Impact of integrating traditional care with the modern healthcare system in reducing tuberculosis diagnosis delays in Ethiopia: a clustered randomized controlled study

Abstract

Background Diagnosis and treatment initiation delays for tuberculosis (TB) are significant challenges in resource-limited settings. These delays can result in poor treatment outcomes, disease transmission, and increased costs. This study aimed to assess the effect of integrating traditional care with modern healthcare systems on reducing TB diagnosis delay. Methods A cluster randomized controlled trial was conducted among TB patients, with 510 participants, 255 individuals were assigned to the intervention group and 255 to the control group. Training in the intervention group was provided for both traditional and modern healthcare providers in three rounds to enhance their knowledge, attitudes, and skills in TB screening and referral. A non-parametric independent sample test was used to compare the baseline and end-line data. The effect size was determined using Cohen’s d. To account for individual and cluster-level variations, a mixed-effect parametric survival model was employed. Furthermore, conditional (fixed only) and marginal (random effects) graphs were used to compare between the intervention and control groups. Results A total of 510 participants were included in the baseline study, with a similar number of participants included in the endline study. In the intervention group, the delay in diagnosis was 4.185 per 1000 person-days post-intervention, compared to 4.608 per 1000 person-days pre-intervention. In the control group, the delay for diagnosis was 4.759 per 1000 person-days pre-intervention and 5.031 per 1000 person-days post-intervention. The median time to diagnosis was 135 days. The non-parametric comparison showed that the intervention significantly reduced patient delays in the intervention group compared to the control group (p = 0.006), with a Cohen's d effect size of 0.246. The intervention also significantly reduced diagnosis delay in the intervention group compared to the control group (p = 0.036), with a Cohen's d effect size of 0.187. The diagnosis of TB was accelerated by 1.076 times due to the integration of traditional care with the modern healthcare system in the intervention group compared to the control group (δ: 1.076; 95% CI 1.021, 1.134). Conclusions The involvement of traditional care providers in TB control programs significantly reduced diagnosis delays in Ethiopia. These findings suggest the need for integrating traditional care with modern healthcare systems for the effective prevention of TB in high-burden countries. Clinical trial registration ClinicalTrials.gov ID: NCT05236452.