1. www.ias2011.org Community adherence support sustains improved three year outcomes for children on ART A. Grimwood 1, G. Fatti 1, E. Mothibi 1, M. Malahlela 1, A. Jason 1 1 Kheth'Impilo, Cape Town, South Africa TUAD0103 : 19 July 2011 Rome

2. www.ias2011.org Kheth’Impilo SA NGO supports district scale up quality services for the management of HIV/AIDS at PHC level, focusing on providing a family centered comprehensive & integrated service; KI operates in: 96 sites in the Eastern Cape, KwaZulu Natal, Mpumalanga & the Western Cape with >90 000 patients RIC; Programmes: Health Services Cluster (HSC) – ART (Adults & children), TB, HCT & PMTCT linked to Community Services Cluster (CSC) – for Adherence & Psychosocial support

3. www.ias2011.org We know Optimum adherence to ART is critical for effective HIV/AIDS management Poor adherence is a major predictor of treatment failure, progression to AIDS and death High levels of virologic failures (>40%) noted in many ART programmes in resource-poor settings Primary care sites have better RIC than district and regional hospitals (Fatti G et al PLoS ONE 2010)

4. www.ias2011.org We also know Sites with <950pts enrolled have less LTFU than larger sites (Fatti et al J Acqu Imm Def Synd 2011, in press) Sites with PA support have significantly better adult RIC, VL suppression & LTFU ( Igumbor et al AIDS Care,2011) Rural children travelling to urban treatment centres have increased LTFU (Fatti et al JAIDS 2010) Males & adolescents have a greater risk of LTFU (Fatti G et al PLoS ONE 2010)

5. www.ias2011.org Community Adherence Community based adherence support health care workers called Patient Advocates (PAs) were introduced in 2004; Link clinical services & community; trained in the basics of HIV, patient rights, confidentiality, ethics, etc. Ensure ongoing treatment, counselling and psycho-social support at the community level; Special attention paid to very important patients (VIPs); the ill, pregnant, TB, children & adolescents, those who have not disclosed & those showing early signs of defaulting; VIPs make up 40% of PA’s workload; Patients encouraged to contract with themselves & get a treatment buddy to facilitate adherence to positive lifestyle choices that include the taking of treatment & keeping appointments

6. www.ias2011.org PA Support Structure AREA COORDINATOR PA PRIMARY HEALTH CARE CENTRE (Clinics) DISTRICT OFFICE NATIONAL OFFICE COMMUNITY HEALTH CENTRE  Site Facilitator  CSC District Coordinator  CSC Trainer  Doctor  Nurse  Pharmacist  PMTCQuality Mentor  Social Worker  Data Quality Manager Roving SWAT TEAM  Site Facilitator

7. www.ias2011.org PAs assist with patient treatment readiness & assess: 1.Psychosocial barriers to adherence including non-disclosure are identified 2.Pre-treatment initiation education to ensure the understanding around the need for adherence 3.Plan support services to suit individual client needs through planned home visits and clinic support 4.Regular follow-ups. Information gathered is presented at the treatment initiation Multidisciplinary Team meetings.

8. www.ias2011.org VIP – visiting regime Week 1 – at least 3 times 1st week; Weekly visits for the rest of the month; If patient is adherent, psycho-social circumstances fine, monthly visits X 5 months. At 6 mths - If adherent, no missed pick up dates, no side effects –VL undetectable & CD4 count up – no further home visits; Continue with monthly checks per phone or clinic; –If non-adherent / not improving – review with Clinical team –If PMTCT, visits continue monthly to when baby is 18 months. If HIV tests are negative, follow-up discontinues. Mother would fall into stable regime or if non-adherent remains VIP.

9. Methods ART-naïve children enrolling at 47 facilities in four provinces in SA between Jan 2004- Sept 2009 included. Outcomes: Mortality and patient attrition. The vital status of children LTFU were cross-checked with national death records. Corrected mortality was estimated using Kaplan-Meier and multivariable Cox regression to determine effect of children who were attached to PAs.

10. www.ias2011.org Results Database records for a total of 6442 children < 16 years of age were screened for eligibility for the study –1134 Children excluded commenced ART within 6m of closure of the site database, –1381 ART experienced, –269 who had zero days of follow-up time & –95 children for whom it could not be definitively ascertained as to whether they had support from a PA or not.

11. Results II 3563 ART-naïve children were included in analyses 323 (9.1%) with PAs, 3240 (90.9%) without PAs Baseline characteristics: Median age 6.3 years (P=0.49); CD4% 12.0% (P=0.18); advanced clinical stage 60.0% (P=0.18) between groups At the start of treatment, children with PAs had a higher proportion below one year of age, a higher proportion who received treatment at PHC facilities and a higher proportion who received AZT instead of d4T

12. www.ias2011.org Results III Missing baseline TB treatment information 2 (0.6%) vs. 276 (8.5%) (P <0.0005) & missing baseline immunological values significantly lower in patients supported by PAs; 44 (13.6%) vs. 766 (23.6%) (P <0.0005) Total observation time was 4848 person-years.

13. www.ias2011.org 2. KI patient advocates (PAs) assigned to children on ART: logrank P = 0.027 with PAs without PAs Retention in care logrank P = 0.027 Adjusted hazard of attrition of patients with PAs: 0.57 (CI: 0.35–0.94)

14. www.ias2011.org 2. KI patient advocates (PAs) assigned to children on ART: Mortality without PAs Adjusted hazard of mortality of patients with PAs: 0.40 (CI: 0.15–1.06) Corrected mortality without PAs with PAs logrank P = 0.060

15. www.ias2011.org Results IV –Other baseline factors independently associated with mortality were age below two years, WAZ-scores below -3, severe immunodeficiency receiving treatment for tuberculosis

16. www.ias2011.org Results V Virologic suppression was equivalent between the two groups, with 78.9% (95% CI: 70.8–85.6; n=128) & 82.4% (95% CI: 80.2–84.4; n =1351) of patients achieving viral suppression at six months in children with and without PAs respectively (P =0.33). Multivariate analyses to 36 months of ART showed equivalence in virologic response between the groups (P =0.18). The median CD4 percentage increase after six months of treatment was 7.2% (IQR: 3.2–13.4) & 8.0% (IQR: 4.0–12.6), in children with and without PAs, respectively, (P =0.57). Multivariate analyses of CD4 percentage until 36 months revealed no significant difference between the groups (P =0.20).

17. www.ias2011.org Conclusions Community adherence support is critical for ensuring good survival and remaining in care outcomes & need to be closely linked and coordinated with primary level health services Monitoring & evaluation is key to quality of delivery as services are dynamic & require constant review & support Community adherence worker training should be seen as part of career pathing in an environment of critical health care shortages & can be a cost effective support of the delivery of primary health services in the community

18. www.ias2011.org Thanks to Our patients Patient Advocates DonorsStaff