Research ObjectivePoint‐of‐care (POC) assays for early infant diagnosis of HIV (EID) increase access to testing, shorten time to results, and expedite initiation of antiretroviral therapy when compared with laboratory‐based (centralized) assays. However, there is a significant gap in our understanding of its human resource impact at the facility‐level. This study evaluated the human resource labor time associated with the use of EID platforms by facility‐based health workers (HWs).Study DesignUsing time‐motion methodology, we collected direct observation and self‐reported time‐use data on EID tasks performed by HWs at health facilities in Zimbabwe. Data collection occurred at three EID facility types in Zimbabwe—five POC hubs (test samples from the hub and lower‐volume, nearby facilities), nine POC spokes (lower‐volume), and eleven facilities that use centralized laboratories.EID activities were grouped into pre‐result (counseling, administrative, blood sampling, and preparation) and total EID time (counseling to result communication). Total EID time was observed in POC hub facilities. Since EID tests were not performed on‐site at facilities using centralized laboratories and POC spoke facilities, tasks related to logging and communicating results were not observed. Instead, direct observation data were complemented with HW surveys at those facilities.Mean difference comparisons between centralized and POC EID services were performed with a two‐sample t test with allowance for unequal variances. Bootstrapping with 10 000 replicates was used to estimate the bias‐corrected and accelerated 95% confidence intervals for each measure and difference in means test. To assess whether results were robust, group comparison was performed with two‐sided permutation tests with 10 000 replicates, with same inference to those from the bootstrapped two‐sample t tests. A P‐value of <0.05 was considered statistically significant.Population StudiedFacility‐based health workers in Zimbabwe.Principal FindingsData collectors observed 30 EID processes and 30 HWs provided self‐reported time. Observed average total labor time per EID test at POC hubs was 28 minutes, 22 seconds (95% CI, 22:51‐35:48). HWs performed other tasks while the POC EID machine processed assays.Observed average pre‐result time (counseling to sample preparation) was 18 minutes, 6 seconds (95% CI, 13:00‐23:42) for POC compared with 27 minutes, 48 seconds (95% CI, 23:48‐32:50) for facilities using centralized laboratories. The mean difference of 9 minutes, 42 seconds (95% CI, 03:04‐16:18) was statistically significant. The major driver of this difference was the longer time to prepare dried blood spots (DBS) compared with POC assays.Self‐reported average labor time per negative EID was 49 minutes, 30 seconds (95% CI, 0:42:56‐0:57:08) for facilities using centralized laboratories compared with 53 minutes, 4 seconds (95% CI, 0:40:24‐1:04:44) for POC assays. The differences for self‐reported time were not statistically significant, and results were similar for positive EIDs.ConclusionsPOC EID did not incur significant additional facility‐based human resource time compared to centralized EID.Implications for Policy or PracticeThis study, combined with previously published data demonstrating that POC EID significantly improves timely ART initiation and is cost‐effective, furthers the evidence that POC EID is a feasible intervention for low‐resource contexts, where health facilities are short‐staffed.Primary Funding SourceUNITAID.
Read full abstract