Link4Health Study : - PowerPoint Presentation

Link4Health Study: A combination strategy for linkage to and retention in HIV care in Swaziland Margaret L. McNairy, Matthew R. Lamb, Averie B. Gachuhi, Harriet Nuwagaba-Biribonwoha, Sean Burke, Sikhathele Mazibuko, Velephi Okello, Peter Ehrenkranz, Ruben Sahabo, and Wafaa M. El-SadrJuly 20, 2016 1

BackgroundAll steps in the HIV care cascade must be achieved to optimize health outcomes and decrease HIV transmission. 2*Bicego et al, Plos One 2013; Swaziland MOH 2012Swaziland has a severe generalized HIV epidemic.Adult HIV prevalence 31%* HIV incidence 2.4%* Barriers to engagement across the cascade include structural, behavioral and biomedical factors .

Link4Health: PurposeTo evaluate the effectiveness of a combination strategy of multiple interventions targeting barriers at various cascade steps on linkage and retention3

Assessed for ART eligibilityRetained in careLink4Health: InterventionCombination Intervention Strategy (CIS) 4Cascade StepLinked to CareAssessed for ART EligibilityInitiated ARTRetained in CareCIS Components1. Point-of-care CD4 at HIV testingXXX2. Accelerated ART for CD4 < 350XX 3. SMS appointment reminders X X 4. Basic care and prevention packageX X 5. Non-cash financial incentive X X

Link4Health: Study DesignImplementation science designCluster-randomized controlled trial10 study units randomized to CIS versus Standard of CareStudy unit included 1 secondary and 1 affiliated primary-level health clinic Stratified by location (rural/urban) and implementing partnerStudy population:> 18 years identified as HIV+Willing to receive HIV care at study unitNo prior ART in past 6 monthsNot currently pregnantNot planning to leave catchment areaStudy follow-up : 12 months5

Link4Health: Study OutcomesPrimary outcome: Linkage: HIV clinic visit within 1 month after HIV testing at assigned study unit andRetention: HIV clinic visit within 12 months (+/- 3 months) after HIV testing at assigned study unitSecondary outcomes:Linkage at 1 monthRetention at 12 monthsAssessment for ART eligibility (i.e. CD4 cell count/WHO Stage) ART initiationMortality6

Link4Health: Data Collection & AnalysisData abstracted from routine health records including clinic registers and patient files and patient trackingIntent-to-treat analysisComparison of Relative Risk (RR) for achieving primary outcome among all participants in CIS versus SOC Per-protocol analysisComparison of RR for achieving primary outcome among participants who received full package of CIS versus SOC for duration of study participationInteraction analysisAssessment of RR for achieving primary outcome by key sub-group using Interaction Contrast Ratios7

CIS (5 Study Units)SOC (5 Study Units)Link4Health: Study Population Screened for study eligibility(N = 1,234)Refused (N = 23)Refused (N = 114)Ineligible (N = 111)Ineligible (N = 101)Enrolled in study (N = 1,100)Reasons ineligible: CIS arm: < 18 yrs (N = 4), not HIV-positive (N = 3), refuse referral to SU (N = 26), refuse to provide contact information (N = 4), plans to move in next year (N = 22), recent history of HIV care or ART (N = 12), other (N = 40) SOC arm: < 18 yrs (N = 2), not HIV-positive (N = 1), refuse referral to SU (N = 28), plans to move in next year (N = 21), recent history of HIV care or ART (N = 13), other (N = 34) Screened for study eligibility (N = 1,316) Enrolled in study (N = 1,101) Primary Outcome Assessment (N = 1,036) Primary Outcome Assessment (N = 932) 8 10 Study Units

Link4Health: Site and Participant CharacteristicsSite Characteristics60% urban, 100% supported by MOHAll 4 regions of SwazilandParticipant Characteristics2,201 HIV+ adults enrolled August 2013-November 201459% femaleMedian age 32 years (IQR 26-40), 20% <25 years37% married53% unemployed; median income $9 (IQR 0-32)/weekHIV testing site: 80% VCT, 20% PICT 9

CISN = 1100SOCN = 1101RR (95 % CI)P valueIntent-to-Treat 705 (64%)477 (43%)1.48 (1.36-1.60)<0.001Accounting for Clustering*1.52 (1.18-1.95)0.003Per-protocol*#672 (72%)477 (43%)1.68 (1.53-1.79)<0.001 * Accounting for within study unit clustering using a random-intercept log- poisson regression model with robust standard error# Adjusted participants who received CIS against those in SOC according to study protocol^ Considered achievement of outcome if participant reported linkage and retention in their study unit or in follow-up questionnaire to a different clinic Link4Health: Primary Outcome Linkage within 1 month & Retention at 12 months after HIV Testing 10

CIS Arm: N = 705 (64%)SOC Arm: N = 477 (43%)Intent-to-treat:Relative Risk 1.48 (95% CI 1.36-1.60 p< 0.0001)Accounting for Clustering:Relative Risk 1.52 (95% CI 1.18-1.95 p= 0.0026)Link4Health: Primary OutcomeLinkage within 1 month & Retention at 12 months after HIV Testing 11

CIS (N =1100)SOC(N= 1101)RR^ (95 % CI)P valueLinkage1 month 1010 (92%)918 (83%)1.12 (0.96-1.30)0.15Mean time (std)2.5 days (19.5)7.4 days (46.6) 0.003 Retention 12 months 720 (65%) 498 (45%) 1.47 (1.17-1.85) 0.002 ART Eligibility Assessment Ever 1100 (100%) 920 (84%) 1.20 (1.07-1.34) 0.004 ART Initiation Total 710 (65%)* 635 (58%)** 1.16 (0.96-1.40) 0.13 Median time (IQR) 8 days (3-37) 17 days (7-57) <0.001 Death Before ART Initiation 10 (1%) 23 (2%) 0.44 (0.19-1.00) 0.05 After ART Initiation 25 (2%) 20 (2%) 1.18 (0.57-2.46) 0.64 Link4Health: Secondary Outcomes ^ Accounting for clustering using random-intercept model with robust error estimates * (85% among eligible) **(88% among eligible) 12

Link4Health: HIV Care Cascade53%onART12% Pre-ART 39%onART6% Pre-ARTP = 0.15P 0.04P = 0.13P = 0.00213Retention 12 months after HIV testing HIV + Test(Study Enrollment) Linked within 1 month

Subgroup Analysis of Primary OutcomeAge (yrs) Sex Daily Income (USD) Employment Marital Status Away from home in last year Travel Time First HIV+ Test Household member with HIV Facility Type Relative Risk for Primary Outcome

Strengths and LimitationsStrengthsPragmatic study design integrated in clinical care Multiple interventions along the HIV care cascade evaluated Majority of clinics in Swaziland included LimitationsPrimary outcome from routinely collected data in medical recordsInability to evaluate effectiveness of individual components of CIS intervention15

ConclusionsThe Link4Health combination strategy of evidenced-based interventions, aimed at gaps in the HIV care cascade, was associated with a 50% increase in prompt linkage to care and 12-month retention. The combination strategy improved ART eligibility assessment, reduced median time to ART initiation, and reduced mortality among participants prior to ART initiation.This effective and pragmatic strategy can be adapted to other settings and to specific patient populations to enhance the trajectory of patients across the HIV care cascade.16

Many thanks to all of the collaborators, health workers, and patients who made this study possible.Investigators Wafaa El-SadrPeter EhrenkranzMatthew LambMargaret McNairySikhathele MazibukoHarriet Nuwagaba-BiribonwohaVelephi OkelloRuben SahaboICAP NY Study StaffAverie GachuhiNeena PhilipThis study is funded by NIH R01AI100059 and conducted in collaboration with CDC and PEPFAR ICAP Swaziland Study Staff Pido BongominSean BurkeCebsile DlaminiThabiso Fakudze Henry Ginindza Altaye Kidane Veli Madau Yvonne Mavengere Nozipho Ndlovu Fortune Ndhlovu Rita Nunu Zweli Simelane Nkosikhona Dlamini Ministry of Health Nomthandazo Lukhele Simangele Mthethwa Rejoice Nkambule Simon Zwane Health facility staff 17