Reimagining Global Health Zero TB : What models do Search-Treat-Prevent Implementors want?

Aamir Khan MD, PhD Executive Director, IRD Zero TB CITIES is a global initiative targeted at creating “islands of elimination” with strong local ownership in high burden settings

Zero TB implements evidence-based Search, Treat & Prevent approaches from around the world shown to be successful in controlling TB Zero TB Cities and districts – Current, new and potential:

Almaty, Kazakhstan** Manila, Philippines** Balti, Moldova Melbourne, Australia** Chennai, India Mexicali, Mexico Dhaka, Bangladesh* Mthata, South Africa** Geissen, Germany Mumbai, India** Hai Phong, Vietnam Muscat, Oman** Hanoi, Vietnam Odessa, Ukraine Ho Chi Mihn City, Vietnam Pattaya, Thailand** Indore, India** Peshawar, Karachi, Pakistan Quetta, Pakistan** Kathmandu, Nepal** Shenzhen, China* Kisumu, Kenya Sofia, Bulgaria** Lagos, Nigeria ** Tbilisi, Georgia** Lima (Carabayllo), Peru Ulaanbaatar, Mongolia Vladimir, Moscow

* First of multiple cities planned ** Discussions of alignment underway Pakistan’s Zero TB Initiative Global Fund Support 2016-2017 USD 40m – Zero TB Karachi +32 districts 2018-2020 USD 40m – Zero TB Karachi, Peshawar, Quetta + 32 districts

>3m 55 mobile >61,000 3 Zero TB >1.5m ~1200 screened X-ray vans/ patients Cities / 32 Chest X-ray staff on verbal 70 fixed with TB districts screens symptoms X-rays notified

12

Active Case Finding ≠ ‘Active Case Finding’

• What is the focus and intensity of case finding? – Which populations are targeted for systematic screening? • Low-income communities, outpatients, prison inmates, factory workers • Adults vs children – How does one measure intensity of ACF? • Number of bacteriological tests done • Yield of patients from screening at specific settings • Proportion of target population screened or tested – Cost of CAD and GeneXpert cartridges as impediment to scale Zero TB Karachi: All Forms TB Notifications 2008 – 2018*

30000

25,064 25,339 25000

20,560 19,336 20000 18,693 18,270 17,545 18,115 16,104 15,993

15000 14,470 Notifications 10000

5000

0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

*Q4 2018 numbers are self counted from TB03 registers Zero TB Karachi: Mobile Chest X-ray Screening among Adults Jan 2018 – Dec 2018

Other X-ray based Chest X-ray / Computer Aided Detection Screening (CAD > 70) Case-finding Models Large Community Community GP-Linked Household Screening Site Hospital Camps Camps Prisons Factories Camps Contacts OPDs (Not for Profit) (Social Enterprise)

Screened 197,346 86,622 111,909 10,071 10,221 41,276 5,069

Presumptive 17, 839 6,147 9,584 597 686 5,174 -

B+ve (Rif+ Included) 1,708 396 172 23 5 57 16

Cases 2,332 475 669 106 13 283 49

Yield from 13% 8% 7% 18% 2% 6% - presumptive Yield from screened 1.2 % 0.5 % 0.6 % 1 % 0.1 % 0.7 % 1 % Number Needed to 85 182 167 95 786 146 103 Screen (NNS) Childhood TB Screening: Karachi, Lahore, Peshawar: July 2016 – June 2018

Zero TB Karachi Lahore Zero TB Peshawar

Screened 212,132 81,705 36,966

13,631 Presumptive 3,997 2,909 (6%) (5%) (8%)

1,568 758 297 TB Diagnosed (12%) (19%) (14%) Zero TB Karachi: Childhood TB as a Proportion of All Forms Case Notifications 2010 – 2018 30000 100%

25,480 90% 25000 80%

70% 20000 60%

15000 50%

40% Notifications 10000 30%

20% Notifications of all Proportion 5000 12% 13% 9% 10% 11% 9% 6% 7% 6% 10% 3,298 0 0% 2010 2011 2012 2013 2014 2015 2016 2017 2018 Karachi Karachi Peads % Peads Zero TB Karachi: GeneXpert Cartridge Use 2014 - 2018

55000 48,324 50000 46,490 45000 42,417 40000 35,492 35000 28,476 30000 26,643 25,983 25000 22,263 22,541 20000 14,766 15000 11,149 13,449 12,049 10,818 10,379 9,159 10000 3,863 5,868 3,162 3,889

Number of GeneXpert Tests GeneXpert of Number 5000 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 2014 2015 2016 2017 2018 Karachi: Quarterly tests and trend of MTB+/Rif+ Detected Q3 2013 – Q12019

60000 35.0 31.0 30.0 50000

24.7 25.0 24.0 23.7 25.0 40000

18.2 20.0 17.2 17.2 30000 15.0 15.0 15.0 16.1 13.6 15.0 12.1 11.1 20000 10.5 9.2 11.5 11.7 8.6 8.7 8.0 10.0 6.5 6.4 7.0 9.0 5.7 7.7 10000 7.0 7.1 6.3 6.6 6.3 6.2 5.0 6.0 6.1 5.7 5.7 5.5 4.5 5.0 4.9 4.6 4.9 4.6 5.1

0 0.0 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 2013 2014 2015 2016 2017 2018 2019

Total Xpert Test % Total MTB+ Det/Total Tested % RR + Det/Total MTB+Detected Karachi Residents: DR-TB Enrolments and Proportion of New Cases 2009 – May 2019

350 100% 319 312 90% 300 80% 244 250 70%

202 60% 200

TB Enrolments TB 176 - 50% 150 40% 38% 34% 40% 32% 32% 118 29% 100 89 30% 19% 73 Cases of New Proportion

Number of DR of Number 66 57 20% 50 14 5% 10% 7% 3% 2% 0 0% 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Total Karachi Residents % New Karachi City: Active Case Finding Resources Invested and Case Notifications 2010 - 2018

250 6000

5000 200

4000 150 3000 100 2000

Case Notifications 100,000per Notifications Case 50 1000 100,000 per Invested Resources

0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018

Adult DSTB Case Notificaton Rates All Ages B+ DSTB Case Notification Rate Xpert Tests/Population *100000 XRAYs/Population *100000

XRAYs/Population = Fixed and mobile Contact Tracing and Prevention

• What is the focus and coverage of contact tracing and prophylaxis? • Household contact tracing vs all contact tracing • Drug susceptible vs drug resistant TB contacts – Uptake of prevention regimens among contacts • 3HP vs others – Cost of Rifapentine as impediment to scale Drug Susceptible TB Prevention Regimen: 6 months INH vs 3HP Zero TB Karachi 2017

6 months INH 3HP Indicators N % N % Contacts offered treatment 1258 956 Contacts started treatment 1029 82 715 75 Contacts refused after initiating treatment 247 24 41 6 Contacts completed treatment 171 17 194 27 Contacts not completing treatment 125 12 28 4 Contacts with unknown outcomes 3 0 2 0 Contacts still on treatment 483 47 464 65 TB Prevention Cascade – Drug Resistant TB

Total contacts (N=800)

<5 years 5-17 years >17 years (N=94) (N=258) (N=448)

Screened Screened Screened (N=88, 93%) (N=238, 93%) (N=411, 93%)

Evaluated Evaluated Evaluated (N=76, 86%) (N=196, 82%) (N=389, 95%)

PT eligible (N=76, 100%) PT eligible (N=96, 49%) PT eligible (N=42, 11%) Eligibility reason: Eligibility reason: Eligibility reason: Age (n=76) TST+ (n=5) Malnourished (n=41) Malnourished (n=91) Diabetes (n=1)

Initiated Initiated Initiated (N=61, 80%) (N=85, 89%) (N=25, 60%)

Completed Completed Completed (N=46, 75%) (N=58, 68%) (N=17, 68%) Zero TB Karachi: Cumulative probabilities for evaluation, prescription, uptake and completion of treatment for presumed DR-TB infection treatment by age group (N=792, Karachi Zero TB, unpublished Amyn Malik, Mercedes Becerra, Hamidah Hussain et al) Prevention Cascade (CP) 100.0%

90.0%

80.0%

70.0%

60.0%

50.0%

40.0%

30.0% Cumulative Probability Cumulative 20.0%

10.0%

0.0% <5 years 5-17 years >17 years Total Evaluated Prescribed Uptake Completion Modeling Zero TB Karachi Impact

• Modeling from 2013 (David Dowdy and Andrew Azman) • Significant declines in 5-year mortality and incidence (2013 Dowdy paper)

• Modeling from Zero TB baseline (Sourya Shreshta and David Dowdy) • Targeted case finding can double reductions in TB incidence

• But what is the combined impact of targeted Active Case Finding and targeted Preventive Therapy? (Sourya Shreshta and David Dowdy)

Korangi Town: All Ages All Forms Notifications 2010 – Q4 2018

5000 4,642 20% decrease in 4500 4,305 notifications 4,123 reported from 4000 3,735 Town 3500 3,392

3000 2,850 2,598 2500 2,223

Notification 2000

1500 1,314 1,043 1000 793

500

0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 includes 5 BMUs: Indus Hospital, SZC Korangi, SGH Korangi, Anti TB Association, and Baldia Maternity Home Korangi Town: Active Case Finding Resources Invested and Case Notifications 2010 - 2018

600 25000

500 20000

400 15000 300 10000 200

5000 Case Notifications per 100,000 Notifications Case

100 100,000 per Invested Resources

0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018

All Ages B+ DSTB Case Notification Rate Adult Case Notificaton Rates Xpert Tests/Population *100000 XRAYs/Population *100000

XRAYs/Population = Fixed and mobile Karachi : All Forms TB Yield through Community Chest X-ray Camps 2017 and 2018

2017

*

* Korangi Town 2018

* Zero TB Interventions in Korangi: CXRs and Increasing NNS Q1 2017 – Q4 2018

25000 160 CXRs NNS (All Forms) NNS (B+)

140 2017 Q1 ------20000 120 2017 Q2 8,294 25 47

15000 100 2017 Q3 18,639 39 97 80

NNS 2017 Q4 21,909 62 137 10000

60 2018 Q1 21,004 65 91 Number Number ofCXRs 40 5000 2018 Q2 16,251 56 82 20 2018 Q3 19,293 67 89 0 0 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 2018 Q4 23,749 92 117 2017 2018 Chest X-Rays NNS (All Forms) NNS (B+)

NNS: Number Needed to Screen # of CXRs/# of cases ZTB Interventions include GHD performed CXRs (community and facility), CHS run community camps and CHS center CXRs Indus Hospital: All Forms Notifications and Proportion of Korangi Residents 2008 – 2018

4500 76% 80%

4000 68% 70% 61% 3500 57% 60%

3000 48% 45% 50% 43% 43% 2500 43% 40% 39% 40% 2000

Notifications 30% 1500 20%

1000 Residents of Korangi Proportion

500 10%

0 0% 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Korangi Town Residents Other Town Residents % of Korangi Residents Korangi Adult Tuberculosis Case Notification Rate Korangi Pediatric Tuberculosis Case Notification Rate 500 compared to the rest of Karachi 90 compared to the rest of Karachi

450 80

400 70

350 60 300 50 250 40

200

Rate per 100,000 Rate per 100,000 Rate 30 150

20 100

50 10

0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2010 2011 2012 2013 2014 2015 2016 2017 2018 Year Year Koangi- TB (all forms) Case Notification Rate Korangi - TB (all forms) Case Notification Rate Karachi w/o Korangi - TB (all forms) Case Notification Rate Karachi w/o Korangi - TB (all forms) Case Notification Rate TB Modeling and Universal Health Care

• Modeling TB investments as a conduit to delivering screening and linkage to care for other public health priority diseases – Capturing benefits to patients and costs saved to UHC by early diagnosis and referral • Diabetes • Depression and anxiety • Hepatitis C • COPD Zero TB Karachi: HCV Testing and Treatment Cascade Jan to July 2019

25000

20371 20000

15000

10078 9598 (49%) (95%)

10000 Number of Individuals

5000 806 745 475 324 245 (8%) (92%) (64%) (68%) (76%) 0 Eligible for Verbal RDT Anti HCV +ve PCR testing PCR positive Baseline Treatment screening screening screening evaluation initiation TB Program: Screening for Diabetes and Depression Integrated Practice Units (IPU)

Supported by the Harvard Medical School-Center for Global Health Delivery-Dubai

Integration of mental health and diabetes services with existing TB treatment sites to improve adherence, treatment outcomes and provide holistic care

Mass screening for TB Adherence & Mental Psychologist-Severe Diabetes Screening Diabetes Medication and Depression & Anxiety Health Counseling Case Consultation and Consultation Counseling TB Program: Screening for Diabetes and Depression Integrated Practice Units (IPU)

Psycho-Social Support Interventions (PSSI)

6 10 3,500 1,012 3,441 IPUs developed Counselors trained Screened for Enrolled and given Screened for across public and and deployed Depression and baseline adherence Diabetes. 21% private hospital Anxiety. 30% counselling. (known diabetics settings Symptomatic 562 completed and RBS+) linked to intervention HbA1c testing/care Key Finding from Integrated Practice Unit (IPU) for Mental Health and TB

92% of patients who completed the mental health intervention also completed TB treatment,

compared to a 75% TB treatment completion rate in those who did not complete the mental

health intervention (Unpublished data) Summary (of sorts)

• Data availability drives the development of TB models – Increasingly models should drive good practice • Good epidemic control practice = Search, Treat, Prevent – Models can shine a bright light on the Search, Treat, Prevent path – Implementers will tell you what is useful and what isn’t (for them) • Effective modeling can help advocate for – price reductions of mobile X-ray vans, CAD software, GeneXpert, Rifapentine – high prices are barriers to scale – Use of TB investments for delivering UHC and social protection services Acknowledgement of Zero TB Partners Discussion Slides (only if needed) Why stop at UHC? Why not Social Protection?

• Should we model TB programs as conduit for social protection services? – Improved targeting of poorest for social protection services – Cost savings in integrating social protection services • Rationale for TB Program – Improved uptake of services – Improved patient outcomes Psycho-Social Support Interventions (PSSI)

Improve treatment adherence by addressing the social determinants of health. Pilot conducted with 250 MDR-TB patients

Life-Skills Based Financial Security/ Well-being Kits Home Renovation Counseling Education Microfinance Psycho-Social Support Interventions (PSSI)

Improve treatment adherence by addressing the social determinants of health. Pilot conducted with MDR-TB patients 345 20 233 23 239 Patients and Family Patients Referred for Well-being Kits Homes Assessed for DR-TB Patients Members given Life- Microfinance Loans Distributed Infection Control Identified for Mental Skills Education Health Counseling Korangi Town Residents: MDR-TB Enrolments and Proportion that are New Infections 2009 – May 2019 70 100% 62 90% 60 80%

50 47 70% 60% 40 47% 47% 50% 31 30 30 27 38% 33% 40% 21 22 42% 37% 19 30% Number of Enrolments of Number 20 18% Cases New of Percentage 17% 13 20% 10 12 2 5% 10% 0% 0 0% 0% 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Korangi Town % New Should we account for emerging models of private care in costing diagnostics and treatment? • Out-of-pocket costs for diagnosis and care (TB and other diseases) often not captured • Social enterprises can reduce costs to donors/governments by cross-subsidizing TB care from revenues generated by general patient services (e.g. laboratory tests) – Still a cost to society, even if no longer being borne by donor/government Social Enterprise Scale-up in Pakistan (2017-Current) 51 private provider network diagnostic centers in 27 districts TB Services Model and Aggregator Platform

Private Sector Network Provincial Tuberculosis Referrals generated through Program network of 100 health providers All TB cases registered at SZ centers are around each of the 61 centers of notified to the Department of Health TB excellence program,

Active Case Finding TB Treatment All patients detected with TB Free TB Screening through GP are registered on free 1st line clinic camps and Mobile X- treatment, counselled and Ray camps in the community followed-up

Drug Resistant TB Walk-Ins Management sites Patients referred by people All cases with DR-TB are referred to they trust who have Programmatic Management of experienced SZ services DR-TB sites Social Enterprise Models for Sustainable TB Care -A capitation-based reimbursement for TB under UHC can create incentives for increased cost-effectiveness and cost-sharing through cross-subsidization models

45,000 350 4,500,000

40,000 4,000,000 300 35,000 3,500,000 250 30,000 3,000,000

25,000 200 2,500,000 2,000,000 20,000 150 1,500,000 15,000 100 1,000,000

10,000

Annual Cases Notified Cases Annual Share Projections with GF with Projections Share 50 - 500,000 5,000 - - 0 Cost 2017 2018 2019 2020 2021 2022 2023 2017 2018 2019 2020 2021 2022 2023 Cost Share by CHS Costs by Donor Annual Cases Notified to NTP Cost Per Case to Donor (USD)

$4,000 $1,500 $700 Mega Cities Monthly Revenue Urban and Rural Average monthly revenue Generated for supporting from 36 new centers in Average monthly revenue operational costs by each of first 3 Karachi and Lahore from 22 new centers in smaller centers in 2017 cities and rural districts