33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 267

Interventions Targeting HIV­Infected Risky Drinkers

Drops in the Bottle

Jeffrey H. Samet, M.D., M.A., M.P.H., and Alexander Y. Walley, M.D., M.Sc.

Alcohol use is common among people infected with HIV and may contribute to adverse consequences such as reduced adherence to treatment regimens and increased likelihood of risky sexual behaviors. Therefore, researchers and clinicians are looking for treatment approaches to reduce harmful alcohol consumption in this population. However, clinical trials of existing treatment models are scarce. A literature review identified only 11 studies that included HIV­infected patients with past or current risky alcohol use and which targeted alcohol use and other health behaviors. Four studies focusing on HIV­infected participants with alcohol problems found mixed effects on adherence and on alcohol use. Five clinical trials included at least 10 percent of HIV­infected subjects who use alcohol; of these, only one reported significant evidence of a favorable impact on alcohol consumption. Finally, two trials targeting alcohol users at high risk for HIV infection identified treatment effects that were not sustained. Taken together, these findings provide limited evidence of the benefit of behavioral interventions in this population. Nevertheless, these studies give some guidance for future interventions in HIV­infected patients with alcohol problems. KEY WORDS: Alcohol and other drug use; alcohol consumption; alcohol use disorder; human immunodeficiency virus; HIV­infected patients; sexually transmitted disease; unsafe sex; treatment method; treatment outcome; intervention; clinical trial; literature review

n the United States, people infected • Lack of a health care provider for the 1 According to the National Institute on Alcohol Abuse and HIV infection (Metsch et al. 2009); Alcoholism (2007), women who drink more than 3 drinks per with the human immunodeficiency day or more than 7 drinks per week and men who drink more virus (HIV) drink more alcohol than 4 drinks per day or more than 14 drinks per week are at increased risk for alcohol­related problems. Alcohol consumption I • Delayed linkage to HIV medical levels above these limits are considered risky drinking. than people in the general population. Specifically, a higher proportion drink care (Samet et al. 1998); risky amounts1 or have an alcohol use • Increase in risky sexual behaviors JEFFREY H. SAMET, M.D., M.A., disorder (i.e., abuse or dependence) (Kalichman et al. 2002; Metsch M.P.H., is a professor in the Clinical (Conigliaro et al. 2003; Galvan et al. et al. 2009); Addiction Research and Education 2002; Lefevre et al. 1995; Samet et (CARE) Unit, Section of General • Increased transmission of sexually Internal Medicine, Department of al. 2003a,b, 2004). Risky alcohol transmitted infections (Kalichman Medicine, Boston University School of use in HIV­infected people has been et al. 2000); and Medicine, and in the Department of associated with the following range Social and Behavioral Sciences, Boston of adverse effects: • Progression of HIV disease University School of Public Health, (Conigliaro et al. 2003; Miguez both in Boston, Massachusetts. • Reduced adherence to medication et al. 2003; Samet et al. 2007). LEXANDER Y. WALLEY, M.D., M.SC., regimens for treatment of HIV A Given the spectrum of problems is an assistant professor in the CARE infection (Chander et al. 2006; associated with such alcohol use among Unit, Section of General Internal Conen et al. 2009; Cook et al. HIV­infected patients, one important Medicine, Department of Medicine, 2001; Golin et al. 2002; Halkitis avenue to improving the health of this Boston University School of Medicine, et al. 2003; Samet et al. 2004); population is to develop interventions Boston, Massachusetts.

Vol. 33, No. 3, 2010 267 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 268

that target alcohol use and its associated (e.g., cognitive–behavioral coping • HIV risk behaviors; consequences. Accordingly, interventions skills, motivational enhancement, have been designed to both decrease 12­step facilitation) can be effective • Acquisition of sexually transmitted alcohol consumption and address the (Project MATCH Research Group infections; and specific adverse health consequences. 1997). In addition, several medica­ The concept that negative conse­ tions (i.e., disulfiram, naltrexone, and • Alcohol use. quences of alcohol use can be reduced acamprosate) are approved for the in patients with HIV infection is treatment of alcohol dependence, and To be included in the review, the based on research demonstrating the other medications (e.g., topiramate) studies had to report alcohol­specific impact of clinical interventions on are being further evaluated (Anton et outcomes. Beyond that, the studies alcohol consumption and associated al. 2006; Garbutt et al. 2005; Kranzler were classified into three categories of negative consequences in patients and Van Kirk 2001; Olmsted and specificity. The most specific category without HIV infection (Institute of Kockler 2008; Rubio et al. 2001). comprised clinical trials that included Medicine 1990; Kristenson et al. 1983). Given the strong evidence that only HIV­infected people with past Alcohol research over the past three alcohol consumption is an important or current unhealthy alcohol use. The decades has demonstrated that behav­ health issue for many people with second category comprised clinical ioral interventions can be effective, HIV infection, efforts to potentially trials that included only HIV­infected with benefits varying based on setting, ameliorate these problems by addressing people but in which not all of the severity of alcohol problems, and alcohol use are of great interest. The participants exhibited unhealthy alcohol patient characteristics. For example, studies in non–HIV­infected people use. For a study to be included in this meta­analyses of randomized controlled reviewed above suggest that interven­ category, at least 10 percent of partic­ trials (RCTs)2 of interventions to tions among HIV­infected people with ipants had to report current alcohol reduce risky alcohol use demonstrated alcohol problems could be beneficial. use. The third category of studies decreased drinking for patients in pri­ However, the wide range of results comprised trials that were aimed at mary care settings (Beich et al. 2003; in these intervention studies based on preventing alcohol use and sexual Kaner et al. 2007). However, no such setting and disease severity argues for behaviors that put people at risk of effects were found in meta­analyses of the need to carefully assess efforts to HIV infection among alcohol­using interventions delivered in hospital mitigate alcohol’s deleterious impact on people. Although these studies did settings (Emmen et al. 2004), possi­ health in HIV­infected patients. As not include HIV­infected participants

bly because inpatients typically have an important step in this direction, this or did not report the HIV status of

greater severity of alcohol problems article summarizes the findings of a the participants, they were reviewed because they may inform future research (i.e., most are alcohol dependent) (Saitz review of the clinical trial literature on on people at risk of HIV transmission et al. 2007, 2008). Several high­quality interventions addressing alcohol con­ in the setting of alcohol use. RCTs of brief interventions delivered sumption and its consequences among Initially, the review intended to in emergency departments also detected HIV­infected patients. After describing include only RCTs. However, very no or limited benefit (D’Onofrio and the design of the literature search and few studies were identified that met Degutis 2002; Daeppen et al. 2007; evaluation, the article reviews the findings Longabaugh et al. 2001; Monti et al. this criterion in the first two categories. of the studies identified and discusses Therefore, the search was expanded 1999). The influence of the patient’s the implications of those findings. consumption levels also was demon­ to include nonrandomized and non­ strated in several studies. For example, controlled clinical intervention trials in two separate RCTs in the primary­ Design of the Literature in categories 1 and 2. care setting (Fleming et al. 1997; Review To identify relevant studies, the Ockene et al. 1999), where patients literature database MEDLINE was were seeking medical care but not The literature review sought to identify searched through September 30, necessarily for an alcohol problem, clinical trials of interventions among 2009, using the search terms “HIV, implementation of a 5­ to 15­minute HIV­infected people with past or current alcohol, hazardous drinking, risky discussion reduced alcohol consump­ unhealthy alcohol use (i.e., the spectrum drinking, problem drinking, counseling, tion in patients who met the criteria from risky drinking to alcohol dependence brief intervention, 12 step, pharma­ for risky drinking. Studies of such [Saitz 2005]) that reported effects on cotherapy, naltrexone, acamprosate, brief interventions among patients who any of the following outcomes: disulfiram, topiramate, and clinical met the criteria for alcohol dependence, trial.” For all articles identified using however, have shown no benefit • HIV disease progression; this approach, the reference lists also (Kaner et al. 2007; Whitlock et al. 2004; Wutzke et al. 2002). • Receipt of HIV treatment; 2 RCTs are clinical studies in which patients randomly are assigned to either one or more groups receiving the treatment For alcohol­dependent patients, under investigation or to a control group receiving no treatment more extensive behavioral interventions • HIV medication adherence; or a treatment of known efficacy.

268 Alcohol Research & Health 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 269

Interventions for HIV­Infected Risky Drinkers

were scanned, as were related articles the four medications recommended findings is limited by the fact that there identified by the search engine for by NIAAA (2007) for the treatment was differential loss to follow­up—that the MEDLINE data base to look for of alcohol dependence (i.e., disulfiram, is, the analyses included only 81 percent additional studies. Reference lists for naltrexone, acamprosate, and topira­ of participants randomized to the articles that were closely related, but mate) have been conducted in HIV­ intervention group and 90 percent of did not meet the criteria, also were infected patients. subjects randomized to the control group. reviewed. Finally, articles referenced in Thus, one cannot exclude the possibility relevant review articles were examined. that particularly in the intervention group, Titles of all articles were reviewed to Clinical Trials Among HIV­ participants with worse outcomes were determine if the articles met the selec­ Infected People With Past not included in the analysis. tion criteria. If the nature of the study or Current Unhealthy could not be discerned through the Alcohol Use Aharonovich and Colleagues (2006) title, the abstract and/or full text of Study. In this pilot study, 31 HIV­ the article was retrieved and reviewed. Velasquez and Colleagues (2009) infected primary­care patients with For all studies that met the criteria Study. These investigators conducted heavy alcohol use received one session for one of the three categories, infor­ an RCT among 253 HIV­infected men of MI from a trained counselor, fol­ mation on the setting, study design, who had had sex with men in the pre­ lowed by daily telephone­based interac­ methodological quality, type of inter­ vious 3 months and who scored more tive voice response (IVR) assessments vention, outcomes reported, period than eight points on the AUDIT ques­ of drinking amounts and graphic feed­ of follow­up, and results was extracted. tionnaire (Babor et al. 2001). The back of changes in drinking at 30 and The following sections summarize the intervention group received four manual­ 60 days. This intervention resulted in a findings of these analyses. They are guided individual sessions and four decrease in the number of drinks per presented as a descriptive narrative manual­guided peer education and day at 30 and 60 days (from 3.2 drinks synthesis because studies were too support group sessions that utilized per day at baseline to 1.7 drinks at 30 few and heterogeneous to perform motivational interviewing (MI) coun­ days and 1.2 drinks at 60 days). The a standard meta­analysis. seling strategies (Miller and Rollnick IVR system was utilized; 77 percent of 2002) to guide participants through all possible daily calls were completed at 30 days. However, these improve­ Results of the Literature the stages of change of Prochaska and ments can not be attributed to the Review DiClemente’s Trans­Theoretical Model3 (Prochaska and DiClemente 1982). In intervention with confidence because The search strategy described above contrast, the control group received there was no control group. identified 241 potentially relevant studies educational materials on HIV and that were evaluated further. Of these, alcohol, referral information, and advice Parsons and Colleagues (2007) four studies including a total of 578 to stop or cut back on their alcohol Study. These investigators conducted patients (Aharonovich et al. 2006; use. At the 12­month follow­up, the an RCT among 143 HIV­infected peo­ Parsons et al. 2007; Samet et al. 2005; investigators determined some benefits ple with “hazardous drinking” (defined Velasquez et al. 2009) met the selection of the intervention on some of the as more than 16 standard drinks per criteria for the first category (see table measures evaluated. For example, the week for men or more than 12 standard 1). Another five clinical trials that control group had 1.4 times the num­ drinks per week for women), assessing included 1,311 patients (Gilbert et al. ber of drinks per 30 days and 1.5 times treatment effects on HIV medication 2008; Naar­King et al. 2006, 2008; the number of heavy­drinking days per adherence and alcohol outcomes. The Rotheram­Borus et al. 2001, 2009; 30 days compared with the interven­ intervention involved eight 1­hour Sorensen et al. 2003) fell into the second tion group. For other measures (e.g., individual sessions of MI and cognitive category. In addition, two informative having anal sex without a condom, behavioral skills training over 3 months and was compared with a time­ and studies of interventions among people number of drinking days, or number 4 at high­risk for HIV reported outcomes of days on which both drinking and content­equivalent control. Over the

specific to alcohol use (Kalichman et al. sex occurred), however, no significant 3 The transtheoretical model (TTM) is a health behavior theory 2008; Morgenstern et al. 2007). All of difference existed between the two that assesses the individual’s readiness to change a particular these studies are reviewed below. Some groups. Only when the analysis of behavior in order to facilitate the desired behavior change. The stages of change are: precontemplation, contemplation, other studies that involved alcohol­ same­day drinking and sex was restricted preparation, action, and maintenance. using, HIV­infected patients, but were to participants who had shown this 4 With a time­ and content­equivalent control group, participants excluded from this discussion because behavior at baseline, did those in the in that group spend the same amount of time with a health care of serious design or methodological control group have significantly (i.e., provider/therapist as the intervention group, and they receive the limitations, are listed in Table 2 because 2.19 times) more days on which drink­ same type of information. The only difference between the inter­ vention and control groups is the method used to deliver the they may inform additional research. ing and sex occurred than the interven­ information, allowing researchers to determine whether one Interestingly, no controlled trials of tion group. The interpretation of these approach is more effective than the other.

Vol. 33, No. 3, 2010 269 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 270

follow­up period (3 and 6 months), intervention did improve medication Samet and Colleagues (2005) Study. both groups exhibited substantial adherence, number of virus particles This RCT included 151 HIV­infected improvement for both total alcohol detectable in the blood (i.e., viral load), patients on antiretroviral therapy (ART) 5 drinks over 14 days or drinks per and CD4 cell counts at 3 months. who had a history of alcohol problems.

drinking day, although no significant These statistically significant improve­ The participants received either four ments were not sustained at 6 months. differences existed between the inter­ nurse­delivered, 30­ to 60­minute sessions vention and the control group. However, 5 CD4 cells are a type of white blood cell that is the main target of the focusing on HIV medication adherence HIV virus; accordingly, levels of these cells in the blood decline with compared with the control group, the progressing HIV infection and are a marker for disease progression. and alcohol counseling, both in a clinic

Table 1 Studies Identified During a Literature Search on Interventions to Decrease Alcohol Use and Related Behaviors among

HIV­Infected People and Alcohol Users at High Risk for Infection

Study Population/Setting Design Outcomes/Results Comments

Category 1: Clinical trials among HIV­infected people with past or current unhealthy alcohol use

Velasquez Population: 253 Intervention: Randomized Alcohol use: Control group Alcohol measures:

et al. 2009 HIV­infected men Controlled Trial (RCT) of four had 1.38 times the number AUDIT, 90­day timeline

who had sex with men sessions of motivational of drinks per 30 days and follow­back (TLFB) at

(MSM) in the previous 3 interviewing (MI)­based 1.50 times the number of follow­up assessments.

months and an AUDIT individual counseling and heavy drinking days per 30

score of more than 8. four sessions of transtheoretical days compared with the Differential loss to follow­

Setting: Recruited from model–based peer­group intervention group. up at 12 months (34%

HIV organizations, education/support. Sex risk: No significant in intervention group and

advertising, and Control: HIV and alcohol effect was demonstrated 26% in control group). Only

social venues educational materials, for anal sex without a 95 of 118 (81%) of the

between 1999 and resource referrals, and advice condom or number of days intervention group and

2003. to stop or reduce drinking. on which drinking and sex 121 of 135 (90%) of the

Assessment: Baseline, 3, occurred. control group were included

6, 9, and 12 months. in the analyses.

Aharonovich Population: 31 Intervention: 30­minute MI Drinks per day: Using 7­day Alcohol measures:

et al. 2006 HIV­infected men and session on reducing alcohol recall, mean drinks per Quantity and frequency in

women engaged in use by counselor trained in day was 3.2 at baseline, past week and past month.

HIV primary care. MI plus an automated daily 1.7 at 30 days, and 1.2 at

Alcohol use: All had telephone self­monitoring 60 days. Mean highest Qualitative assessment of

four or more drinks interactive voice response drinks per day was 8.4, 4.1, the program demonstrated

at least once in the (IVR) system with graphical and 3.8, respectively. satisfaction with daily

past 30 days, 55% feedback at 30­day follow­up Cocaine use: Decreased calling and the feedback

had five or more meetings. significantly at 60 days. graph.

drinks in the last Control: No control group.

week. Assessment: Baseline, 30, Not a randomized

Setting: HIV primary 60, and 90 days. controlled trial.

care clinic.

Parsons Population: 143 Intervention: RCT of eight Alcohol use: No significant Alcohol measure:

et al. 2007 HIV­infected subjects 60­minute MI plus cognitive effects on total drinks over Self­report 14­day TLFB

on antiretroviral therapy behavioral skills training (CBST) 14 days or drinks per drinking to calculate total drinks

(ART) with hazardous session by Masters­level day. Decreases in both and drinks per drinking day.

drinking (more than 16 counselors. groups from baseline to 3 Adherence measures:

drinks per week for Control: Eight 60­minute and 6 months for these two Self­report dose

men, more than 12 time and content­equivalent drinking outcomes. adherence = number of

drinks per week for education sessions by health Medication adherence: doses taken/number of

women) recruited educators. Significant improvement in doses scheduled over 14

through HIV clinics and dose and day adherence days. Self­report day

advertising from 2002 All sessions delivered at 3 months, but difference adherence = number of

to 2005. individually in private office not retained at 6 months. days with perfect

Setting: Behavioral over 12 weeks. HIV viral load/CD4 cell adherence/14 days.

research center. Assessment: Baseline, 3 count: Significant

and 6 months. improvement at 3 months

but not at 6 months.

270 Alcohol Research & Health 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 271

Interventions for HIV­Infected Risky Drinkers

and at home, or no intervention. Clinical Trials Among partners (Wong et al. 2008). The par­ The study found no significant differ­ HIV­Infected People of ticipants received either 15 90­minute ences between groups upon examina­ Whom at Least 10 Percent individual sessions of cognitive–behavioral tion of the following outcomes: 3­day Currently Use Alcohol therapy (CBT) delivered over 15 medication adherence, 30­day adher­ months or usual care. The subanalysis ence, CD4 cell count, viral load, drinks Five studies identified in the literature by Rotheram­Borus and colleagues per day, percent reporting drinking, review fell into this category, and only (2009) was limited to 270 HIV­infected or percent reporting hazardous drink­ one of these (Rotheram­Borus et al. participants who were homeless or ing. Study limitations were that not 2009) demonstrated significant treat­ without stable housing. In this group, all participants were non­adherent to ment effects on alcohol use (see table the intervention was found to reduce their HIV medication at baseline and 1). This study was a subanalysis of a alcohol or marijuana use from 36 to 28 a substantial percentage were not in parent RCT among 936 HIV­infected days in the prior 90 days, whereas in the risky­drinking range of unhealthy people who were sexually active with­ the control group the frequency of alcohol use, the group most amenable out a condom with at least one HIV­ alcohol or marijuana use was unchanged to brief interventions. negative partner or two HIV­infected at 35 of the last 90 days. However, this

Table 1 con’t

Study Population/Setting Design Outcomes/Results Comments

Samet Population: 151 Intervention: RCT of four 15­ Alcohol use: No significant Alcohol measures:

et al. 2005 HIV­infected patients to 60­minute sessions over effects on drinks per day, Self­report 30­day alcohol

on ART, with current 3 months with MI­trained percent reporting any use from the Addiction

or lifetime alcohol nurse who (1) addressed drinking, percent reporting Severity Index.

problems, determined alcohol problems, (2) educated hazardous drinking. Adherence measures:

by two or more about ART efficacy, and Medication adherence: Self­reported AIDS Clinical

positive responses on (3) delivered tailored No significant effects on Trial Group scale with

CAGE questionnaire adherence advice including a 3­day or 30­day adherence. 100% and 95% or more

or clinical diagnosis of reminder watch and a home visit. HIV viral load/ CD4 cell thresholds at 3­day and

alcohol disorder Control: Standard care count: No significant 30­day adherence,

recruited from 1997 Assessment: Baseline, 6, effects on mean CD4 cell respectively.

to 2000. and 12 months. count or mean log HIV RNA.

Setting: Hospital

(patients receiving

HIV medical care).

Category 2: Clinical trials among HIV­infected people of whom at least 10% have current alcohol use

Rotheram­ Population: 270 Intervention: RCT of 15 Alcohol or marijuana use Subanalysis of a clinical

Borus et al. HIV­infected people 90­minute individual counseling in the last 3 months: At 25 trial (Wong et al. 2008):

2009 sexually active without sessions, organized in three months, the intervention 5% used alcohol/

a condom with at least modules (“Coping” at 0–5 group reduced its use marijuana in the parent

one HIV­negative months, “Act Safe” at 5–10 from 36 to 28 days in the study. Proportion of

partner or two months, and “Stay Healthy” prior 90 days, whereas alcohol users at baseline

HIV­infected partners at 10–15 months). the control group was not presented in this study.

who were marginally Control: No intervention, only unchanged at 35 days of

housed and had four assessments the last 90. Parent study reported only

or more assessments; Assessment: Baseline, 15, 20, transmission act outcomes

recruited from 2000 and 25 months. Number of HIV negative and demonstrated an

to 2002. partners and risky sexual effect that was not

Alcohol use: Mean acts also was reduced. maintained at 25 months.

number of days using

alcohol or marijuana Imbalance in transmission

in the last 90 was 37. risk acts at baseline

Setting: Recruited from resulted in ineffective

community agencies, randomization, thus

medical clinics, and propensity scores were

advertisements. used to adjust for

imbalances.

Vol. 33, No. 3, 2010 271 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 272

Table 1 con’t

Study Population/Setting Design Outcomes/Results Comments

Naar­King Population: 65 Intervention: RCT of four No significant effects at Alcohol and drug

et al. 2006, HIV­infected patients, 60­minute sessions of 9­month follow­up. measures: Timeline

2008 aged 16–25 regardless motivational enhancement. Alcohol use: Borderline follow­back, though time

of alcohol use or risk Therapy focused on two of significant reduction in window is not stated.

behaviors. three areas: substance use, number of drinks in the Sex risk measure:

Alcohol use: 77% sexual risk, or medication week containing the Total number of

lifetime, 39% had used adherence over 10 weeks. maximum number of drinks unprotected intercourse

alcohol in last 30 days Control: Wait list and standard (–9.65 vs. –1.3) at 3 months acts without a condom. n at study entry. care. ( = 51).

Setting: Adolescent Assessment: Baseline, 3, 6, Marijuana use: Borderline Note: 3­month outcomes

HIV care clinic within and 9 months. significant reduction in on 51 subjects were

a tertiary care children’s number of times marijuana published in 2006 and 6­ n hospital. was used ( = 65). and 9­month outcomes

Sexual risk: Borderline on 65 subjects published

significant reduction in total in 2008.

number of intercourse acts

without a condom at 6 n months ( = 65).

HIV viral load: Significant

reduction in log viral load at n 6 months ( = 65).

Gilbert et al. Population: 476 Intervention: RCT of two Alcohol use: No significant Alcohol measures:

2008 patients with alcohol sessions of tailored risk­ effects on any risky drinking Self­reported NIAAA risky

risk (38%), defined reduction counseling at study or number of drinks per drinking over 3 months.

as exceeding NIAAA entry and 3 months using a week. Drug use measures:

safe drinking limits MI “Video Doctor” via laptop Drug use: Significantly Self­reported drug use

or drug risk (42%), computer, printed educational decreased 30­day illicit over 30 days included any

or sex risk (60%), worksheet, and delivery of drug use at 3 and 6 months cocaine, methamphetamine,

were recruited a cueing sheet on reported and fewer days of illicit or heroin or 3 or more days

between 2003 and risks to clinic care providers. drug use at 6 months. of barbiturates, prescription

2006. Control: Standard care. Sex risk: Significantly opiates, hallucinogens,

Setting: Outpatient Assessment: Baseline, 3, decreased 3­month inhalants, or methylene­

HIV clinics. and 6 months. unprotected sex at 3 and dioxymethamphetamine

6 months and fewer casual (MDMA).

sex partners at 6 months.

No effects on condom use.

Sorensen Population: 190 Intervention: RCT of 12 No outcomes showed Summary/index

et al. 2003 HIV­infected patients months of case management significant change between score is shown without

with substance by certified substance study groups at any time explanation of the raw

dependence; recruited counselors in the community points, except decreased measure.

from inpatient medical with caseload of 1:20 sex risk index.

wards, detoxification Control: Single brief contact Outcomes measured:

clinic, and the with education about Addiction severity index

emergency department reducing HIV risk, information composite scores, AIDS risk

from 1994 to 1996. on HIV services, referrals to assessment scores, Beck

Alcohol use: 61% in addiction treatment, social depression inventory, health

the last 30 days. services. status questionnaire, and

Setting: Public general Assessment: Baseline, 6, support evaluation list.

hospital. 12, and 18 months.

272 Alcohol Research & Health 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 273

Interventions for HIV­Infected Risky Drinkers

Table 1 con’t

Study Population/Setting Design Outcomes/Results Comments

Rotheram­ Population: 310 Intervention: 23 group Alcohol/marijuana use: Sequential assignment of

Borus et al. HIV­infected patients sessions of two modules 63% for attendees vs. 67% 15 youths to intervention

2001 (age 13–24) from nine (“Stay Healthy” and “Act for control vs. 84% for versus control groups (not

adolescent clinics Safe”). nonattendees at 15 months. randomized).

recruited from 1994 Control: Standard care.

to 1996. Eligible for receiving the The reported comparisons

Alcohol use: 67% intervention at the were attendees versus

nonabstinent at conclusion of the study. non­attendees versus

baseline. Assessment: Baseline, 9, control subjects. No

Setting: Adolescent and 15 months. intention­to­treat analysis

clinics. was reported.

Differential loss to

follow­up. No alcohol­

specific outcome was

reported.

Category 3: Randomized controlled trials among alcohol users at high risk for HIV infection

Morgenstern Population: 198 MSM Intervention: 12 sessions of Drinks per day: At 12 Alcohol measures:

et al. 2007 with current alcohol combined MI and coping weeks, the MI group had CIDI at baseline.

user disorder. skills training (MI+CBT) over greater decreases in drinks TLFB and short n Alcohol use: 88% with 12 weeks ( = 47). per day than the MI+CBT inventory of problems

alcohol dependence. Control: Four sessions of group. This difference was at followup. n Mean drinks per MI over 12 weeks ( = 42). not sustained at 12 months.

drinking day was 10.4. Non–help­seeking (NHS) Both intervention groups Potential subjects with n Setting: Subjects control group ( = 109). had greater decreases then drug use more severe

recruited through Assessment: Baseline, 12 the NHS group, but the than alcohol use disorder

advertisements in gay weeks, and 12 months. NHS group also had were excluded. Less than

media, internet chat substantial decreases in 10% HIV infected.

rooms, outreach to drinking.

gay bars and clubs. Subjects lost to follow­up

not included in the

analysis.

Kalichman Population: 342 men Intervention: 3­hour skills­ The following behaviors Alcohol measures: AUDIT,

et al. 2008 and women who drink based HIV–alcohol risk­ were improved significantly frequency of drinking

in South African reduction group session. at 3 months among the before sex in previous

shebeens. Control: 1­hour HIV­alcohol intervention group: month. Change in AUDIT ● Setting: Informal information group session. alcohol use before sex scores not reported. ● alcohol establishments Assessment: Baseline, 3, unprotected intercourse ● (shebeens). and 6 months. percent of sex with condoms 7% HIV infected in ● number of sex partners. intervention group. 4% HIV

infected in control group.

Intervention effects were

significantly stronger in

those drinking less and

dissipated at 6 months.

Vol. 33, No. 3, 2010 273 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 274

study had substantial methodological completed four follow­ups and were • In a preliminary analysis of 3­ limitations, some of which pertain to homeless or without stable housing. month outcomes among 51 subjects the parent study. For example, in the Finally, the outcome was alcohol or randomized to four 1­hour motiva­ parent study, random assignment of marijuana use in the last 3 months with tional enhancement therapy sessions participants to the groups resulted in no alcohol­specific results provided. in an adolescent clinic, Naar­King an imbalance between the groups with The four other studies in this cate­ and colleagues (2006) observed a respect to baseline HIV risk behaviors gory did not demonstrate any signifi­ trend, but no statistically significant or demographics. Moreover, the sub­ cant effects of the interventions tested reduction, in the number of drinks analysis was limited to participants who on alcohol use: per week during the week with the

Table 2 Studies Identified but not Selected for the Literature Review

Citation Population Reason Excluded

Golin et al. 2003 140 HIV­infected patients. No data on the proportion of drinkers at

Setting: Hospital HIV clinic. baseline.

Goujard et al. 2003 326 HIV­infected patients. No specific alcohol outcomes; alcohol group

Setting: Hospital­ and university­based centers. not analyzed independently.

Jones et al. 2003 174 women with AIDS from three U.S. cities No alcohol­specific outcomes reported.

recruited in 1997 from outpatient clinics,

community health centers and agencies, and

participant referrals.

Alcohol use: 32% with history of alcohol.

Setting: Primarily recruited from outpatient clinics,

community health centers, and participant referrals.

Pradier et al. 2003 244 HAART­treated patients. No specific alcohol outcomes; alcohol group

Setting: Hospital not analyzed independently.

Samet et al. 2008 181 Russian men and women who reported any No alcohol­specific outcomes reported.

alcohol or drug dependence and who reported at Although both HIV­infected and alcohol­

least one incidence of unprotected sex in the past dependent patients were included in this

6 months. study, the HIV­infected patients were not the

Setting: Narcology hospitals alcohol­dependent patients.

Sampaio­Sa et al. 2008 107 HIV­infected, antiretroviral­naïve patients at an Alcohol­specific outcomes not reported.

Brazilian HIV clinic for whom antiretrovirals were

indicated were recruited from 2003 to 2004.

45% with alcohol use in the last 3 months.

Simoni et al. 2007 136 HIV­infected men and women. No information on current use; no specific

Setting: Outpatient clinic alcohol outcomes.

Wong et al. 2008 936 HIV­infected from four U.S. cities recruited Alcohol­specific outcomes not reported;

between 2000 and 2002. absolute numbers for outcome not presented.

Setting: Community agencies, AIDS service

organizations, and medical clinics

SOURCES : Golin, C.E.; Earp, J.; Tien, H.C.; et al. A 2­arm, Journalrandomized, of Acquiredcontrolled Immune trialDeficiencyof a motivational Syndromesinterviewing­based intervention to improve adherence to antiretroviral therapy (ART) among patients failing or initiating ART. Journal of Acquired42:42–51, Immune2006; DeficiencyGoujard, SyndromesC.; Bernard, N.; Sohier, N.; et al. Impact of a patient education program on adherence to HIV medication: A randomized clinical trial. AIDS Care 34:191–194, 2003; Jones, D.L.; Ishii, M.; LaPerriere, A.; et al. Influencing medication adherence among women with AIDS. 15:463–474, 2003; Pradier, HIVC.; ClinicalBentz, L.;TrialsSpire, B.; et al. Efficacy of an educational and counseling intervention on adherence to highly active antiretroviral therapy: French prospective controlled study. 4:121–131, 2003; Samet, J.H.; Krupitsky, E.M.; Cheng, D.M.; et al. Mitigating risky sexual behaviors Addictionamong Russian narcology hospital patients: The PREVENT (Partnership to Reduce the Epidemic Via Engagement in Narcology Treatment) randomized controlled trial. 103:1474–1483, 2008; Sampaio­Sa, M.; Page­Shafer, AIDS and BehaviorK.; Bangsberg, D.R.; et al. 100% adherence study: Educational workshops vs. video sessions to improve adherence among ART­naive patients in Salvador, Brazil. 12:S54–S62, 2008; Simoni, J.M.; Pantalone, D.W.; Plummer, M.D.; and Huang, B. A Healthrandomized Psychologycontrolled trial of a peer support intervention targeting antiretroviral medication adherence and depressive symptomatology in HIV­positive men and women. 26:488–495, 2007; Wong, F.L.; Rotheram­Borus, Addiction M.J.; Lightfoot, M.; et al. Effects of behavioral interven­ tion on substance use among people living with HIV: The Healthy Living Project randomized controlled study. 103:1206­1214, 2008.

274 Alcohol Research & Health 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 275

Interventions for HIV­Infected Risky Drinkers

maximum number of drinks. diagnosed with alcohol abuse or depen­ clinical intervention studies focusing Moreover, in the final analysis of the dence but were seeking to moderate exclusively on HIV­infected patients study, which included 65 subjects, their alcohol use. The investigators with current or past unhealthy alcohol 39 percent of whom used alcohol, compared the effects of 12 weekly MI use; five other clinical trials included and this difference was not sustained at sessions augmented with CBT with 4 documented the alcohol use of some of 6 or 9 months (Naar­King et al. 2008). sessions of MI alone. Unexpectedly, the their HIV­infected participants. Overall, investigators found that the nonaug­ the current state of research strongly • Gilbert and colleagues (2008) ran­ mented MI group had less drinking suggests that although the problems domized 476 HIV­infected patients, and fewer alcohol­related drinking related to alcohol in HIV­infected 38 percent of whom reported risky problems than the MI­plus­CBT group people are abundant, effective interven­ drinking, to an MI­based “Video during the 12 weeks of the interven­ tions are few and new ones are urgently Doctor” intervention via laptop tion and that there were no significant needed. Hence, addressing alcohol computer or a control group receiving differences at 12­month follow­up. problems remains an important issue usual care. The intervention Thus, the addition of CBT to MI tech­ in HIV research. resulted in decreased 30­day illicit niques provided no additional benefit Not only are studies among alcohol­ drug use, lower mean number of regarding alcohol outcomes and poten­ abusing, HIV­infected patients scarce, drug use days, and a modest reduc­ tially even diminished effects in this but the existing studies also yielded tion of unprotected sex at 3 and 6 population. Subgroup analyses demon­ mixed results. Two of the four studies months. However, no differences in strated that the detrimental effect of that specifically targeted HIV­infected alcohol use existed between the augmentation occurred particularly in people with alcohol problems showed intervention and control groups. participants with a concomitant drug improvement in drinking outcomes. use disorder. Velasquez and colleagues (2009) • Sorensen and colleagues (2003) Another RCT (Kalichman et al. demonstrated reduced drinking levels randomly assigned HIV­infected 2008) compared a 3­hour, skills­based over 12 months after an intervention patients with drug dependence, 61 HIV and alcohol risk reduction group that included both MI and peer sup­ percent of whom reported current session with a 1­hour HIV/alcohol port. The intervention was particularly alcohol use, to 1 year of continuous information group session among 342 strong in reducing same­day drinking case management or to a brief con­ South Africans frequenting drinking and sex, which compels further research tact (i.e., one HIV risk education establishments. In this study, the on interventions targeting alcohol use session and printed information). extended session resulted in decreases at the time of HIV risk behaviors No differences were noted in in alcohol use before sex and unpro­ (Velasquez et al. 2009). Although the alcohol outcomes at 6, 12, or 18 tected intercourse at 3 month but intervention types used in the study months. not at 6 month follow­up. Moreover, only were shown to be effective in a intervention effects were stronger in sample of men who have sex with • A study among HIV­infected participants drinking less at baseline. men, they warrant study among other youths compared the effects of 23 populations. In the other study, 2­hour group sessions and usual Aharanovich and colleagues (2006) care on risk behaviors (Rotheram­ Discussion demonstrated the feasibility of ongoing Borus et al. 2001). The investigators telephone­based interactive voice Given the high prevalence of unhealthy found no changes from baseline on response and graphic feedback, which alcohol use among HIV­infected people a measure reflecting alcohol and should inspire the inclusion of auto­ and its associated adverse health conse­ marijuana use and no difference mated, tailored, ongoing intervention quences, development of clinical and between the intervention and boosting as part of behavioral inter­ public health interventions that seek to control groups. ventions. It is important to note, address alcohol use and improve health however, that both these studies had outcomes in this population is a priority. methodological limitations (e.g., sub­ In recognition of this, NIAAA, as early stantial or differential loss to follow­ RCTs Among Alcohol as 1996, issued a request for applications up, incomplete assessments) and their Users at High­Risk for entitled “Developing Alcohol­Related findings therefore are not definitive. HIV Infection HIV Preventive Interventions (AA–97 Nevertheless, they provide some –03).” Since then, several studies have guidance for future more rigorous Two informative RCTs have been con­ been published that describe clinical clinical trials. ducted among alcohol drinkers at high outcomes of interventions in this pop­ The other two clinical trials (Parsons risk for HIV infection. Morgenstern ulation. However, as this article has et al. 2007; Samet et al. 2005) among and colleagues (2007) performed a study demonstrated, the literature on this alcohol­abusing HIV­infected people with 198 high­risk, HIV­negative men important topic still is not extensive. attempted to improve ART adherence. who had sex with men and who were A literature search revealed only four This is an appropriate target of alcohol

Vol. 33, No. 3, 2010 275 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 276

intervention studies in this popula­ given that pharmacotherapy plays a decreasing alcohol use in order to suc­ tion because medication adherence is major role in addressing the AIDS cessfully achieve behavior change is cru­ of utmost importance for achieving epidemic by improving outcomes cial for developing future interventions. good HIV disease outcomes, and of HIV­infected subjects. Moreover, One interesting development noted alcohol­using patients have been doc­ some preclinical research has demon­ in the studies reviewed here was the umented to exhibit suboptimal ART strated that naltrexone, an effective use of new technology (e.g., interac­ adherence (Braithwaite et al. 2005; medication for alcohol dependence, tive voice­response systems) in two of Chander et al. 2006; Conen et al. inhibits alcohol­mediated enhance­ the studies (Aharonovich et al. 2006; 2009; Samet et al. 2004). The results ment of HIV infection (Wang et al. Gilbert et al. 2008). These approaches of both of these trials are discouraging, 2006) and may potentiate the anti­ to delivering a behavioral intervention however, because although they HIV effects of antiretroviral medica­ merit further exploration because explicitly addressed both alcohol use tions (Gekker et al. 2001). Therefore, they have the potential for providing and medication adherence, one study testing the effectiveness of naltrexone scalable, ongoing delivery of tailored (Samet et al. 2005) found no impact and other medications in alcohol­ automated messages that may boost a on adherence, alcohol consumption, dependent HIV­infected patients is an more intensive directly administered or any HIV outcome, and the other important current research direction. intervention. (Parsons et al. 2007) only detected Two of the studies reviewed here When assessing the relevance of the short­lived improvements (i.e., they that included HIV­infected patients studies reviewed here, particularly were evident at 3 months, but not at among whom at least 10 percent those conducted among HIV­infected 6 months). Thus, these two high­quality currently used alcohol, targeted risky patients with past or current unhealthy studies suggest that achieving clinically sexual behaviors rather than alcohol alcohol use, it is important to consider beneficial outcomes in HIV­infected consumption. Assessing treatment the methodological quality of the work people with alcohol problems is more effects on sex risk factors is appropriate (i.e., the potential for bias, design difficult than has been the case with for studies among HIV­infected limitations, and outcome measures). populations of HIV­infected without drinkers because several studies have The report by Velasquez and col­ diagnosed unhealthy alcohol use demonstrated an association between leagues (2009) is the only controlled (Amico et al. 2006; Simoni et al. alcohol use and risky sex (Purcell et study demonstrating a sustained clini­ 2006). Among the latter group, RCTs al. 2001; Stein et al. 2009). In both cally significant treatment effect on to improve adherence that used inter­ the study by Gilbert and colleagues an alcohol­specific outcome, making ventions with a range of intensities (2008) and the study by Naar­King publication bias (i.e., the preferential did reveal improvements in adherence and colleagues (2006, 2008), sex risk publication of studies that find signif­ which were sustained for up to 12 behaviors were decreased in the group icant differences) unlikely. months, as well as in HIV viral load randomized to the intervention at 3 Regarding their design, most, but and CD4 counts (Tuldra et al. 2000). and 6 months, but there were no or not all, of these studies met important The difficulty of achieving positive only transient effects on alcohol use. design criteria, such as random allo­ benefits (e.g., improved ART adher­ These findings suggest that behavioral cation of participants to treatment ence) through interventions among interventions which are not specifically groups and intention­to­treat analyses6 HIV­infected people who have alcohol tailored to address alcohol use are in the presentation of results. As with problems also is evidenced by the study unlikely to impact alcohol problems all behavioral intervention studies, by Kalichman and colleagues (2008) in a sustained fashion. keeping participants in the dark about among drinkers who were not infected The dearth of studies focusing on which treatment they receive (i.e., with HIV. The findings of that study alcohol consumption among HIV­ blinding of participants to their treat­ suggest that, as in brief intervention infected people is understandable. ment) is not possible. However, both studies, intervention effectiveness Although the spectrum of unhealthy Parsons and colleagues (2007) and varies by severity of alcohol use, with alcohol use ranging from risky use Gilbert and colleagues (2008) utilized less improvement noted in dependent to alcohol dependence occurs in this time­ and content­equivalent controls than in nondependent drinkers. Thus, population, other pressing health to allow for the detection of effects levels of alcohol consumption, alcohol concerns (e.g., ART adherence, risky use disorder severity, and alcohol­ sexual behaviors, or engagement in related consequences are important 6 An intention­to­treat analysis is based on the initial treatment HIV care) appropriately become the intent, not on the treatment actually administered. Thus, every covariates to be assessed and reported main focus of clinical trials that also participant who begins the treatment is considered to be part of in HIV intervention studies. may address alcohol consumption in the trial, whether they finish it or not. This is done to avoid vari­ ous misleading artifacts that can arise in a study. For example, if A notable finding of this literature their intervention arms. Developing participants who have a more serious problem tend to drop out review was that as of 2009, no study interventions that target a specific at a higher rate, even an ineffective treatment may appear to of pharmacotherapy for alcohol behavior (e.g., sex) at the time of provide benefits if one only compares the condition before and after the treatment among participants who finish the treatment dependence in HIV­infected patients alcohol use is a worthy pursuit, and and ignores participants who were enrolled originally but did not had been published. This is surprising understanding the importance of finish the treatment.

276 Alcohol Research & Health 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 277

Interventions for HIV­Infected Risky Drinkers

specific to the counseling method challenges when attempting to in Primary Care. 2nd ed. Geneva: World Health studied. improve clinical outcomes? Organization, 2001. The outcome measures reported BEICH, A.; THORSEN, T.; AND ROLLNICK, S. were not consistent across studies and • How can individual, network, or Screening in brief intervention trials targeting not always meaningful, limiting the community interventions in people excessive drinkers in general practice: Systematic review and meta­analysis. BMJ 327:536–542, comparability of study outcomes. For with multiple overlapping prob­ 2003. PMID: 12958114 example, Naar­King and colleagues lems, including alcohol use, opti­ (2006) used an alcohol­specific mea­ mally reduce unhealthy behaviors? BRAITHWAITE, R.S.; MCGINNIS, K.A.; sure—the number of drinks per week CONIGLIARO, J.; ET AL. A temporal and dose­ response association between alcohol consumption during the week with the maximum • How might combined pharma­ and medication adherence among veterans in care. number of drinks at 3 months—that cotherapy and behavioral therapy be Alcoholism: Clinical and Experimental Research is not widely used and of question­ utilized to address the spectrum of 29:1190–1197, 2005. PMID: 16046874

able clinical meaning. Sorensen and clinical consequences that accom­ CHANDER, G.; LAU, B.; AND MOORE, R.D. colleagues (2003) only report a measure pany heavy alcohol consumption? Hazardous alcohol use: A risk factor for non­adher­ called the Addiction Severity Index ence and lack of suppression in HIV infection. Alcohol Composite Score, without Journal of Acquired Immune Deficiency Syndromes Obtaining answers to these ques­ 43:411–417, 2006. PMID: 17099312 any explanation or reporting of the tions is the key next step in the suc­ individual components, complicating cessful development of clinical and CONEN, A.; FEHR, J.; GLASS, T.R.; ET AL. Self­ judgment of its clinical meaning. reported alcohol consumption and its association public health interventions to mitigate with adherence and outcome of antiretroviral ther­ Finally, Samet and colleagues (2005) the adverse outcomes from alcohol focused on ART adherence as an out­ apy in the Swiss HIV Cohort Study. Antiviral use in HIV­infected patients. ■ Therapy 14:349–357, 2009. PMID: 19474469 come, yet this study may underesti­ mate the effectiveness of the interven­ CONIGLIARO, J.; GORDON, A.J.; MCGINNIS, K.A.; tion because the criteria for eligibility ET AL. How harmful is hazardous alcohol use and Acknowledgements abuse in HIV infection: Do health care providers to participate in the study did not know who is at risk? Journal of Acquired Immune exclude patients with already good The authors acknowledge Victoria Deficiency Syndromes 33:521–525, 2003. PMID: adherence. Thus, participants with Churchill, M.P.H, for her thoughtful 12869842 good adherence at baseline provided lit­ assistance in the preparation of this COOK, R.L.; SEREIKA, S.M.; HUNT, S.C.; ET AL. tle opportunity for an intervention to manuscript and Carly Bridden, M.A., Problem drinking and medication adherence reveal a clinically meaningful impact. M.P.H. for reviewing the manuscript. among persons with HIV infection. Journal of In summary, as of 2009 the med­ General Internal Medicine 16:83–88, 2001. PMID: ical literature on clinical trials focused 11251758 on people with HIV infection and Financial Disclosure DAEPPEN, J.B.; GAUME, J.; BADY, P.; ET AL. Brief unhealthy alcohol use is limited (i.e., alcohol intervention and alcohol assessment do not “drops in a bottle”). Few of these The authors declare that they have no influence alcohol use in injured patients treated in studies were able to document competing financial interests. the emergency department: A randomized con­ trolled clinical trial. Addiction 102:1224–1233, improved outcomes, and any effects 2007. PMID: 17565563 observed generally were modest and transitory. Based on these findings References D’ONOFRIO, G., AND DEGUTIS, L.C. Preventive care in the emergency department: Screening and and current knowledge, the following AHARONOVICH, E.; HATZENBUEHLER, M.L.; brief intervention for alcohol problems in the emer­ questions need to be addressed: JOHNSTON, B.; ET AL. A low­cost, sustainable inter­ gency department: A systematic review. Academic vention for drinking reduction in the HIV primary Emergency Medicine 9:627–638, 2002. PMID: • What are the characteristics of care setting. AIDS Care 18:561–568, 2006. PMID: 12045080 16831783 interventions that mitigate the EMMEN, M.J.; SCHIPPERS, G.M.; BLEIJENBERG, G.; health consequences of alcohol AMICO, K.R.; HARMAN, J.J.; AND JOHNSON, B.T. AND WOLLERSHEIM, H. Effectiveness of oppor­ use in HIV­infected people? Efficacy of antiretroviral therapy adherence inter­ tunistic brief interventions for problem drinking in ventions: A research synthesis of trials. 1996 to a general hospital setting: Systematic review. BMJ 2004. Journal of Acquired Immune Deficiency 328:318–322, 2004. PMID: 14729657 • How does the treatment setting Syndromes 41:285–297, 2006. PMID: 16540929 impact the effectiveness of behavioral FLEMING, M.F.; BARRY, K.L.; MANWELL, L.B.; ET interventions? ANTON, R.F.; O’MALLEY, S.S.; CIRAULO, D.A.; ET AL. Brief physician advice for problem alcohol AL. Combined pharmacotherapies and behavioral drinkers: A randomized controlled trial in commu­ interventions for alcohol dependence: The COM­ nity­based primary care practices. JAMA: Journal of • How can technology best be used BINE study: A randomized controlled trial. JAMA: the American Medical Association 277:1039–1045, to extend and enhance intervention Journal of the American Medical Association 1997. PMID: 9091691 effects? 295:2003–2017, 2006. PMID: 16670409 GALVAN, F.H.; BING, E.G.; FLEISHMAN, J.A.; ET BABOR, T.F.; HIGGINS­BIDDLE, J.C.; SAUNDERS, AL. The prevalence of alcohol consumption and • What characteristics of HIV­ J.B.; AND MONTEIRO, M.G. AUDIT the Alcohol heavy drinking among people with HIV in the infected drinkers suggest greater Use Disorders Identification Test: Guidelines for use United States: Results from the HIV Cost and

Vol. 33, No. 3, 2010 277 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 278

Services Utilization Study. Journal of Studies on follow­up of 24­60 months of long­term study with PARSONS, J.T.; GOLUB, S.A.; ROSOF, E.; AND Alcohol 63:179–186, 2002. PMID: 12033694 randomized controls. Alcoholism: Clinical and HOLDER, C. Motivational interviewing and cognitive­ Experimental Research 7:203–209, 1983. PMID: behavioral intervention to improve HIV medication ARBUTT RANZLER ALLEY G , J.C.; K , H.R.; O’M , S.S.; 6135365 adherence among hazardous drinkers: A random­ ET AL. Efficacy and tolerability of long­acting ized controlled trial. Journal of Acquired Immune injectable naltrexone for alcohol dependence: A LEFEVRE, F.; O’LEARY, B.; MORAN, M.; ET AL. Deficiency Syndromes 46:443–450, 2007. PMID: randomized controlled trial. JAMA: Journal of the Alcohol consumption among HIV­infected 18077833 American Medical Association 293:1617–1625, patients. Journal of General Internal Medicine 2005. PMID: 15811981 10:458–460, 1995. PMID: 7472704 PROCHASKA, J.O, AND DICLEMENTE, C.C. Trans­ theoretical therapy—Toward a more integrative GEKKER, G.; LOKENSGARD, J.R.; AND PETERSON, LONGABAUGH, R.; WOOLARD, R.E.; NIRENBERG, model of change. Psychotherapy: Theory, Research P.K. Naltrexone potentiates anti­HIV­1 activity of T.D.; ET AL. Evaluating the effects of a brief moti­ and Practice 19:276­288, 1982. antiretroviral drugs in CD4+ lymphocyte cultures. vational intervention for injured drinkers in the Drug and Alcohol Dependence 64:257–263, 2001. emergency department. Journal of Studies on Alcohol Project Match Research Group. Matching alcoholism PMID: 11672940 62:806–816, 2001. PMID: 11838918 treatments to client heterogeneity: Project MATCH posttreatment drinking outcomes. Journal of Studies GILBERT, P.; CICCARONE, D.; GANSKY, S.A.; ET AL. METSCH, L.R.; BELL, C.; PEREYRA, M.; ET AL. on Alcohol 58:7–29, 1997. PMID: 8979210 Interactive “Video Doctor” counseling reduces drug Hospitalized HIV­infected patients in the era of and sexual risk behaviors among HIV­positive PURCELL, D.W.; PARSONS, J.T.; HALKITIS, P.N.; ET highly active antiretroviral therapy. American patients in diverse outpatient settings. PLoS One AL. Substance use and sexual transmission risk Journal of Public Health 99:1045–1049, 2009. 3:e1988, 2008. PMID: 18431475 behavior of HIV­positive men who have sex with

PMID: 19372520 men. Journal of Substance Abuse 13:185–200, 2001. GOLIN, C.E.; LIU, H.; HAYS, R.D.; ET AL.A PMID: 11547619 prospective study of predictors of adherence to MIGUEZ, M.J.; SHOR­POSNER, G.; MORALES, G.; ET AL. HIV treatment in drug abusers: Impact of combination antiretroviral medication. Journal of ROTHERAM­BORUS, M.J.; DESMOND, K.;

General Internal Medicine 17:756–765, 2002. alcohol use. Addiction Biology 8:33–37, 2003. COMULADA, W.S.; ET AL. Reducing risky sexual PMID: 12390551 PMID: 12745413 behavior and substance use among currently and formerly homeless adults living with HIV. MILLER, W.R., AND ROLLNICK, S. Motivational HALKITIS, P.N.; PARSONS, J.T.; WOLITSKI, R.J.; American Journal of Public Health 99:1100–1107, Interviewing: Preparing People to Change. New AND REMIEN, R.H. Characteristics of HIV 2009. PMID: 18799777 antiretroviral treatments, access and adherence in York: Guilford Press, 2002. an ethnically diverse sample of men who have sex ROTHERAM­BORUS, M.J.; LEE, M.B.; MURPHY, MONTI, P.M.; COLBY, S.M.; BARNETT, N.P.; ET AL. with men. AIDS Care 15:89–102, 2003. PMID: D.A.; ET AL. Efficacy of a preventive intervention Brief intervention for harm reduction with alcohol­ 12655837 for youths living with HIV. American Journal of

positive older adolescents in a hospital emergency Public Health 91:400–405, 2001. PMID: Institute of Medicine. Broadening the Base of department. Journal of Consulting and Clinical 11236404 Treatment for Alcohol Problems: Report of a Study by Psychology 67:989–994, 1999. PMID: 10596521 a Committee of the Institute of Medicine Division of RUBIO, G.; JIMENEZ­ARRIERO, M.A.; PONCE, G.; Mental Health and Behavioral Medicine. MORGENSTERN, J.; IRWIN, T.W.; WAINBERG, M.L.; AND PALOMO, T. Naltrexone versus acamprosate: Washington, DC: National Academy Press, 1990. ET AL. A randomized controlled trial of goal choice One year follow­up of alcohol dependence treat­ interventions for alcohol use disorders among men ment. Alcohol and Alcoholism 36:419–425, 2001. KALICHMAN, S.C.; ROMPA, D.; AND CAGE, M. who have sex with men. Journal of Consulting and PMID: 11524308 Sexually transmitted infections among HIV Clinical Psychology 75:72–84, 2007. PMID: seropositive men and women. Sexually Transmitted 17295566 SAITZ, R. Clinical practice: Unhealthy alcohol use. Infections 76:350–354, 2000. PMID: 11141850 New England Journal of Medicine 352:596–607, NAAR­KING, S.; LAM, P.; WANG, B.; ET AL. Brief 2005. PMID: 15703424 KALICHMAN, S.C.; SIMBAYI, L.C.; VERMAAK, R.; ET report: Maintenance of effects of motivational AL. Randomized trial of a community­based alco­ enhancement therapy to improve risk behaviors SAITZ, R.; HORTON, N.J.; CHENG, D.M.; AND hol­related HIV risk­reduction intervention for and HIV­related health in a randomized controlled SAMET, J.H. Alcohol counseling reflects higher men and women in Cape Town South Africa. trial of youth living with HIV. Journal of Pediatric quality of primary care. Journal of General Internal Annals of Behavioral Medicine 36:270–279, 2008. Psychology 33:441–445, 2008. PMID: 17905800 Medicine 23:1482–1486, 2008. PMID: 18618204 PMID: 18836789 SAITZ, R.; PALFAI, T.P.; CHENG, D.M.; ET AL. Brief NAAR­KING, S.; WRIGHT, K.; PARSONS, J.T.; ET AL. KALICHMAN, S.C.; WEINHARDT, L.; DIFONZO, K.; intervention for medical inpatients with unhealthy Healthy choices: Motivational enhancement ther­ ET AL. Sensation seeking and alcohol use as markers alcohol use: A randomized, controlled trial. Annals apy for health risk behaviors in HIV­positive youth. of sexual transmission risk behavior in HIV­positive of Internal Medicine 146:167–176, 2007. PMID: AIDS Education and Prevention 18:1–11, 2006. men. Annals of Behavioral Medicine 24:229–235, 17283347

2002. PMID: 12173680 PMID: 16539572 SAMET, J.H.; CHENG, D.M.; LIBMAN, H.; ET AL. National Institute on Alcohol Abuse and KANER, E.F.; BEYER, F.; DICKINSON, H.O.; ET AL. Alcohol consumption and HIV disease progression. Effectiveness of brief alcohol interventions in pri­ Alcoholism. Helping Patients Who Drink Too Much: Journal of Acquired Immune Deficiency Syndromes mary care populations. Cochrane Database of A Clinician’s Guide: Updated 2005 Edition. 46:194–199, 2007. PMID: 17667330 Systematic Reviews CD004148, 2007. PMID: Bethesda, MD: National Institutes of Health, 17443541 2007. SAMET, J.H.; FREEDBERG, K.A.; STEIN, M.D.; ET AL. Trillion virion delay: Time from testing positive KRANZLER, H.R., AND VAN KIRK, J. Efficacy of nal­ OCKENE, J.K.; ADAMS, A.; HURLEY, T.G.; ET AL. for HIV to presentation for primary care. Archives trexone and acamprosate for alcoholism treatment: Brief physician­ and nurse practitioner­delivered of Internal Medicine 158:734–740, 1998. PMID: A meta­analysis. Alcoholism: Clinical and counseling for high­risk drinkers: Does it work? 9554679 Experimental Research 25:1335–1341, 2001. Archives of Internal Medicine 159:2198–2205, PMID: 11584154 1999. PMID: 10527297 SAMET, J.H.; HORTON, N.J.; MELI, S.; ET AL.A randomized controlled trial to enhance antiretrovi­ KRISTENSON, H.; OHLIN, H.; HULTEN­NOSSLIN, OLMSTED, C.L., AND KOCKLER, D.R. Topiramate ral therapy adherence in patients with a history of M.B.; ET AL. Identification and intervention of for alcohol dependence. Annals of Pharmacotherapy alcohol problems. Antiviral Therapy 10:83–93, heavy drinking in middle­aged men: Results and 42:1475–1480, 2008. PMID: 18698008 2005. PMID: 15751766

278 Alcohol Research & Health 33.3_9.1.10.qxd:32(1).qxp 9/8/10 11:44 AM Page 279

Interventions for HIV­Infected Risky Drinkers

SAMET, J.H.; HORTON, N.J.; MELI, S.; ET AL. ized controlled trials. Journal of Acquired Immune trial. Journal of Consulting and Clinical Psychology Alcohol consumption and antiretroviral adherence Deficiency Syndromes 43(Suppl):S23–S35, 2006. 77:657–667, 2009. PMID: 19634959 among HIV­infected persons with alcohol prob­ PMID: 17133201 lems. Alcoholism: Clinical and Experimental Research WANG, X.; DOUGLAS, S.D.; PENG, J.S.; ET AL. 28:572–577, 2004. PMID: 15100608 SORENSEN, J.L.; DILLEY, J.; LONDON, J.; ET AL. Naltrexone inhibits alcohol­mediated enhancement Case management for substance abusers with of HIV infection of T lymphocytes. Journal of SAMET, J.H.; HORTON, N.J.; TRAPHAGEN, E.T.; ET HIV/AIDS: A randomized clinical trial. American Leukocyte Biology 79:1166–1172, 2006. PMID: AL. Alcohol consumption and HIV disease progres­ Journal of Drug and Alcohol Abuse 29:133–150, 16574767 sion: Are they related? Alcoholism: Clinical and 2003. PMID: 12731685 Experimental Research 27:862–867, 2003a. PMID: WHITLOCK, E.P.; POLEN, M.R.; GREEN, C.A.; ET 12766632 STEIN, M.D.; ANDERSON, B.J.; CAVINESS, C.M.; ET AL. Behavioral counseling interventions in primary AL. Relationship of alcohol use and sexual risk taking care to reduce risky/harmful alcohol use by adults: AMET ARSON ORTON ET AL S , J.H.; L , M.J.; H , N.J.; . among hazardously drinking incarcerated women: A summary of the evidence for the U.S. Preventive Linking alcohol­ and drug­dependent adults to pri­ An event­level analysis. Journal of Studies on Alcohol Services Task Force. Annals of Internal Medicine mary medical care: A randomized controlled trial of and Drugs 70:508–515, 2009. PMID: 19515290 140:557–568, 2004. PMID: 15068985 a multi­disciplinary health intervention in a detoxi­ fication unit. Addiction 98:509–516, 2003b. TULDRA, A.; FUMAZ, C.R.; FERRER, M.J.; ET AL. WONG, F.L.; ROTHERAM­BORUS, M.J.; PMID: 12653820 Prospective randomized two­arm controlled study LIGHTFOOT, M.; ET AL. Effects of behavioral inter­

SAMET, J.H.; PHILLIPS, S.J.; HORTON, N.J.; ET AL. to determine the efficacy of a specific intervention vention on substance use among people living with Detecting alcohol problems in HIV­infected to improve long­term adherence to highly active HIV: The Healthy Living Project randomized con­ patients: Use of the CAGE questionnaire. AIDS antiretroviral therapy. Journal of Acquired Immune trolled study. Addiction 103:1206–1214, 2008. Research and Human Retroviruses 20:151–155, Deficiency Syndromes 25:221–228, 2000. PMID: PMID: 18494840 2004. PMID: 15018702 11115952 WUTZKE, S.E.; CONIGRAVE, K.M.; SAUNDERS, J.B.; SIMONI, J.M.; PEARSON, C.R.; PANTALONE, D.W.; VELASQUEZ, M.M.; VON STERNBERG, K.; AND HALL, W.D. The long­term effectiveness of ET AL. Efficacy of interventions in improving highly JOHNSON, D.H.; ET AL. Reducing sexual risk brief interventions for unsafe alcohol consumption: active antiretroviral therapy adherence and HIV­1 behaviors and alcohol use among HIV­positive A 10­year follow­up. Addiction 97:665–675, 2002. RNA viral load: A meta­analytic review of random­ men who have sex with men: A randomized clinical PMID: 12084136

Vol. 33, No. 3, 2010 279