1. Background

Pneumonia remains a leading cause of death in children under five, and survival depends on prompt, appropriate care [1]. In rural settings, most delay occurs before the child reaches a facility: caregivers may not recognise danger signs, and referral is often not completed [2]. Community health workers (CHWs), equipped with a structured triage aid, may shorten this pre-facility delay. We evaluated such a programme across 60 villages in western Kenya.

2. Methods

CHWs used a symptom-based triage aid aligned with WHO integrated management of childhood illness danger signs [3], together with referral vouchers that covered transport. We recorded, for each episode, the time from symptom onset to appropriate care and whether the referral was completed, in the 6 months before and the 6 months after rollout.

We summarise the effect on delay as the relative reduction

$$R = \frac{t_{\text{before}} - t_{\text{after}}}{t_{\text{before}}},$$

and report referral completion as a proportion. A mixed-effects model with a village random intercept was used to account for clustering.

3. Results

Median time to appropriate care fell from 52 to 29 hours, a relative reduction of $R = 0.44$, and referral completion rose from 48 to 74 percent (Table 1). The improvement was largest where CHWs received at least monthly supervisory contact.

4. Discussion and conclusion

A simple triage aid coupled with referral support substantially shortened the pre-facility delay that drives poor pneumonia outcomes, and raised referral completion. As a before-after design, the study cannot fully exclude secular trends, and outcomes were caregiver-reported; a stepped-wedge trial would strengthen causal claims. Even so, the size and consistency of the effect, and its dependence on supervision, support scaling structured CHW triage with adequate supervisory contact.