1. The good news

2. THE GOOD NEWS: Expansion of policies and services and the rapidly shrinking burden of HIV mortality Malebogo Tlhajoane Biomedical Research Training Institute, Zimbabwe Georges Reniers London School of Hygiene & Tropical Medicine, UK

3. Adoption of national HIV policies to promote ART uptake Implementation across ALPHA partner sites – Zimbabwe case study Populations-wide changes mortality – Mortality rates among PLHIV – Impact of ART on life expectancy Outline

4. Has explicit policyHas implicit policy unclear Does not have policy Selected policy indicators influencing ART access Policy WHOKenyaMalawi South Africa TanzaniaUgandaZimbabwe Free ART at public facilities 2011 2004 20102007 2011 2007 Clinical officers or nurses initiate ART 2005 2004 20102005 2011 2010 6 monthly CD4 testing in pre-ART with CD4 < 500 2011 2010 2005 20032010 Lab tests not required to start ART (e.g. FBC, LFTS) Strongly desirable 2005 2004 2013 Option B+ for pregnant women 2012 2011 2011 Selected facilities

5. Implementation of policies influencing ART access

6. Implementation of Option B+

7. Zimbabwe: Complete HIV testing policies WHO 2013 Manicaland 2013 WHO 2015 Manicaland 2015 Service Access and Coverage Free HCT services PITC to ANC Clients PITC to TB clients PITC to STI or FP clients Testing offered to high risk groups Quality of Care ≥1 staff member received HIV testing training in past 2 years Quality of care audit at least once a year Maximum 15 clients per day per counsellor Co-ordination of Care and Patient Tracking Repeat test after 3 month window period Support to PLHIV (%) Pre-test counselling always provided Post-test counselling always provided No Explicit Policy Explicit Policy

8. Consistently wide coverage of provider-initiated HIV testing Manicaland 2013 Manicaland 2015 Service Access and Coverage Free HCT services86100 PITC to ANC Clients97*100 PITC to TB, STI or FP clients100* Testing offered to high risk groups0*0*12 Quality of Care ≥1 staff member received HIV testing training in past 2 years 63*50 Quality of care audit at least once a year94*97 Maximum 15 clients per day per counsellor0 *28 Co-ordination of Care and Patient Tracking Repeat test after 3 month window period9772 Support to PLHIV (%) Pre-test counselling always provided9794 Post-test counselling always provided97 Explicit PolicyNo Explicit Policy Minimal coverage 0-25% Partial Coverage 26-74% Wide Coverage 75-100% ‡ Missing data for over 10% of facilities * Missing data for under 10% of facilities ** Missing data for all observations

9. Successful uptake of WHO recommended HIV treatment policies in Zimbabwe Explicit PolicyNo Explicit Policy Minimal coverage 0-25% Partial Coverage 26-74% Wide Coverage 75-100% ‡ Missing data for over 10% of facilities * Missing data for under 10% of facilities ** Missing data for all observations

10. Increased facility provision of TB screening and treatment Manicaland 2013Manicaland 2015 Quality of Care Quality of care audit at least once a year 100100* ≥1 staff member received HIV treatment training in past 2 years 2554 * Co-ordination of Care and Patient Tracking Drugs collectable by designee 10091 * Adherence monitored among ART patients 100‡100 * Home visits conducted following poor adherence 27 ‡34 * Home visits conducted following missed visit 31‡29 Support for PLHIV Support groups available at facility 40 ‡37 * Medical Management 6 monthly (min) CD4 on ART stable patients 100 ‡100 * Prophylactic IPT in stock 5343 * TB screening at every ART visit 85*100 * TB Treatment offered at facility 100 ‡100 * Pill counts and every visit 93‡100 *

11. uMkhanyakude Masaka Rakai Karonga Reniers et al (2014) AIDS (updated estimates) ~80% decline in the mortality rates of PLHIV since ART Both on/off treatment: higher coverage of ART earlier engagement of PLHIV Relatively high mortality on treatment shortly after the rollout of ART Rapid declines in the mortality rates of PLHIV

12. Adult LE Gains since ART : 10-14 years LE Deficit as of mid 2014 Shortfall of the population- wide adult LE compared to HIV negatives < 2 years in eastern Africa 4 to 6 years in South Africa Reniers, Eaton et al (2015) CROI Life expectancy (LE) gains in excess of 10 years in all study sites LE gains Remaining LE deficit Years

13. Age-cause decomposition of the LE gains between 2001/’04 and 2011/’14, uMkhanyakude (RSA) IAS Poster TUPEC132 Interpretation: bars quantify the contribution of each age and cause group to the overall LE gain in years Findings: HIV and pulmonary TB account for >80% of the LE gains since ART 10% is due to a reduction in injuries (men) Reduction in HIV and TB mortality account for almost all of the LE gains

14. 2016 IAS Poster TUPEC132 HIV and TB mortality still account for most of the LE deficit Age-cause decomposition of the LE deficit in 2011-’14, uMkhanyakude (RSA) HIV and pulmonary TB account for >80% of the remaining LE deficit

15. LE change in the absence of ART: Changes in background mortality Historical changes in HIV incidence Model no-ART counterfactual LE, and estimate: LE change without ART The net contribution of ART Findings: Without ART LE would have decreased further in RSA, but increased in eastern Africa Net LE gains due to ART – ~20 years in South Africa – ~8 years in eastern African studies Reniers, Eaton et al (2015) CROI The LE gains attributable to ART are largest in South Africa Net LE gains attributable to ART LE change without ART

16. ALPHA adult LE trends compared with national-level estimates (UNPD & IHME) (much) larger LE gains in the ALPHA Network partner sites: Severe epidemics (e.g., Kisumu, uMkhanyakude) High uptake of HTC Nonetheless … is the impact of ART sufficiently appreciated in estimates of UNPD & IHME ? A comparison worth making …

17. National policies concerning ART generally exceed WHO guidelines Considerable heterogeneity and mixed progress in implementation at health facility level 80% decline in the mortality rates of PLHIV (also in the ART naïve population) Population-wide gains in the LE at age 15 >1 years p.a. for total gains since ART between 10 and 14 years – cf. post WWII Japan: 0.5 years p.a. – Net impact attributable to ART even larger in South Africa Residual burden of HIV mortality is now under 6 years in South Africa, and under 2 years in eastern African populations The good news