Costs and quality of services in public hospitals in Zimbabwe : implications for hospital reform
Hospitals come under the focus of health planners and policy makers because they invariably consume large and increasing amounts of health care resources and performance is commonly believed to fall short of that possible. The common response by governments to this situation has been to implement hospital reforms. However, emerging evidence from impact evaluations of such reforms shows little clear evidence of performance enhancement. It is argued in this study that hospital reforms in most countries are implemented without enough understanding of current performance, or knowledge of hospital behaviour. Such information is necessary for effective design, implementation and evaluation of reforms. The aim of the study was to measure hospital performance and contribute to the understanding of its determinants. The role of internal organisation and management to hospital performance has been underplayed in most studies such that the workings of the hospital remain a "black box". The study sought to demonstrate that understanding hospital performance entails understanding not only the technical relationships of dimensions of hospital performance but also the institutional context, and behaviour of individuals or groups within it who ultimately shape hospital behaviour. A multiple case study approach was used to study six tertiary hospitals in Zimbabwe. Hospital performance was first assessed through analysis of utilisation statistics. This was followed by an assessment of two dimensions of hospital performance: costs and quality of inpatient services. Costs were measured using standard cost accounting methods at hospital, ward and patient level. At patient level, a combination of. prospective micro-costing and top-down costing methods was applied to cohorts of patients suffering from selected tracer diseases: 207 malaria and 158 pulmonary tuberculosis cases. The quality of hospital inpatient services was also measured at hospital and patient level using structural and process approaches. The relationship between cost and quality of services was then explored at patient level using tracer conditions. A triangulation of methods was then used to explore internal organisation and management: staff interviews, observations, attendance at hospital meetings and review of administrative records. Analysis of activity statistics showed that the six hospitals had different levels of activity although they had similar roles in the referral hierarchy. Distinctive unit cost patterns were observed across the hospitals. Unit cost variation across hospitals was generally similar at hospital, ward and patient level. The results from the analysis of activity statistics were predictive of hospital cost classifications. The quality of hospital services varied across hospitals from both structural and process perspectives. There was little convergence in results from hospital level structural quality assessment, and process quality assessment. Cost-quality relationships in inpatient care showed a distinct pattern across tracer diseases, which permitted classification of the six hospitals into three performance categories. These classifications were used to relate quantitative and qualitative results of the study. The institutional contexts within which public hospitals in Zimbabwe operate is explored and described. There are fundamental policy design weaknesses related to the way hospitals are financed, governed and managed, which affect hospital performance. Hospital staff appears apathetic about hospital performance because of lack of appropriate incentives. Several hospital internal factors were reported as impinging on hospital performance. These factors can broadly be summarised as lack of management capacity and skills, inappropriate internal organisational and management structures, and staff reward systems. The current incentive structure at individual and institution level does not engender performance improvement. Relative hospital performance did not vary systematically with different institutional characteristics. For instance, compliance or non-compliance with mandated organisation and management structures did not account for performance differences whilst weak associations were found between relative performance, and differences in management capacity and skills. The absence of direct relationships between institutional characteristics and relative performance was not unexpected given the exploratory nature of the study and the possible multiple interrelationships between these factors Nonetheless, the study systematically describes and exposes current weaknesses in the internal structure of public hospitals in Zimbabwe, and identifies those internal organisational and management features considered important to performance. The study concludes that there is considerable scope for improving hospital efficiency and quality of services (with available resources) by changing internal organisation and management of hospitals. Of particular importance is the need to change and align incentives (monetary and nonmonetary) at both individual and institution level in ways that promote performance improvement.