PHS95 - Developing A Community Based Referral And Counter Referral System As A Strategy For Improving Antenatal, Postnatal Care Visits And Health Facility Delivery, Case Of Kisii County, Rural Kenya

Abstract

Referral is a practice of sending patients from a less to a more specialized level of care. Studies have shown that effective referral systems have a beneficial effect and could make significant contribution to a reduction in home deliveries. The study was meant to develop and test a Community Based Referral Model (CBRM) that could be used to improve antenatal, postnatal care visits and hospital deliveries. The study had three outcome measures e.g. Proportion of pregnant women who attended antenatal care clinics, home deliveries who visited postnatal clinic and proportion of women who delivered in the link hospitals. A quasi -experimental study was carried out in two sub- locations linked to a hospital within the sub location. Secondary data on antenatal, postnatal visits and hospital deliveries on the past 10 months was extracted from registers in the two link hospitals before implementation of the referral system. One hundred community health workers were trained on community based referral and counter referral system and issued with referral tools. They were instructed to regularly visit the households in order to identify pregnant women, counsel and refer them to antenatal clinics in the linked hospitals. Each community health worker was assigned 25 households; hence 2209 households were regularly visited for a period of 10 months. Endline information was extracted from the same hospital registers. The two sets of data were compared. Statistically significant differences were observed in postnatal care visits (p-value 0.007) depicting an increase of 166% and hospital deliveries (p-value 0.001) depicting an increase of 135%. No differences were observed in all the 4 antenatal care visits. In order to get a better understanding of the effectiveness of a community based referral and counter referral system in a rural setting, a larger study with randomly selected intervention and control sites is recommended.