Implementing at-birth, point-of-care HIV testing in Kenya: a qualitative study using the Consolidated Framework for Implementation Research.

BACKGROUND: At-birth and point-of-care (POC) testing can expedite early infant diagnosis of HIV and improve infant outcomes. Guided by the Consolidated Framework for Implementation Research (CFIR), this study describes the implementation of an at-birth POC testing pilot from the perspective of implementing providers and identifies the factors that might support and hinder the scale up of these promising interventions. METHODS: We conducted 28 focus group discussions (FGDs) with 48 providers across 4 study sites throughout the course of a pilot study assessing the feasibility and impact of at-birth POC testing. FGDs were audio-recorded, transcribed, and analyzed for a priori themes related to CFIR constructs. This qualitative study was nested within a larger study to pilot and evaluate at-birth and POC HIV testing. RESULTS: Out of the 39 CFIR constructs, 30 were addressed in the FGDs. While all five domains were represented, major themes revolved around constructs related to intervention characteristics, inner setting, and outer setting. Regarding intervention characteristics, the advantages of at-birth POC (rapid turnaround time resulting in improved patient management and enhanced patient motivation) were significant enough to encourage provider uptake and enthusiasm. Challenges at the intervention level (machine breakdown, processing errors), inner settings (workload, limited leadership engagement, challenges with access to information), and outer setting (patient-level challenges, limited engagement with outer setting stakeholders) hindered implementation, frustrated providers, and resulted in missed opportunities for testing. Providers discussed how throughout the course of the study adaptations to implementation (improved channels of communication, modified implementation logistics) were made to overcome some of these challenges. To improve implementation, providers cited the need for enhanced training and for greater involvement among stakeholders outside of the implementing team (i.e., other clinicians, hospital administrators and implementing partners, county and national health officials). Despite provider enthusiasm for the intervention, providers felt that the lack of engagement from leadership within the hospital and in the outer setting would preclude sustained implementation outside of a research setting. CONCLUSION: Despite demonstrated feasibility and enthusiasm among implementing providers, the lack of outer setting support makes sustained implementation of at-birth POC testing unlikely at this time. The findings highlight the multi-dimensional aspect of implementation and the need to consider facilitators and barriers within each of the five CFIR domains. TRIAL REGISTRATION: ClinicalTrials.gov, NCT03435887 . Retrospectively registered on 19 February 2020.

Piloting the Feasibility and Preliminary Impact of Adding Birth HIV Polymerase Chain Reaction Testing to the Early Infant Diagnosis Guidelines in Kenya.

BACKGROUND: In Kenya, standard early infant diagnosis (EID) with polymerase chain reaction (PCR) testing at 6-week postnatal achieves early treatment initiation (<12 weeks) in <20% of HIV+ infants. Kenya's new early infant diagnosis guidelines tentatively proposed adding PCR testing at birth, pending results from pilot studies. METHODS: We piloted birth testing at 4 Kenyan hospitals between November 2017 and November 2018. Eligible HIV-exposed infants were offered both point-of-care and PCR HIV testing at birth (window 0 to <4 weeks) and 6 weeks (window 4-12 weeks). We report the: proportion of infants tested at birth, 6-week, and both birth and 6-week testing; median infant age at results; seropositivity and antiretroviral therapy initiation. RESULTS: Final sample included 624 mother-infant pairs. Mean maternal age was 30.4 years, 73.2% enrolled during antenatal care and 89.9% had hospital deliveries. Among the 590 mother-infants pairs enrolled before 4 weeks postnatal, 452 (76.6%) completed birth testing before 4 weeks, with 360 (79.6%) testing within 2 weeks, and 178 (39.4%) before hospital discharge (0-2 days). Mothers were notified of birth PCR results at a median infant age of 5.4 weeks. Among all 624 enrolled infants, 575 (92.1%) were tested during the 6-week window; 417 (66.8%) received testing at both birth and 6-weeks; and 207 received incomplete testing (93.3% only 1 PCR and 6.7% no PCR). Four infants were diagnosed with HIV, and 3 infants were initiated on antiretroviral therapy early, before 12 weeks of age. CONCLUSIONS: Uptake of PCR testing at birth was high and a majority of infants received repeat testing at 6 weeks of age.

ART Initiation for Infants Diagnosed With HIV Through Point of Care and Conventional Polymerase Chain Reaction Testing in Kenya: A Case Series.

We sought to understand the sequence of testing and treatment among nine infants offered both conventional and point-of-care testing and diagnosed as HIV-positive by 6 months of age in Kenya. One infant received per protocol testing and treatment. Patient-level (late presentation and disengagement), provider-level (reluctance and error/oversight) and system-level (stock outs, errors) challenges delayed diagnosis and treatment. Early point-of-care testing can streamline testing; however, challenges mitigate benefits.

Programmatic evaluation of feasibility and efficiency of at birth and 6-week, point of care HIV testing in Kenyan infant.

BACKGROUND: Testing infants at birth and with more efficient point of care (POC) HIV diagnostic can streamline EID and expedite infant ART initiation. We evaluated the implementation of at birth and 6-week POC testing to assess the effectiveness and feasibility when implemented by existing hospital staff in Kenya. METHODS: Four government hospitals were randomly assigned to receive a GeneXpert HIV-1 Qual (n = 2) or Alere m-PIMA (n = 2) machine for POC testing. All HIV-exposed infants enrolled were eligible to receive POC testing at birth and 6-weeks of age. The primary outcome was repeat POC testing, defined as testing both at birth and 6-weeks of age. Secondary outcomes included predictors of repeat POC testing, POC efficiency (turnaround times of key services), and operations (failed POC results, missed opportunities). RESULTS: Of 626 enrolled infants, 309 (49.4%) received repeat POC testing, 115 (18.4%) were lost to follow up after an at-birth test, 120 (19.2%) received POC testing at 6-weeks only, 80 (12.8%) received no POC testing, and 2 (0.3%) received delayed POC testing (>12 weeks of age). Three (0.4%) were identified as HIV-positive. Of the total 853 POC tests run at birth (n = 424) or 6-weeks (n = 429), 806 (94.5%) had a valid result documented and 792 (98.3%) results had documented maternal notification. Mean time from sample collection to notification was 1.08 days, with 751 (94.8%) notifications on the same day as sample collection. Machine error rates at birth and 6-weeks were 8.5% and 2.5%, respectively. A total of 198 infants presented for care (48 at birth; 150 at 6-weeks) without receiving a POC test, representing missed opportunities for testing. DISCUSSION: At birth POC testing can streamline infant HIV diagnosis, expedite ART initiation and can be implemented by existing hospital staff. However, maternal disengagement and missed opportunities for testing must be addressed to realize the full benefits of at birth POC testing.

Implementation planning for community-based point-of-care HIV testing for infants: Recommendations from community leaders in Kenya.

BACKGROUND: Early infant diagnosis (EID) establishes the presence of HIV infection in HIV-exposed infants and children younger than 18 months of age. EID services are hospital-based, and thus fail to capture HIV-exposed infants who are not brought to the hospital for care. Point-of-care (POC) diagnostic systems deployed in the community could increase the proportion tested and linked to treatment, but little feasibility and acceptability data is available. METHODS: Semi-structured interviews (n = 74) were conducted by a Kenyan team with community members (Community Health Workers/Volunteers [CHW/CHV], Traditional Birth Attendants [TBAs], community leaders) and parents of HIV-exposed infants at four study sites in Kenya to elicit feedback on the acceptability and feasibility of community-based POC HIV testing. RESULTS: Participants described existing community health resources that could be leveraged to support integration of community-based POC HIV testing; however, the added demand placed on CHW/CHV could pose a challenge. Participants indicated that other potential barriers (concerns about confidentiality, disclosure, and HIV stigma) could be overcome with strong engagement from trusted community leaders and health providers, community sensitization, and strategic location and timing of testing. These steps were seen to improve acceptability and maximize the recognized benefits (rapid results, improved reach) of community-based testing. CONCLUSION: Community members felt that with strategic planning and engagement, community-based POC HIV testing could be a feasible and acceptable strategy to overcome the existing barriers of hospital-based EID.

Maternal viral load monitoring: Coverage and clinical action at 4 Kenyan hospitals.

BACKGROUND: Kenya's guidelines for prevention of mother-to-child transmission of HIV (PMTCT) recommend routine viral load (VL) monitoring for pregnant and breastfeeding women. METHOD: We assessed PMTCT VL monitoring and clinical action occurring between last menstrual period (LMP) and 6 months postpartum at 4 Kenyan government hospitals. Pregnant women enrolled in the HIV Infant Tracking System from May 2016-March 2018 were included. We computed proportions who received VL testing within recommended timeframes and who received clinical action after unsuppressed VL result. RESULTS: Of 424 participants, any VL testing was documented for 305 (72%) women and repeat VL testing was documented for 79 (19%). Only 115 women (27%) received a guideline-adherent baseline VL test and 27 (6%) received a guideline-adherent baseline and repeat VL test sequence. Return of baseline and repeat VL test results to the facility was high (average 96%), but patient notification of VL results was low (36% baseline and 49% repeat). Clinical action for unsuppressed VL results was even lower: 11 of 38 (29%) unsuppressed baseline results and 2 of 14 (14%) unsuppressed repeat results triggered clinical action. DISCUSSION: Guideline-adherent VL testing and clinical intervention during PMTCT must be prioritized to improve maternal care and reduce the risk of HIV transmission to infants.