PEDIATRIC UTILIZATION OF A TEACHING HOSPITAL AND A COMMUNITY-HEALTH CENTER - PREDICTORS OF LEVEL OF CARE BY CHILDREN FROM KHAYELITSHA, CAPE-TOWN

Rationale. Inappropriate utilisation of hospital services for primary curative care aggravates inefficiencies and inadequacies in health car e delivery. Identification of reasons for such malutilisation may assi st the development of appropriate strategies for development of ration ally organised primary and secondary care services that will provide i mproved quality of care. Subjects. Children under 6 years of age livin g in Khayelitsha. Objectives. To ascertain: (i) the proportion of visi ts made to Red Gross Children's Hospital (RXH) that could be more appr opriately handled at primary care level; (ii) reasons for attendance a t RXH, compared with attendance at a large community health centre (Si te B day hospital) in Khayelitsha; (iii) predictors of inappropriate a ttendance at the teaching hospital for primary care problems. Methods. A case-referent study design was used to compare children attending R XH with children attending Site B day hospital. All care-givers attend ing the respective outpatient departments on 2 randomly selected days were included in the study sample, Data were collected by semi-structu red interview and record review, on reasons for attendance, demographi c and social variables relating to the child and care-giver, as well a s clinical data on the final diagnosis contained in the patients' fold ers. Criteria for determining appropriateness of attendance by level o f care were developed a priori via a modification of published measure s. Main outcome measures. Reasons for attendance at the facility appro priateness of the visit by level cif care and predictors of inappropri ate attendance at the teaching hospital. Results. Sixty-nine per cent of RXH visits were identified as inappropriate for a tertiary institut ion, The main reasons given by care-givers for attending Site B were c onvenience and the prohibitive cost of travel to RXH. Mothers intervie wed at RXH reported problems with failure of treatment at primary care clinics, and being turned away at Site B because of overcrowding as t he main reasons for attending RXH. Attendance at RXH was predicted by: (i) children who were infants; (ii) no other domestic child care resp onsibilities for the care-giver; and (iii) no previous attendance docu mented in the hospital folder. Conclusion. Appropriate service utilisa tion by level of care needs to be improved. Users' choice of service a ppears to be a rational decision based on the accessibility of local p rimary care services and perceptions of the quality of these services, More appropriate use of primary care facilities therefore requires be tter access and perceived quality.