Evaluating the implementation of the Primary Health Integrated Care Project for Chronic Conditions: a cohort study from Kenya

In Kenya, non-communicable diseases (NCDs) are estimated to account for almost one-third of all deaths and this is likely to rise by over 50% in the next 10 years. The Primary Health Integrated Care for Chronic Conditions (PIC4C) project aims to strengthen primary care by integrating comprehensive NCD care into existing HIV primary care platform. This paper evaluates the association of PIC4C implementation on clinical outcomes.Outcomes included proportion of new patients, systolic blood pressure (SBP), fasting plasma glucose (FPG), diastolic blood pressure, hypertension control, random plasma glucose, diabetes control, viral load and HIV viral suppression. We used interrupted time series and binomial regression with random effects for facility-level data and generalised mixed-effects regression for visit-level data to examine the association between PIC4C and outcomes between January 2017 and December 2021. We conducted sensitivity analysis with restrictions on sites and the number of visits.Data from 66 641 visits of 13 046 patients with hypertension, 24 005 visits of 7267 patients with diabetes and 84 855 visits of 21 186 people with HIV were analysed. We found evidence of association between PIC4C and increase in proportion of new patients per month with hypertension (adjusted OR (aOR) 1.57, 95% CI 1.39 to 1.78) and diabetes (aOR 1.31, 95% CI 1.19 to 1.45), small increase in SBP (adjusted beta (aB) 1.7, 95% CI 0.8 to 2.7) and FPG (aB 0.6, 95% CI 0.0 to 1.1). There was no strong evidence of association between PIC4C and viral suppression (aOR 1.20, 95% CI 0.98 to 1.47). In sensitivity analysis, there was no strong evidence of association between PIC4C and SBP (aB 1.74, 95% CI −0.70 to 4.17) or FPG (aB 0.52, 95% CI −0.64 to 1.67)PIC4C implementation was associated with increase in proportion of new patients attending clinics and a slight increase in SBP and FPG. The immediate post-PIC4C implementation period coincided with the COVID-19 pandemic, which is likely to explain some of our findings.