# Evaluation of Midwife‐Led Colposcopy for Female Genital Schistosomiasis Screening at Primary Level of Care in Rural Madagascar: A Cross‐Sectional Study

## TL;DR

Midwives in rural Madagascar can effectively detect genital schistosomiasis with high accuracy, but their results vary depending on training and environment.

## Contribution

This study evaluates midwife-led colposcopy for genital schistosomiasis in low-resource settings, showing its feasibility with needed quality measures.

## Key findings

- Midwife-led colposcopy showed high sensitivity (96.4%) but low specificity (28.7%) for detecting genital schistosomiasis.
- Midwife experience and healthcare environment significantly influenced diagnostic accuracy and agreement with gynaecologists.
- The study suggests midwife-led screening is feasible but requires quality assurance to ensure reliable results.

## Abstract

Female genital schistosomiasis is a condition with a complex diagnosis and severe consequences such as infertility. In the absence of a reliable biomarker, in endemic settings the World Health Organization recommends colposcopy as a diagnostic tool for the detection of female genital schistosomiasis lesions. Nevertheless, it is seldom performed in low‐resource contexts due to a lack of expertise or insufficient infrastructure. This study aims to assess Female Genital Schistosomiasis colposcopy at the primary level of care, evaluating its diagnostic accuracy in reference to gynaecologist diagnosis in a highly endemic context.

This is a secondary analysis of a cross‐sectional study conducted in the Boeny region of Madagascar, which collected colposcopy images and Female Genital Schistosomiasis decision at the primary health care level with re‐evaluation by gynaecologists. Statistical analysis using R included descriptive statistics, measures of diagnostic accuracy with 95% confidence intervals and binary Poisson regression with robust standard errors, while reporting followed the STARD statement.

Among 495 included participants, a high sensitivity [96.4% (95% CI 93.7–98.0)] and relatively low specificity [28.7% (95% CI 21.8–36.5)], with a fair agreement [κ 0.30 (95% CI 0.22–0.39)], was observed for midwife‐led colposcopy. Practice of midwives (3.5 months) was associated with reduced concordance [APR 0.88 (95% CI 0.79–0.98)] and specificity [APR 0.27 (95% CI 0.15–0.49)]. The environment of one health care centre negatively influenced concordance and specificity of midwife‐led colposcopy.

Midwives can detect female genital schistosomiasis with high sensitivity but limited specificity when compared to expert gynaecologists, revealing variation in performance between environments as well as the influence of practice and workload. This study suggests that implementing midwife‐led colposcopy at primary care level for female genital schistosomiasis screening is feasible but requires appropriate quality assurance measures.