- Methods: HIVcontinuum.org is a collaboration of academic, private, mission rates, the efficacy of treatment-for-prevention approaches and public health partners that has developed an online resource for for HIV-positive (HIV+) YBMSM with serodiscordant partners may be visualizing HIV continuum elements by geography. Philadelphia HIV limited due to poor engagement in the CC, and developing efficacious surveillance/laboratory data for adult/adolescent residents who were interventions for improving their care engagement is paramount. The diagnosed with HIV during 2006-2010 were utilized. The zip-code level present study aimed to examine differences in engagement in care maps include outcomes of new HIV diagnoses (5-year risk of new HIV outcomes among HIV+ YBMSM with and without primary partners

POSTER diagnosis), late HIV diagnoses (proportion who had an AIDS diagno- (i.e., boyfriends), and HIV+ YBMSM in HIV seroconcordant vs. sis within 12 months of initial HIV diagnosis), linkage to HIV care (pro- serodiscordant relationships. portion linked to care within 3 months of HIV diagnosis), engagement in HIV care (proportion living in 2011 who had a CD4 count and/or a Methods: A sample of young (ages 18-29) HIV+ BMSM in Dallas and viral load), viral suppression among those engaged in care (engaged Houston, TX. Participants completed measures regarding Linkage to in care in 2011 and most recent viral load <200 copies/mL) and viral care (have a primary health care provider), any care in past 6 months, suppression among the newly diagnosed (proportion newly diag- adherence to ART (any skipped doses in past 30 days), and demo- nosed whose most recent viral load was <200 copies/mL). graphic information. Results: HIVContinuum.org allows viewers to look at ZIP-code/ward- Results: Data across single and partnered men (N = 290) were com- level HIV surveillance data by race/ethnicity, sex and age groups. pared on engagement in the CC variables. Chi-square difference The presentation will review the features of this new online tool tests showed that men with a primary partner reported being more using data from Philadelphia, a city highly impacted by HIV. engaged in HIV care than single men in terms of receiving HIV treatment in past 6 months, X2(1) = 6.68, p = .01, and in having a primary Conclusion: Geographic analysis of the HIV care continuum can be health care provider, X2(1) = 7.01, p = .01. No statistically significant instrumental in targeting areas that need additional/improved servic- differences were found between partnered and single men on ART es, and may help to mobilize public health, policy, and health care adherence. However, Chi-square difference tests comparing men resources. with seroconcordant positive partners with those with serodiscordant partners showed that men with seroconcordant partners reported worse adherence in the last 30 days than men with serodiscordant partners, X2(1) = 6.65, p = .01, while engagement in the CC did not vary by partner serostatus.

Conclusions: These data suggest that partners may play an integral role in engagement in the HIV CC and adherence to treatment, and that optimizing care and treatment outcomes may be improved through intervention strategies that incorporate a couples approach. Further, interventions to improve HIV care engagement and ART adherence for partnered HIV+ BMSM should be tailored to whether their partner is HIV- or HIV+.

126 9th International Conference on HIV Treatment and Prevention Adherence 491 an optiontonotifyabuddyiftheuserfallsofftheirregimen. importance oftakingeverydoseday “reason photo,”whichservesasaninspirationalreminderofthe tion oftheirhealthcareteam.Theappallowstheusertouploada effects, tipstostayadherent,andaplacerecordcontactinforma- counts, acomponenttorecordmisseddosesandmedicationside dose andappointmentreminders,trackingviralloadCD4cell among PL support medicationadherence.Aprototypewascreatedandpiloted conducted anenvironmentalscanoftheliteratureonmobileappsto along withadditionalelementstoimproveuserinteractivity. We also create anappthatpossessedkeyfeaturesofthepagerintervention researcher ofthepagerinterventionandamobileappdeveloperto Prevention, CapacityBuildingBranchpartneredwiththeoriginal ent totheirHIVtreatment. mobile apptoassistpeoplelivingwithHIV(PLWH) instayingadher- project thatupdatedanevidence-basedpagerinterventionintoa ingly beingusedforthepreventionandcareofHIV. We describea based interventionsandimproveuserinteractivity. developers toadequatelyincorporatekeyelementsofevidenced- tions. Furthermore,publichealthorganizationsshouldworkwithapp formative workiscriticaltocustomizingfeaturesforspecificpopula- Recommendations: that theappdidmorethantheyinitiallyexpected. they wouldnotusethisfeature.Morethan50%ofparticipantsfound half ofparticipantsstatedwhilethebuddyfeaturemaybebeneficial, eral, participantsfoundinitialset-upandloggingdoseseasy. About thought theappwouldserveasavaluabletoolforadherence.Ingen Lessons Learned: Description: Introduction: 1 Angel Ortiz-Ricard Kathleen Green to AntiretroviralTreatment a MobileApplicationtoImproveAdherence 9th InternationalConference on Centers forDiseaseControlandPrevention,Atlanta,GA,USA WH (N=9)during60-minuteinterviews.Theappfeatures Mobile andsmartphoneapplications(app)areincreas- Intervention PagerMessaginginto Updating theEvidenced-Based In August2011,theCDC’s DivisionofHIV/AIDS 1 (presenting) 1 Pilot interviewsrevealedthatallparticipants In developingmobileinterventionsforPL , RhondetteJones HIV T reatment andPrevention Adherence . Lastly 1 , , theappfeatures WH, - States on aSocialMediaPlatformintheUnited Messages aboutPre-ExposureProphylaxis what typeofinformationisbeingcommunicated. media platform,isbeingusedtodiscussPrEPand,ifso,bywhomand exploratory studytodetermineifT media isbeingusedtocommunicateaboutPrEP, weconductedan quently enhancePrEPuptake.Becauseitisunknownwhethersocial such associalmediamayhelptoincreaseawarenessandconse- about PrEPamongpotentialusers.Usingcommunicationchannels low intheUnitedStates,whichmaypartlybeduetoawareness Background: 3 2 1 Viraj Patel Conclusion: Results: Methods: awareness andinfluencePrEPuptake. implement anddisseminateinterventionsthataimtoincreasePrEP given theirbroadandefficientreach,maybeanimportant‘venue’to group ofstakeholdersonsocialmedia.Twitter andsimilarplatforms, Oni Blackstock identified emergingthemes. coded each140-charactertweet,andthroughaniterativeprocess and sizeoftheauthor’s socialnetwork.Usingcontentanalysis,we combinations. We collectedinformationabouteachtweet’s author “pre-exposure prophylaxis,”“PrEP[and]HIV and postedfromJune2012to2013.We usedthesearchterms sages (“tweets”)referencingPrEPoriginatingintheUnitedStates towards PrEPorreportedonpersonalexperienceswith(<5%). cost) aboutPrEP, andaminorityexpressednegative attitudes promoted PrEP, 13%containedquestionsorconcerns(e.g.,safety, links tootherinformationaboutPrEP shared linkstoscientificarticlesornewsaboutPrEPand22% related publications(20.9%).Thelargestproportionoftweets(29%) individuals (35.7%),non-profitorganizations(32.9%),andhealth- users (rangeofsocialnetworksize:0-835,517).Userswereprimarily (after removingduplicatesandre-tweets)postedby1,149unique Hunter College,CUNY, NewYork, NY, USA Albert EinsteinCollegeofMedicine,Bronx,NY, USA Bronx, NY, USA Montefiore MedicalCenter;AlbertEinsteinCollegeofMedicine, 492 W Using T 1 e identified2,822messagesaboutPrEPandcoded1,358 (presenting) There ispublicdiscourseaboutPrEPinvolvingadiverse HIV pre-exposureprophylaxis(PrEP)uptakeremains 1 witter’ A ContentAnalysisof Who’s Tweeting about#PrEP? s publicsearchtool,weextractedallmes , Vanessa Eddy witter . Anadditional18%oftweets 2 , SaritGolub , thesecondlargestsocial ,” andothersimilar 3 , -

127 POSTER ABSTRACTS Paying for PrEP: Eligibility for The Challenges of Re-Engaging 494 Assistance Programs and 495 the Lost to Follow-Up: Early Willingness to Contribute to Cost among Lessons from a Social Innovation NYC Health Department STD Clinic Patients, Demonstration Project 2013-4 David Batey1 (presenting), William Anderson1, Jeremiah Rastegar2, Andrew Douglas1, Karen Musgrove2, Michael Mugavero1, Zoe Edelstein1 (presenting), Christine Borges2, Emily Westheimer2, James Raper1 Sarit Golub3, Julie Myers1 1 University of Alabama at Birmingham, Birmingham, AL, USA 1 New York City Department of Health and Mental Hygiene, 2 Birmingham AIDS Outreach, Birmingham, AL, USA Queens, NY, USA 2 New York City Department of Health and Mental Hygiene, New Introduction: Fifty percent of HIV-diagnosed individuals are eventu- York, NY, USA ally lost to care. Limited research has provided insight into strategies ABSTRACTS 3 Hunter College, CUNY, New York, NY, USA by which to identify and re-engage, yet little is known about this population and best practices for locating them. Background: The cost of HIV pre-exposure prophylaxis (PrEP) poses a challenge for publically-supported programs. Uninsured patients Description: Birmingham Access to Care (BA2C) is a partnership earning <$20,000/year are eligible for current drug manufacturer between a university and an AIDS service organization, formed to assistance programs. Those earning more or those without insur- characterize and re-engage the lost to follow-up population of an ance may have to pay for some or all of the medication. We examined urban HIV clinic. Eligible persons were identified using the HRSA whether patient contribution might be an acceptable way to defray HIV/AIDS Bureau (HAB) measure of kept clinic visits, people who costs among New York City (NYC) STD clinic attendees. had not arrived at clinic in the past 7 months, and those who missed 2 visits within the last 12 months. Contact procedures included a 12- Methods: Data were derived from a 2013-2014 survey conducted to ³

POSTER month series of telephone calls to all known patient numbers, as well assess acceptance of a proposed PrEP program. Respondents were as certified letters to last known addresses. NYC residents, ³18 years who met the criteria for acute HIV testing (predominantly men with male sex partners) (n = 389). This analysis Lessons Learned: Application of the HRSA-HAB measure yielded 816 was restricted to respondents who reported being somewhat/very individuals; applying the 7-month with no visit and ³2 missed visits likely to take PrEP and participate in the proposed program. We during the past 12 months criteria increased the sample by 255 and described self-reported insurance status, household income, and 152, respectively. Early results indicate that waiting for patients to maximum amount willing to pay for PrEP at the clinic ($0;$5-15;$15- become eligible under more definitive out of care criteria results in 30;$30-50;$50-100;$100+). We also evaluated differences in propor- greater difficulty contacting patients and may jeopardize subsequent tion willing to pay by age, race, insurance status, income, attempts to intervene. current/prior STD diagnosis, perceived HIV risk, and prior awareness of PrEP using Poisson regression. Recommendations: Using both kept- and missed-visit measures of retention that illustrate patients’ visit profiles early on increases the Results: Among respondents likely to take PrEP and participate in the likelihood that contact information will still be valid and established proposed program (n = 231), 45% were uninsured and 82% earned locator strategies will be successful. Alternative means of locating ³$20,000/year, including 43% who were also uninsured. Most (91%) the lost to follow-up population such as those utilizing shared surveil- were willing to pay for PrEP; 33% were willing to pay ³$30/month. lance may improve re-engagement efforts. These recommendations Overall willingness to pay did not differ significantly by factors exam- point to the need to intervene earlier after the patient falls out of care. ined; willingness to pay ³$30/month was associated with income ³$20,000/year and prior awareness of PrEP. Conclusions: In NYC, 9 of 10 high-risk STD clinic patients interested in PrEP reported willingness to pay and one-third were willing to pay ³$30/month. This information is important given high interest in PrEP and apparent lack of payment options (insurance or assistance eligibility) for many. Patient contribution may be an acceptable way to reduce programmatic cost and increase access to publically-funded PrEP programs.

128 9th International Conference on HIV Treatment and Prevention Adherence 496 Women: ExploringthePaucityofEvidence 1 San Francisco. views withpotentialPrEPproviderswhoworktranswomenin additional focusgroupsandindividualinterviews,includinginter- barriers andfacilitatorstoadherence.Nextstepsincludeconducting spectives onPrEP, includingattitudestowarduptakeandpotential ferences betweenMSMandtranswomenintermsofcommunityper- high levelsofmedicalmistrust. concerns aboutsafety potential preventiontoolweremixed,withmostfrequentlyvoiced participants hadneverheardofPrEP PrEP orhavingbeenofferedbyaprovider. Thevastmajorityof Recommendations: Lessons Learned: Spanish (n=8). we havecompleted2focusgroups,1inEnglish(n=11)and including barriersandfacilitatorstouptakeadherence.To date knowledge, beliefs,andattitudestowardPrEPasapreventiontool, of 5)HIV-negative transwomeninSanFranciscotoexploretheir Description: ing transwomeninthestudy(Cohen,2014). stration projectinSanFranciscohaddifficultyrecruitingandretain- transwomen inthestudy(www.bodypro.com, 2013).ThePrEPdemon- ed thatthiswaslikelyduetobarriersadherenceamong enrolled iniPrEx(Mascolini,2011).investigatorshavespeculat- that therewasnoevidenceofPrEPefficacyamongtranswomen grouped transwomenwithMSM,subanalysisofthatdatarevealed PrEP demonstrationsiteshavebeenbasedoniPrExresultsthat sure prophylaxis(PrEP)asapreventiontool.Whileguidelinesfor research, includingrecenteffortstoinvestigatetheuseofpre-expo- tently groupedwithmenwhohavesexinHIV-related and culturalcontext,transgenderwomen(‘transwomen’)areconsis- Introduction: Jae Sevelius 9th InternationalConference on Transgender Health,SanFrancisco,CA,USA University ofCalifornia,SanFrancisco,CenterExcellencefor W 1 Despite significantdifferencesinbehavioralriskfactors Prevention amongTransgender Pre-Exposure ProphylaxisforHIV (presenting) e haveinitiatedaseriesoffocusgroups(completed2 None oftheparticipantsreportedhavingtaken Research isneededtofurtherelucidatethedif , sideeffects,andcost.Participantsexpressed , JoAnneKeatley HIV T . LevelsofinterestinPrEPasa reatment andPrevention Adherence 1 - 497 Background: 1 Megan McCormickKing Conclusions: Results: Methods: ical information. for longerwouldshowbetterHL,givenincreasedexposuretomed- addition, weanticipatedthatpatientswhohadreceivedmedicalcare Reading Comprehension)woulddemonstratelower(better)VL.In of FunctionalHealthLiteracyinAdults(TOFHLA;Numeracyand ed thatpatientsreceivinghigherscoresonbothsubtestsoftheTest examine HLandVLinasampleofyouthlivingwithHIV. We predict- ble predictorofadherenceandVL.Thecurrentstudysoughtto Health literacy(HL)is1factorthathasreceivedattentionasapossi- those patientsatriskfornegativeoutcomesassociatedwithhighVL. ence. UnderstandingfactorsrelatedtoVLisimportantforidentifying marker ofdiseaseactivity, treatmentsuccess,andmedicationadher- and treatmentofthisconstructinmedicalpractice. our understandingof“healthliteracy and exposuretomedicalsettings.Thesedatahaveimplicationsfor more generalcognitiveabilitiesthatarerelativelystableacrosstime need notbespecifictohealth-relatedinformation,butratherreflect ical calculationabilities.Inaddition,itispossiblethattheseskills understanding and/orfollowingmedicalinformationthanarenumer comprehension, andlogicalreasoningskillsmaybemorehelpfulfor tests oftheTOHFLAindicatepossibilitythatliteracy, reading time sincediagnosis. subtest. Inaddition,TOFHLAscoreswerenotcorrelatedwithageor test oftheTOFHLA.Itwasnot,however, correlatedtheNumeracy (r =-.269,p.035)withscoresontheReadingComprehensionsub of routinemedicalcareinanadolescentHIVclinic. male, 99%AfricanAmerican)wereadministeredtheTOFHLAaspart Children’s NationalMedicalCenter, Washington, DC,USA As expected,VLwassignificantlyandnegativelycorrelated Sixty-two patientsbetweentheagesof16and24(63% The novelanddifferentialfindingsbetweenthe2sub- For peoplelivingwithHIV, viralload(VL)isacritical Youth LivingwithHIV Health LiteracyandViral Loadin 1 (presenting) ,” aswelltheassessment - -

129 POSTER ABSTRACTS Poverty - A Risk Environment for Predictors of Participant Dropout in 498 Suboptimal Engagement in HIV 499 ART Adherence RCTs Care among Low-Income Women Living with Kathy Goggin1 (presenting), Dwight Yin1, Yan Wang2, HIV in the Deep South Glenn Wagner3, Ira Wilson4, Marc Rosen5, Carol Golin6, Nancy Reynolds7, Robert Remien8, Jane Simoni9, Judith Erlen10, Melonie Walcott1 (presenting) Julia Arnsten11, Robert Gross12, Neil Schneiderman13, David Bangsberg14, Honghu Liu15 1 University of Alabama at Birmingham, Birmingham, AL, USA 1 Children’s Mercy Hospital/University of Missouri, Background: The HIV epidemic among women in the United States is Kansas City, MO, USA concentrated in the South, especially in areas with high rates of 2 University of California, Los Angeles, CA, USA poverty. Among HIV-infected women in the United States, only 53% 3 RAND Corporation, Santa Monica, CA, USA are in care and only 43% have controlled viral loads. The relationship 4 Brown University, Providence, RI, USA between poverty and engagement in care is complex and merits fur- 5 Yale University School of Medicine, New Haven, CT, USA ABSTRACTS ther investigation. 6 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA 7 Yale University, West Haven, CT, USA Methods: We conducted a qualitative study to describe the role of 8 HIV Center for Clinical and Behavioral Studies, New York, NY, USA poverty in engagement in HIV care among low-income HIV-infected 9 University of Washington, Seattle, WA, USA women in Birmingham, Alabama. We conducted focus groups dis- 10 University of Pittsburgh, Pittsburgh, PA, USA cussions (7) with HIV-infected women (46), and in-depth interviews 11 Albert Einstein College of Medicine, Montefiore Medical Center, with stakeholders (14) including service providers, and community Bronx, NY, USA based organizations. The data were analyzed using a modified 12 University of Pennsylvania, Philadelphia, PA, USA Grounded Theory approach. NVivo qualitative software was used in 13 University of Miami, Miami, FL, USA the management and analysis of the data. 14 Massachusetts General Hospital, Boston, MA, USA 15

POSTER UCLA School of Dentistry, Los Angeles, CA, USA Results: Most of the women were African American (89%) and earned <$11,000.00 annually (64%). Participants described how Background: Participant dropout in randomized controlled trials poverty creates a risk environment for sub-optimal engagement in (RCTs) is costly, reduces statistical power, and may introduce selec- care at different levels. Mechanisms at the community level included tion bias. Few studies have investigated baseline predictors of transportation challenges, easy access to drugs, high rates of unem- dropout in multiple RCTs of behavioral interventions for antiretroviral ployment, stigma, and poor access to health care. Health insurance therapy (ART) adherence, however identification could inform strate- status often dictated where the women went for HIV care - public gies to retain and oversample participants at risk for dropout. infectious disease clinic vs doctor’s office, with implications for engagement in care. Participants reported stigma associated with Methods: Utilizing data from the Multi-site Adherence Collaboration accessing care at facilities for low-income individuals and facilities on HIV among 14 institutes (MACH14), this study explored the asso- designated for HIV care contributed to sub-optimal engagement in ciation between baseline demographics (age, gender, race, educa- care. Mechanisms at the individual level included lack of education, tion, employment, and housing status), depression, substance use, inability to meet basic needs (resulting in health not being a priority), and participant burden in predicting dropout over time. Participants and substance use. were coded as dropouts when they ceased to provide electronic drug monitoring (EDM) data confirmed by ensuring that self-report Conclusion: For these women, the urgency of meeting basic survival data were also never provided at a subsequent visit. needs and avoiding stigma appeared to supersede adopting behaviors that promote engagement in HIV care. The importance of Results: Data from 1,272 participants enrolled in 9 studies were ana- addressing both financial needs and stigma as experienced by HIV- lyzed. Participants had a mean age of 42 years (SD = 8.3), 68% male, infected women in HIV programs is highlighted. 12% were Latino, and 18% dropped out [95% CI (14-30)] prior to 48 weeks. Median time of follow-up was 12 months (interquartile range 9 to 12 months). We conducted a series of Cox proportional hazard models that were unadjusted, then adjusted only for demographics, to identify the candidate variables for the final parsimonious model. This final Cox proportional hazards model revealed that dropout was significantly associated with being male [hazard ratio 1.49, 95% CI (1.09, 2.03)] and Latino [1.94, CI (1.28, 2.96)].

Conclusions: To our knowledge, this study is the first to explore predictors of dropout in behavioral interventions for ART adherence RCTs that utilized EDM. Men were 1.5 times more likely to drop out than females and Latinos were nearly twice as likely to drop out than White participants. Researchers should focus retention and/or over- sampling efforts on Latinos and men to protect against dropout.

130 9th International Conference on HIV Treatment and Prevention Adherence Nova SoutheasternUniversity for Medicine Postgraduate Institute Association ofNursesinAIDSCare ACCREDITED PROVIDERS Providers ofAIDSCare International Associationof SPONSOR 9th InternationalConference on HIV T reatment andPrevention Adherence PLANNING COMMITTEE Nairobi, Kenya National Hospital University ofNairobiandKenyatta Nellie R.Mugo,MBChB,MMed, MPH Washington, DC,USA Research The AmericanFoundationforAIDS Gregorio Millett,MPH Boston, MA,USA Harvard University Kenneth Mayer, MD Pavia, Italy University ofPavia Franco Maggiolo,MD West Yorkshire, UK University ofHuddersfield Ian Hodgson,MA,PhD Houston, TX,USA Baylor University Thomas P Atlanta, GA,USA Sister Love Dazon DixonDiallo,MPH Amsterdam, Netherlands University ofAberdeen Marijn deBruen,PhD Denver, CO,USA University ofColorado,Denver Marla A.Corwin,LCSW, CACIII San Francisco,CA,USA School ofPharmacy University ofCaliforniaSanFrancisco Jennifer Cocohoba,PharmD,MAS Lima, Peru Universidad PeruanaCayetanoHeredia Carlos Cáceres,MD,PhD,MPH Storrs, CT, USA University ofConnecticut K. RivetAmico,PhD CO-CHAIR . Giordano,MD,MPH Pomona, CA,USA Western University James D.Scott,PharmD,MEd Boston, MA,USA Harvard MedicalSchool Steven A.Safren,PhD Stellenbosch, SouthAfrica Stellenbosch University Jean B.Nachega,MD,PhD Birmingham, AL,USA University ofAlabama Michael J.Mugavero,MD,MHSc CO-CHAIR Washington, DC,USA AIDS Care International AssociationofProviders José M.Zuniga,PhD,MPH Waterloo, ON,Canada Wilfrid LaurierUniversity Alan Whiteside,DEcon Akron, OH,USA Association ofNursesinAIDSCare Allison Webel, PhD,RN Geneva, Switzerland W Marco Vitória, MD Bethesda, MA,USA National InstituteofMentalHealth Michael J.Stirratt,PhD Seattle, WA, USA University ofWashington Jane M.Simoni,PhD Fort Lauderdale,FL,USA Nova SoutheasternUniversity Elizabeth Sherman,PharmD orld HealthOrganization

131 ACKNOWLEDGEMENTS 132 ACKNOWLEDGEMENTS San Francisco,CA,USA University ofCalifornia Robert M.Grant,MD,PhD W National InstituteofMentalHealth Christopher M.Gordon,PhD Houston, TX,USA Baylor University Thomas P. Giordano,MD,MPH Denver Denver PublicHealth Edward M.Gardner Bethesda, MA,USA Infectious Diseases National InstituteofAllergyand Vanessa Elharrar, MD,MPH Atlanta, GA,USA Sister Love Dazon DixonDiallo,MPH Amsterdam, Netherlands University ofAberdeen Marijn deBruen,PhD Denver, CO,USA University ofColorado,Denver Marla A.Corwin,LCSW, CACIII Durham, NC,USA FHI360 Amy Corneli,PhD San Francisco,CA,USA School ofPharmacy University ofCaliforniaSanFrancisco Jennifer Cocohoba,PharmD,MAS Washington, DC,USA George Washington University Amanda Castel,MD,MPH Lima, Peru Universidad PeruanaCayetanoHeredia Carlos Cáceres,MD,PhD,MPH Washington, DC,USA National MinorityAIDSCouncil Moises Agosto Storrs, CT University ofConnecticut K. RivetAmico,PhD ABSTRACT REVIEWCOMMITTEE ashington, DC,USA , CO,USA , USA , MD Entebbe, Uganda Medical ResearchCouncil Paula Munderi,MD Birmingham, AL,USA University ofAlabama Michael J.Mugavero,MD,MHSc Nairobi, Kenya National Hospital University ofNairobiandKenyatta Nellie R.Mugo,MBChB,MMed,MPH Bethesda, MD,USA Infectious Diseases National InstituteofAllergyand Tia Morton,RN, MS Washington, DC,USA The AmericanFoundationforAIDSResearch Gregorio Millett,MPH W Pan AmericanHealthOrganization Rafael Mazin,MD,MPH Boston, MA,USA Harvard University Kenneth Mayer, MD Rockville, MD,USA Administration Health ResourcesandServices Marlene Matosky Pavia, Italy University ofPavia Franco Maggiolo,MD Jackson, MS,USA University ofMississippiMedicalCenter Deborah Konkle-Parker Atlanta, GA,USA Centers forDiseaseControlandPrevention Linda Koenig,PhD,MS Bethesda, MD,USA National InstituteonDrugAbuse Shoshana Kahana,PhD West Yorkshire, UK University ofHuddersfield Ian Hodgson,MA,PhD Philadelphia, P University ofPennsylvania Robert Gross,MD,MSCE ashington, DC,USA A, USA 9th InternationalConference on , MPH,RN , PhD,FNP Pomona, CA,USA W James D.Scott,PharmD,MEd Boston, MA,USA Harvard MedicalSchool Steven A.Safren,PhD Miami, FL,USA University ofMiami Allan E.Rodriguez,MD Cape Town, SouthAfrica Desmond Tutu HIVCentre MMed Catherine Orrell,MBChB,MSc, Stellenbosch, SouthAfrica Stellenbosch University Jean B.Nachega,MD,PhD Washington, DC,USA AIDS Care International AssociationofProviders of Benjamin Young, MD,PhD Waterloo, ON,Canada Wilfrid LaurierUniversity Alan Whiteside,DEcon Akron, OH,USA Association ofNursesinAIDSCare Allison W New Y AIDS V Mitchell Warren Geneva, Switzerland World HealthOrganization Marco Vitória, MD W National AssociationofSocialWorkers Evelyn T Bethesda, MA,USA National InstituteofMentalHealth Michael J.Stirratt,PhD Seattle, WA, USA University ofWashington Jane M.Simoni,PhD Fort Lauderdale,FL,USA Nova SoutheasternUniversity Elizabeth Sherman,PharmD HIV estern University ashington, DC,USA T reatment andPrevention Adherence ork, NY accine AdvocacyCoalition omaszewski, MSW ebel, PhD,RN , USA , ACSW June 8-10, 2014

To Whom It May Concern:

This letter is a confirmation that ______attended the 9th International Conference on HIV Treatment and Prevention Adherence, held June 8-10, 2014, at the Loews Miami Beach Hotel in Miami, FL, USA. This 2.5-day conference was jointly sponsored by the International Association of Providers of AIDS Care (IAPAC) and the Postgraduate Institute for Medicine (PIM).

Sincerely,

José M. Zuniga, PhD, MPH President/CEO International Association of Providers of AIDS Care ¡

Institutional Supporters

Commercial Supporters

Gilead ViiV AbbVie Merck & Co. Sciences Healthcare