Implementation fidelity of the cervical cancer screening guidelines for HIV-infected women in Ekurhuleni, South Africa

Abstract

Background: The South African national cervical screening guideline contains recommendations to guide the screening of HIV-infected women. However, despite the availability of this evidence-based guideline, the cervical cancer screening coverage remains low amongst HIV-infected women. Whether health workers adhere to this guideline has not been determined. This study, therefore, aims to assess health workers’ adherence to the national guideline on cervical cancer screening of HIV-infected women (fidelity of implementation), and evaluate factors associated with fidelity level. Methods: This was a cross-sectional study with an analytical component. It was conducted in primary healthcare facilities in the Ekurhuleni Metropolitan Municipality in 2019. A survey of 177 primary healthcare providers was conveniently sampled from primary healthcare facilities in Ekurhuleni health sub-districts North (N1) and South (S2). A self-administered questionnaire collected data on four implementation fidelity constructs as well as data on potential determinants of fidelity (provider characteristics, guideline characteristics, system characteristics, and implementation characteristics). Fidelity level was calculated as a score derived by summing the scores from each of the four constructs. The score was converted to a percentage (maximum possible score 100). Exploratory factor analysis was used to categorise the fidelity score into a binary variable (high and low fidelity). Logistic regression was used to analyse for an association between each determinant and the implementation fidelity level, using odds ratios as the measure of association. A generalised structural equation model (GSEM) was used to investigate the indirect relationships between the determinants and implementation fidelity of the cervical screening of HIV-infected women by the healthcare providers. This model adjusted for random effects at the facility level, using a mixed-effects modelling approach.Results: The median implementation fidelity percentage score was 84.85, with an interquartile range of 75.76 – 87.88. In this study, after categorising fidelity, 62.15% of primary healthcare workers (PHCWs) were seen to have high fidelity, while the remaining 37.85% had low fidelity. The median (IQR) provider characteristics per cent score, guideline characteristics per cent score, system characteristics per cent score and implementation characteristics per cent score were 85.56 (80 – 92.22), 86.15 (76.92 –92.31), 76.67 (66.67 – 86.67) and 60 (53.33 – 80), respectively. In the logistic regression analysis – unadjusted model, the provider characteristics (OR 1.07; 95% CI 1.04 to 1.11), guideline characteristics (OR 1.05; 95% CI 1.02 to 1.08), and the participants’ level of knowledge of the updated guidelines (OR 1.04; 95% CI 1.02 to 1.06) positively influenced implementation fidelity. In the logistic regression analysis – adjusted model, the provider characteristics (OR 1.05; 95% CI 1.01 to 1.10), and the participants’ level of knowledge of the updated guidelines (OR 1.02; 95% CI 1.00 to 1.06) were statistically significantly associated with implementation fidelity. From the GSEM path analysis, having the provider characteristic acting as an endogenous (outcome) variable, a significant indirect relationship was found between implementation fidelity and determinants such as; the guideline characteristics (OR 1.03; 95% CI 1.01-1.04), the participant’s level of knowledge of the updated guidelines (OR 1.01; 95% CI 1.01-1.02).Conclusion This study revealed a high level of primary healthcare workers’ fidelity of implementation of the guidelines for the cervical cancer screening of HIV-infected women. Key determinants such as provider characteristics and guideline characteristics were found to be important factors influencing the level of implementation fidelity. These factors should be considered when developing future implementation strategies for improving the level of fidelity.