Effect of Patient Support Groups for Hypertension on Blood Pressure among Patients with and Without Multimorbidity: Findings from a Cohort Study of Patients on a Home-Based Self-Management Program in Kenya.

INTRODUCTION

Patient support group interventions have been widely used to manage chronic diseases in Kenya. However, the potential benefits of these groups on patient health outcomes, and how this is influenced by multimorbidity, have not been rigorously evaluated.

OBJECTIVE

We assessed the effect of a patient support group intervention on blood pressure (BP) management and the potential moderating effect of multimorbidity among low- and middle-income patients with hypertension in Kenya.

METHODS

We analysed data from a non-randomized, quasi-experimental study of 410 patients with hypertension on a home-based self-management program conducted from September 2019 to September 2020. The program included the formation and participation in patient support groups. Using a modified STEPS questionnaire, data were collected on BP, anthropometry and other measurements at enrolment and after 12 months of follow-up. Multimorbidity was defined as the simultaneous presence of hypertension and at least one or more related conditions with similar pathophysiology (concordant multimorbidity) or unrelated chronic conditions (discordant multimorbidity). Propensity score (PS) weighting was used to adjust for baseline differences among 243 patients who participated in the support groups and 167 who did not. We estimated the effects of patient support groups and moderating effects of multimorbidity on BP management using multivariable ordinary linear regression weighted by PS.

FINDINGS: Participation in support groups significantly reduced systolic BP by 5.4 mmHg compared to non-participation in the groups [β = -5.4; 95% CI -1.9 to -8.8]. However, among participants in the support group intervention, the mean systolic BP at follow-up assessment for those with concordant multimorbidity was 8.8 mmHg higher than those with no multimorbidity [β = 8.8; 95% CI 0.8 to 16.8].

CONCLUSION

Although patient support groups are potentially important adjuncts to home-based self-care, multimorbidity attenuates their effectiveness. There is a need to tailor patient support group interventions to match the needs of the people living with multimorbidity in low- and middle-income settings in Kenya.