Cervical cancer screening is an established preventive strategy that depends on timely and consistent follow-up screenings to detect precancerous changes. This study aimed to measure adherence to recommended follow-up cervical cancer screening, describe the timing of adherence, and identify predictors of adherence among women living with HIV (WLHIV) who had their first screening in Moshi municipality between 2019 and 2021.
The investigation was a hospital-based retrospective cohort study conducted between 2019 and 2021. Data were collected from healthcare facilities in Moshi municipality, Kilimanjaro region, northern Tanzania. The study focused on real-world adherence patterns after initial screening in routine facility and outreach contexts.
The cohort comprised 2,030 WLHIV who underwent their first cervical cancer screening in healthcare facilities in Moshi municipality during the study period. The primary outcome was adherence to a follow-up cervical cancer screening, defined as a subsequent screening test performed after the first screening to detect precancerous cervical changes.
For a sub-analysis on timing, adherence was categorised into four groups based on the interval between the first screening and the subsequent screening: early adherence, good adherence, poor adherence, and extremely poor adherence. The study specifically reported the proportion meeting the good adherence criterion, defined as attending the follow-up within one month after the scheduled date.
Predictors of adherence were assessed using multivariable Poisson regression models with robust standard errors. Results were reported as adjusted risk ratios (ARR) with 95% confidence intervals (CIs). This analytic approach estimated relative differences in adherence associated with patient and facility characteristics.
Among the 2,030 WLHIV who completed a first screening, only 33.2% adhered to a follow-up screening. Within the subset who adhered, only 21.4% (95% CI 18.4% to 24.7%) achieved good adherence, meaning they attended their follow-up within one month after the scheduled date. These findings indicate that both the overall rate of subsequent screening and the punctuality of that screening were suboptimal in this cohort.
The multivariable analysis identified several important predictors:
Facility type: Adherence was higher in private healthcare facilities (ARR = 1.57; 95% CI 1.35 to 1.82) compared with the reference category. By contrast, women whose first screening occurred at outreach facilities showed substantially lower adherence (ARR = 0.28; 95% CI 0.19 to 0.39).
First screening result: Women who received an abnormal screening outcome at their first screening were more likely to attend follow-up (ARR = 2.62; 95% CI 2.26 to 3.05) than those with normal results.
Demographics and parity: Adherence increased with age and with higher parity. Conversely, younger women and those with no or fewer children had lower follow-up rates.
These associations quantify which subgroups are more or less likely to obtain recommended follow-up screening after an initial test.
Adherence to recommended follow-up cervical cancer screening among WLHIV in Moshi municipality was suboptimal, with only about one-third of women returning for subsequent screening. Timely adherence was also limited, as reflected by a small proportion meeting the predefined good adherence window.
To improve follow-up rates, the study highlights the need to strengthen recall systems, particularly in public and outreach facilities where adherence was lowest. Targeted interventions should address younger WLHIV, women with no or fewer children, and those with normal screening outcomes at first screening, as these groups were less likely to attend follow-up. The findings point to health-service and patient-level factors that could be prioritized to increase adherence to follow-up cervical cancer screening in this high-risk population.
Note: The source provided summary results, adjusted risk ratios, confidence intervals, study period and setting. Additional methodological details and subgroup estimates were not reported in the provided source text.