Estimating the real-world effects of expanding antiretroviral treatment eligibility: Evidence from a regression discontinuity analysis in Zambia.

BACKGROUND: Although randomized trials have established the clinical efficacy of treating all persons living with HIV (PLWHs), expanding treatment eligibility in the real world may have additional behavioral effects (e.g., changes in retention) or lead to unintended consequences (e.g., crowding out sicker patients owing to increased patient volume). Using a regression discontinuity design, we sought to assess the effects of a previous change to Zambia's HIV treatment guidelines increasing the threshold for treatment eligibility from 350 to 500 cells/µL to anticipate effects of current global efforts to treat all PLWHs. METHODS AND FINDINGS: We analyzed antiretroviral therapy (ART)-naïve adults who newly enrolled in HIV care in a network of 64 clinics operated by the Zambian Ministry of Health and supported by the Centre for Infectious Disease Research in Zambia (CIDRZ). Patients were restricted to those enrolling in a narrow window around the April 1, 2014 change to Zambian HIV treatment guidelines that raised the CD4 threshold for treatment from 350 to 500 cells/µL (i.e., August 1, 2013, to November 1, 2014). Clinical and sociodemographic data were obtained from an electronic medical record system used in routine care. We used a regression discontinuity design to estimate the effects of this change in treatment eligibility on ART initiation within 3 months of enrollment, retention in care at 6 months (defined as clinic attendance between 3 and 9 months after enrollment), and a composite of both ART initiation by 3 months and retention in care at 6 months in all new enrollees. We also performed an instrumental variable (IV) analysis to quantify the effect of actually initiating ART because of this guideline change on retention. Overall, 34,857 ART-naïve patients (39.1% male, median age 34 years [IQR 28-41], median CD4 268 cells/µL [IQR 134-430]) newly enrolled in HIV care during this period; 23,036 were analyzed after excluding patients around the threshold to allow for clinic-to-clinic variations in actual guideline uptake. In all newly enrolling patients, expanding the CD4 threshold for treatment from 350 to 500 cells/µL was associated with a 13.6% absolute increase in ART initiation within 3 months of enrollment (95% CI, 11.1%-16.2%), a 4.1% absolute increase in retention at 6 months (95% CI, 1.6%-6.7%), and a 10.8% absolute increase in the percentage of patients who initiated ART by 3 months and were retained at six months (95% CI, 8.1%-13.5%). These effects were greatest in patients who would have become newly eligible for ART with the change in guidelines: a 43.7% increase in ART initiation by 3 months (95% CI, 37.5%-49.9%), 13.6% increase in retention at six months (95% CI, 7.3%-20.0%), and a 35.5% increase in the percentage of patients on ART at 3 months and still in care at 6 months [95% CI, 29.2%-41.9%). We did not observe decreases in ART initiation or retention in patients not directly targeted by the guideline change. An IV analysis found that initiating ART in response to the guideline change led to a 37.9% (95% CI, 28.8%-46.9%) absolute increase in retention in care. Limitations of this study include uncertain generalizability under newer models of care, lack of laboratory data (e.g., viral load), inability to account for earlier stages in the HIV care cascade (e.g., HIV testing and linkage), and potential for misclassification of eligibility status or outcome. CONCLUSIONS: In this study, guidelines raising the CD4 threshold for treatment from 350 to 500 cells/µL were associated with a rapid rise in ART initiation as well as enhanced retention among newly treatment-eligible patients, without negatively impacting patients with lower CD4 levels. These data suggest that health systems in Zambia and other high-prevalence settings could substantially enhance engagement even among those with high CD4 levels (i.e., above 500 cells/µL) by expanding treatment without undermining existing care standards.

An exploration of multi-level factors affecting routine linkage to HIV care in Zambia's PEPFAR-supported treatment program in the treat all era.

Multiple steps from HIV diagnosis to treatment initiation and confirmed engagement with the health system are required for people living with HIV to establish full linkage to care in the modern treat all era. We undertook a qualitative study to gain an in-depth understanding of the impeding and enabling factors at each step of this linkage pathway. In-depth interviews were conducted with fifty-eight people living with HIV recruited from ten routine HIV care settings supported by the U.S. President's Emergency Plan for AIDS Relief (PEPFAR) in Lusaka, Zambia. Using a semi-structured interview guide informed by an established conceptual framework for linkage to care, questions explored the reasons behind late, missed, and early linkage into HIV treatment, as well as factors influencing the decision to silently transfer to a different clinic after an HIV diagnosis. We identified previously established and intersecting barriers of internal and external HIV-related stigma, concerns about ART side effects, substance use, uncertainties for the future, and a perceived lack of partner and social support that impeded linkage to care at every step of the linkage pathway. However, we also uncovered newer themes specific to the current test and treat era related to the rapidity of ART initiation and insufficient patient-centered post-test counseling that appeared to exacerbate these well-known barriers, including callous health workers and limited time to process a new HIV diagnosis before treatment. Long travel distance to the clinic where they were diagnosed was the most common reason for silently transferring to another clinic for treatment. On the other hand, individual resilience, quality counseling, patient-centered health workers, and a supportive and empathetic social network mitigated these barriers. These findings highlight potential areas for strengthening linkage to care and addressing early treatment interruption and silent transfer in the test and treat era in Zambia.

Programme science in action: lessons from an observational study of HIV prevention programming for key populations in Lusaka, Zambia.

INTRODUCTION: Optimizing uptake of pre-exposure prophylaxis (PrEP) for individuals at risk of HIV acquisition has been challenging despite clear scientific evidence and normative guidelines, particularly for key populations (KPs) such as men who have sex with men (MSM), female sex workers (FSWs), transgender (TG) people and persons who inject drugs (PWID). Applying an iterative Programme Science cycle, building on the effective programme coverage framework, we describe the approach used by the Centre for Infectious Disease Research in Zambia (CIDRZ) to scale up PrEP delivery and address inequities in PrEP access for KP in Lusaka, Zambia. METHODS: In 2019, CIDRZ partnered with 10 local KP civil society organizations (CSOs) and the Ministry of Health (MOH) to offer HIV services within KP-designated community safe spaces. KP CSO partners led KP mobilization, managed safe spaces and delivered peer support; MOH organized clinicians and clinical commodities; and CIDRZ provided technical oversight. In December 2021, we introduced a community-based intervention focused on PrEP delivery in venues where KP socialize. We collected routine programme data from September 2019 to June 2023 using programme-specific tools and the national electronic health record. We estimated the before-after effects of our intervention on PrEP uptake, continuation and equity for KP using descriptive statistics and interrupted time series regression, and used mixed-effects regression to estimate marginal probabilities of PrEP continuity. RESULTS: Most (25,658) of the 38,307 (67.0%) Key Population Investment Fund beneficiaries were reached with HIV prevention services at community-based venues. In total, 23,527 (61.4%) received HIV testing services, with 15,508 (65.9%) testing HIV negative and found PrEP eligible, and 15,241 (98.3%) initiating PrEP. Across all programme quarters and KP types, PrEP uptake was >90%. After introducing venue-based PrEP delivery, PrEP uptake (98.7% after vs. 96.5% before, p < 0.001) and the number of initiations (p = 0.014) increased significantly. The proportion of KP with ≥1 PrEP continuation visit within 6 months of initiation was unchanged post-intervention (46.7%, 95% confidence interval [CI]: 45.7%, 47.6%) versus pre-intervention (47.2%, 95% CI: 45.4%, 49.1%). CONCLUSIONS: Applying Programme Science principles, we demonstrate how decentralizing HIV prevention services to KP venues and safe spaces in partnership with KP CSOs enabled successful community-based PrEP delivery beyond the reach of traditional facility-based services.

Trends in SARS-CoV-2 seroprevalence among pregnant women attending first antenatal care visits in Zambia: A repeated cross-sectional survey, 2021-2022.

SARS-CoV-2 serosurveys help estimate the extent of transmission and guide the allocation of COVID-19 vaccines. We measured SARS-CoV-2 seroprevalence among women attending ANC clinics to assess exposure trends over time in Zambia. We conducted repeated cross-sectional SARS-CoV-2 seroprevalence surveys among pregnant women aged 15-49 years attending their first ANC visits in four districts of Zambia (two urban and two rural) during September 2021-September 2022. Serologic testing was done using a multiplex bead assay which detects IgG antibodies to the nucleocapsid protein and the spike protein receptor-binding domain (RBD). We calculated monthly SARS-CoV-2 seroprevalence by district. We also categorized seropositive results as infection alone, infection and vaccination, or vaccination alone based on anti-RBD and anti-nucleocapsid test results and self-reported COVID-19 vaccination status (vaccinated was having received ≥1 dose). Among 8,304 participants, 5,296 (63.8%) were cumulatively seropositive for SARS-CoV-2 antibodies from September 2021 through September 2022. SARS-CoV-2 seroprevalence primarily increased from September 2021 to September 2022 in three districts (Lusaka: 61.8-100.0%, Chongwe: 39.6-94.7%, Chipata: 56.5-95.0%), but in Chadiza, seroprevalence increased from 27.8% in September 2021 to 77.2% in April 2022 before gradually dropping to 56.6% in July 2022. Among 5,906 participants with a valid COVID-19 vaccination status, infection alone accounted for antibody responses in 77.7% (4,590) of participants. Most women attending ANC had evidence of prior SARS-CoV-2 infection and most SARS-CoV-2 seropositivity was infection-induced. Capturing COVID-19 vaccination status and using a multiplex bead assay with anti-nucleocapsid and anti-RBD targets facilitated distinguishing infection-induced versus vaccine-induced antibody responses during a period of increasing COVID-19 vaccine coverage in Zambia. Declining seroprevalence in Chadiza may indicate waning antibodies and a need for booster vaccines. ANC clinics have a potential role in ongoing SARS-CoV-2 serosurveillance and can continue to provide insights into SARS-CoV-2 antibody dynamics to inform near real-time public health responses.

Estimating potential silent transfer using baseline viral load measures among people presenting as new to HIV care in Lusaka, Zambia: a cross-sectional study.

OBJECTIVES: To estimate potential silent transfer using baseline viral load measures among individuals presenting as new to HIV care in routine HIV clinical settings in Lusaka, Zambia. DESIGN: Cross-sectional study. SETTING: Two large, urban government-operated health facilities supported by the Centre for Infectious Disease Research in Zambia. PARTICIPANTS: A total of 248 participants with an incident positive HIV rapid test. OUTCOME MEASURES: The primary outcome measure was HIV viral suppression at baseline (i.e., potential silent transfer), defined as having a viral load ≤1000 RNA copies(c)/mL at the time of initiating HIV care. We also examined viral suppression at ≤60 c/mL. METHODS: We surveyed and measured baseline HIV viral load as part of the national recent infection testing algorithm among people living with HIV (PLWH) presenting as new to care. Using mixed effects Poisson regression, we identified characteristics among PLWH associated with potential silent transfer. RESULTS: Among the 248 PLWH included, 63% were women with median age of 30, and 66 (27% (66/248)) had viral suppression at ≤1000 c/mL and 53 (21% (53/248)) at ≤60 c/mL thresholds, respectively. Participants aged 40+ years had a significantly higher adjusted prevalence of potential silent transfer (adjusted prevalence ratio (aPR): 2.10; 95% CI: 2.08, 2.13) compared with participants aged 18-24 years. Participants reporting no formal education had a significantly higher adjusted prevalence of potential silent transfer (aPR: 1.63; 95% CI: 1.52, 1.75) compared with those completing primary education. Among 57 potential silent transfers who completed a survey, 44 (77%) indicated having tested positive previously at ≥1 of 38 clinics in Zambia. CONCLUSIONS: The high proportion of PLWH with potential silent transfer points to clinic shopping and/or co-enrolment at multiple care sites simultaneously, suggesting an opportunity to improve care continuity at the time of HIV care entry.

Geospatial Patterns of Progress towards UNAIDS "95-95-95" Targets and Community Vulnerability in Zambia.

In sub-Saharan Africa, HIV/AIDS remains a leading cause of death. The UNAIDS established the "95-95-95" targets to improve HIV care continuum outcomes. Using geospatial data from the Zambia Population-based HIV Impact Assessment (ZAMPHIA), this study aims to investigate geospatial patterns in the "95-95-95" indicators and individual-level determinants that impede HIV care continuum in vulnerable communities, providing insights into the factors associated with gaps. This study used data from the 2016 ZAMPHIA to investigate the geospatial distribution and individual-level determinants of engagement across the HIV care continuum in Zambia. Gaussian kernel interpolation and optimized hotspot analysis were used to identify geospatial patterns in the HIV care continuum, while geospatial k-means clustering was used to partition areas into clusters. The study also assessed healthcare availability, access, and social determinants of healthcare utilization. Multiple logistic regression models were used to examine the association between selected sociodemographic and behavioral covariates and the three main outcomes of study. Varied progress towards the "95-95-95" targets were observed in different regions of Zambia. Each "95" displayed a unique geographic pattern, independent of HIV prevalence, resulting in four distinct geographic clusters. Factors associated with gaps in the "95s" include younger age, male sex, and low wealth, with younger individuals having higher odds of not being on ART and having detectable viral loads. Our study revealed significant spatial heterogeneity in the HIV care continuum in Zambia, with different regions exhibiting unique geographic patterns and levels of performance in the "95-95-95" targets, highlighting the need for geospatial tailored interventions to address the specific needs of different subnational regions. These findings underscore the importance of addressing differential regional gaps in HIV diagnosis, enhancing community-level factors, and developing innovative strategies to improve local HIV care continuum outcomes.

Geospatial patterns of progress towards UNAIDS '95-95-95' targets and community vulnerability in Zambia: insights from population-based HIV impact assessments.

INTRODUCTION: In sub-Saharan Africa, HIV/AIDS remains a leading cause of death. The UNAIDS established the '95-95-95' targets to improve HIV care continuum outcomes. Using geospatial data from the Zambia Population-based HIV Impact Assessment (ZAMPHIA), this study aims to investigate geospatial patterns in the '95-95-95' indicators and individual-level determinants that impede HIV care continuum in vulnerable communities, providing insights into the factors associated with gaps. METHODS: This study used data from the 2016 ZAMPHIA to investigate the geospatial distribution and individual-level determinants of engagement across the HIV care continuum in Zambia. Gaussian kernel interpolation and optimised hotspot analysis were used to identify geospatial patterns in the HIV care continuum, while geospatial k-means clustering was used to partition areas into clusters. The study also assessed healthcare availability, access and social determinants of healthcare utilisation. Multiple logistic regression models were used to examine the association between selected sociodemographic and behavioural covariates and the three main outcomes of study. RESULTS: Varied progress towards the '95-95-95' targets were observed in different regions of Zambia. Each '95' displayed a unique geographical pattern, independent of HIV prevalence, resulting in four distinct geographical clusters. Factors associated with gaps in the '95s' include younger age, male sex, and low wealth, with younger individuals having higher odds of not being on antiretroviral therapy and having detectable viral loads. CONCLUSIONS: Our study revealed significant spatial heterogeneity in the HIV care continuum in Zambia, with different regions exhibiting unique geographical patterns and levels of performance in the '95-95-95' targets, highlighting the need for geospatial tailored interventions to address the specific needs of different subnational regions. These findings underscore the importance of addressing differential regional gaps in HIV diagnosis, enhancing community-level factors and developing innovative strategies to improve local HIV care continuum outcomes.

Temporal changes in paediatric and adolescent HIV outcomes across the care continuum in Zambia: an interrupted time-series analysis.

BACKGROUND: Paediatric and adolescent HIV treatment programmes in sub-Saharan Africa have rapidly expanded and evolved over the past decade. Real-world evidence of how the implementation of new policies over time has affected treatment outcomes is inadequate, but is crucial for guiding the implementation of the next phases of the HIV treatment response for children. We examined how treatment outcomes in Zambia's national paediatric and adolescent HIV treatment programmes have changed over time as new policies were implemented. METHODS: We used data from Zambia's routine electronic health record to analyse children and adolescents living with HIV who were antiretroviral therapy (ART) naive between the ages of 0 and 19 years who were newly enrolled in care between Jan 1, 2011, and March 31, 2019, at 102 health facilities in Lusaka and Western provinces supported by the Centre for Infectious Disease Research in Zambia. Sociodemographic factors, clinical data, facility-level data, and visit history were obtained from the national electronic health record and laboratory systems used in routine HIV care in Zambia. We aimed to characterise the changes in the distribution of the age and sex of new enrolees over time. We used an interrupted time-series design to examine the rates of ART initiation, retention in care, time to ART initiation, and first-line ART regimens among new enrolees across different age strata as they changed over time with the adoption of new ART guidelines in 2014 and 2017. FINDINGS: Between Jan 1, 2011, and March 31, 2019, 26 214 children and adolescents living with HIV who were ART naïve were newly enrolled at one of 102 ART facilities in two provinces in Zambia. Rates of new enrolees increased by 25-35% among children younger than 15 years over time, but by 92·3% between 2011 and 2017 among adolescents, with the largest absolute increase among adolescent girls. Rates of ART initiation increased steadily and in parallel across all age groups from before the implementation of the 2014 guidelines to after the implementation of the 2017 guidelines (<2 years, 42·4% for 2014 and 81·6% for 2017; 2 to <5 years, 39·3% for 2014 and 82·8% for 2017; 5 to <15 years, 49·2% for 2014 and 86·6% for 2017; 15 to 19 years, 52·4% for 2014 and 86·2% for 2017); median time to ART initiation went from 2-3 months to same-day initiation during this same time period. Rates of retention on ART 6 months after linkage saw much smaller improvements over time (<2 years, 35·4% for 2014 and 52·0% for 2017; 2 to <5 years, 40·2% for 2014 and 54·4% for 2017; 5 to <15 years, 46·7% for 2014 and 63·4% for 2017; 15 to 19 years, 40·1% for 2014 and 52·7% for 2017). INTERPRETATION: Improvements in ART initiation occurred largely in parallel across age groups over time, despite universal treatment being implemented at different timepoints for different ages. Although the rates of ART initiation reach high levels, retention on ART was low. This analysis provides a comprehensive examination of how paediatric and adolescent outcomes have evolved over the past decade in Zambia and identifies where more targeted efforts will be needed over the next decade. FUNDING: National Institutes of Health.

Mitigating the effects of COVID-19 on HIV treatment and care in Lusaka, Zambia: a before-after cohort study using mixed effects regression.

INTRODUCTION: The Zambian Ministry of Health (MoH) issued COVID-19 mitigation guidance for HIV care immediately after the first COVID-19 case was confirmed in Zambia on 18 March 2020. The Centre for Infectious Disease Research in Zambia implemented MoH guidance by: 1) extending antiretroviral therapy (ART) refill duration to 6 multi-month dispensation (6MMD) and 2) task-shifting communication and mobilisation of those in HIV care to collect their next ART refill early. We assessed the impact of COVID-19 mitigation guidance on HIV care 3 months before and after guidance implementation. METHODS: We reviewed all ART pharmacy visit data in the national HIV medical record for PLHIV in care having ≥1 visit between 1 January-30 June 2020 at 59 HIV care facilities in Lusaka Province, Zambia. We undertook a before-after evaluation using mixed-effects Poisson regression to examine predictors and marginal probability of early clinic return (pharmacy visit >7 days before next appointment), proportion of late visit (>7 days late for next appointment) and probability of receiving a 6MMD ART refill. RESULTS: A total of 101 371 individuals (64% female, median age 39) with 130 486 pharmacy visits were included in the analysis. We observed a significant increase in the adjusted prevalence ratio (4.63; 95% CI 4.45 to 4.82) of early return before compared with after guidance implementation. Receipt of 6MMD increased from a weekly mean of 47.9% (95% CI 46.6% to 49.2%) before to 73.4% (95% CI 72.0% to 74.9%) after guidance implementation. The proportion of late visits (8-89 days late) was significantly higher before (18.8%, 95% CI17.2%to20.2%) compared with after (15.1%, 95% CI13.8%to16.4%) guidance implementation . CONCLUSIONS: Timely issuance and implementation of COVID-19 mitigation guidance involving task-shifted patient communication and mobilisation alongside 6MMD significantly increased early return to ART clinic, potentially reducing interruptions in HIV care during a global public health emergency.

Effects of implementing universal and rapid HIV treatment on initiation of antiretroviral therapy and retention in care in Zambia: a natural experiment using regression discontinuity.

BACKGROUND: Universal testing and treatment (UTT) for all people living with HIV has only been assessed under experimental conditions in cluster-randomised trials. The public health effectiveness of UTT policies on the HIV care cascade under real-world conditions is not known. We assessed the real-world effectiveness of universal HIV treatment policies that were implemented in Zambia on Jan 1, 2017. METHODS: We used data from Zambia's routine electronic health record system to analyse antiretroviral therapy (ART)-naive adults who newly enrolled in HIV care up to 1 year before and after the implementation of universal treatment (ie, Jan 1, 2016, to Jan 1, 2018) at 117 clinics supported by the Centre for Infectious Disease Research in Zambia. We used a regression discontinuity design to estimate the effects of implementing UTT on same-day ART initiation, ART initiation within 1 month, and retention on ART at 12 months (defined as clinic attendance 9-15 months after enrolment and at least 6 months on ART), under the assumption that patients presenting immediately before and after UTT implementation were balanced on both measured and unmeasured characteristics. We did an instrumental variable analysis to estimate the effect of same-day ART initiation under routine conditions on 12-month retention on ART. FINDINGS: 65 673 newly enrolled patients with HIV (40 858 [62·2%] female, median age 32 years [IQR 26-39], median CD4 count 287 cells per µL [IQR 147-466]) were eligible for inclusion in the analyses; 31 145 enrolled before implementation of UTT, and 34 528 enrolled after UTT. Implementation of universal treatment increased same-day ART initiation from 41·7% to 74·8% (risk difference [RD] 33·1%, 95% CI 30·5-35·7), ART initiation by 1 month from 69·6% to 87·0% (RD 17·4%, 15·5-19·3), and 12-month retention on ART from 56·2% to 63·3% (RD 7·1%, 4·3-9·9). ART initiation rates became more uniform across patient subgroups after implementation of universal treatment, but heterogeneity in 12-month retention on ART between subgroups was unchanged. Instrumental variable analyses indicated that same-day ART initiation in routine settings led to a 15·8% increase (95% CI 12·1-19·5) in 12-month retention on ART. INTERPRETATION: UTT policies implemented in Zambia increased the rapidity and uptake of ART, as well as retention on ART at 12 months, although overall retention on ART remained suboptimal. UTT policies reduced disparities in treatment initiation, but not 12-month retention on ART. Natural experiments reveal both the anticipated and unanticipated effects of real-world implementation and indicate the need for new strategies leveraging the short-term effects of UTT to cultivate long-term treatment success. FUNDING: National Institutes of Health.

Patient-reported reasons for declining same-day antiretroviral therapy initiation in routine HIV care settings in Lusaka, Zambia: results from a mixed-effects regression analysis.

INTRODUCTION: In the current "test and treat" era, HIV programmes are increasingly focusing resources on linkage to care and same-day antiretroviral therapy (ART) initiation to meet UNAIDS 95-95-95 targets. After observing sub-optimal treatment indicators in health facilities supported by the Centre for Infectious Disease Research in Zambia (CIDRZ), we piloted a "linkage assessment" tool in facility-based HIV testing settings to uncover barriers to same-day linkage to care and ART initiation among newly identified people living with HIV (PLHIV) and to guide HIV programme quality improvement efforts. METHODS: The one-page, structured linkage assessment tool was developed to capture patient-reported barriers to same-day linkage and ART initiation using three empirically supported categories of barriers: social, personal and structural. The tool was implemented in three health facilities, two urban and one rural, in Lusaka, Zambia from 1 November 2017 to 31 January 2018, and administered to all newly identified PLHIV declining same-day linkage and ART. Individuals selected as many reasons as relevant. We used mixed-effects logistic regression modelling to evaluate predictors of citing specific barriers to same-day linkage and ART, and Fisher's Exact tests to assess differences in barrier citation by socio-demographics and HIV testing entry point. RESULTS: A total of 1278 people tested HIV positive, of whom 126 (9.9%) declined same-day linkage and ART, reporting a median of three barriers per respondent. Of these 126, 71.4% were female. Females declining same-day ART were younger, on average, (median 28.5 years, interquartile range (IQR): 21 to 37 years) than males (median 34.5 years, IQR: 26 to 44 years). The most commonly reported barrier category was structural, "clinics were too crowded" (n = 33), followed by a social reason, "friends and family will condemn me" (n = 30). The frequency of citing personal barriers differed significantly across HIV testing point (χ2 p = 0.03). Significant predictors for citing ≥1 barrier to same-day ART were >50 years of age (OR: 12.59, 95% CI: 6.00 to 26.41) and testing at a rural facility (OR: 9.92, 95% CI: 4.98 to 19.79). CONCLUSIONS: Given differences observed in barriers to same-day ART initiation reported across sex, age, testing point, and facility type, new, tailored counselling and linkage to care approaches are needed, which should be rigorously evaluated in routine programme settings.

Nonadherence to antiretroviral therapy among HIV-infected patients in Zambia is concentrated among a minority of patients and is highly variable across clinics.

OBJECTIVES: The distribution of adherence to antiretroviral therapy (ART) can indicates whether barriers are concentrated or more distributed. We quantified the medication possession ratio (MPR) and characterized the distribution of medication nonpossession in a network of clinics in Zambia to identify 'hotspots' and predictors of poorer adherence. METHODS: We analyzed a population of adults on ART for more than 3 months who made at least one clinic visit between 1 January 2013 and 28 February 2015. Pharmacy refill and clinical information were obtained through the electronic medical record system used in routine care. We constructed a Lorenz curve to visualize the distribution of poor adherence and used a multilevel logistic regression model to examine factors associated with MPR. RESULTS: Among 131 767 patients in 56 clinics [64% women, median age 34 years (interquartile range (IQR) 29-41), median CD4 cell count at ART initiation 351 cells/µl (IQR 220-517)], the median MPR was 85.8% (IQR 70.8-96.8). During months 7-12 on ART, 45.6% of patients had 100% MPR and 10.5% accounted for 50% of medication nonpossession. Across clinics, median MPR ranged from 49.1 to 98.5, and clinic accounted for 12% of the variability in adherence after adjusting for individual and clinic-level characteristics. CONCLUSION: A small fraction of patients account for the majority of days of medication nonpossession. Further characterization of these subpopulations is needed to target interventions. Clinic also accounted for much variability in MPR. Health systems interventions targeting clinic 'hot spots' may represent an efficient use of resources to improve ART adherence.