Abstract
Quality assessment in family planning (FP) and post-abortion care (PAC) programmes is shaped by workforce capability, programme context, and assessment modality. Longitudinal evaluations may misattribute quality-score differences if repeated measurements and temporal patterning are ignored. To examine associations of workforce characteristics, calendar period, service-delivery channel, and assessment modality with Quality Technical Assessment (QTA) performance in Sierra Leone, 2019-2022. We analysed 108 QTA records from 23 service delivery points using Gaussian generalised estimating equations with facility-level clustering and robust standard errors. Models included year, modality, channel, active staff count, and core-service competency coverage; full-time staffing and assessor identity were examined in sensitivity analyses. In 2021, all facilities received planned in-person and then remote assessments in a fixed programme-wide sequence. Mean QTA scores increased descriptively from 89.8% (SD 7.4) in 2019 to 94.2% (SD 4.7) in 2022. In the primary model, 2020 scores were lower than 2019 (B = -5.15; 95% CI -9.99 to -0.32). Active staff count was inversely associated with QTA scores (B = -2.86; 95% CI -4.47 to -1.26). Remote assessment was associated with higher adjusted scores, but this cannot be interpreted causally because modality was confounded with year and assessment order. QTA performance varied across calendar periods and workforce structure. Longitudinal evaluations should account for within-facility correlation and modality timing before attributing quality differences to the workforce or assessment approach.