An Interrupted Time Series Analysis Using Segmented Regression in Evaluating the Efficacy of Public Health Interventions in Kilifi County

Abstract

Background: Public health interventions may affect a variety of health outcomes. This study developed an Interrupted Time Series model to test its efficacy in evaluating public health interventions. The developed model can be used to forecast future trends in interventions to curb pneumonia. Methods: This study utilized interrupted time-series analysis (ITS) as the study design. The study population comprised children between two months and five years admitted to Kilifi County Hospital from May 2007 to March 2020. The population included a cohort that received the PCV10 vaccine that was introduced in January 2011 for three months. Results: The study findings indicated a downward trajectory with regard to the number of pneumonia cases reported. Further, the segmented regression results show that the intercept (β0) = 823.16, coefficient estimate of time (β1) = -2.72, coefficient estimate of PCV10 intervention (β2) = 59.63, and the coefficient estimate of the time after PCV10 intervention (β3) = -6.03. In addition, the results showed that during the post-intervention period, the response variable had an average value of approximately. 422.02. The 95% interval of this counterfactual prediction is [669.64, 821.18]. Therefore, the adverse effects observed during the intervention period are statistically significant. Conclusion: The overall findings of the segmented regression model imply that public health initiatives in Kilifi County have been successful in enhancing population health outcomes. The study recommends using PCV10 vaccination as an intervention for longevity of good health and reducing the number of pneumonia cases among children under five in Kenya.

Keywords

• interrupted time series
• segmented regression
• pneumonia
• efficacy and public health interventions