Forecasting Demand in Health Services: the Case of University of Nairobi Clinics

Abstract

The provision of health services is an important aspect of socio-economic development and improving them plays an important role in raising people's standard of living. Taking a quick glance at any hospital or clinic in Kenya, ones attention is attracted to whether the management of these health programs has taken due consideration to the challenges the program poses. On several occasions, Kenyans face a number of problems in

trying to use the few facilities available in these hospitals and clinics. This study,set out to model demand patterns for the University of Nairobi Health Cl inics by developing time series forecasting models for·each of the clinics and for all the clinics tqgether. These models, it is hoped, can assist in the prediction of demand for the health services in the University Clinics.

The result of the study indicates that the attendance process can ". be modelled using time series analysis. This is supported by the F- test carried out. This test showed that all the clinics except Upper Kabete (staff) had significant F- ratios (at 95% significance

level) and thus the demand is time dependent. The models were validated and predictive powers of between 95.47% and 71.78% were achieved for the specific models developed. The overall model has a predictive power of 91.7%.