Abstract
This research report is part of the research program entitled “financing study to support the public health security in Thailand” with the primary focus on public hospitals under supervision of the Ministry of Public Health. The objectives of this research are: a) to deepen understanding about cost efficiency of public hospitals in Thailand and to search for those hospitals that are operating at cost frontier, b) to lend support to the database development that would adequately cover the financial aspect of hospital management; c) to apply econometric tools to measure cost and technical efficiency of public hospitals and to perform comparative analyses of efficiency across public hospitals.
This research compiled data and information from public hospitals that are broadly grouped into 3 types, namely, regional-, provincial- and community hospitals and adopts economic theories and models to estimate cost functions, unit cost, economies-of scale, and capacity utilization. With respect to relative efficiency of public hospitals, the present study employs Data Envelopment Analysis (DEA) as tool to measure technical efficiency scores of each hospitals; the model has some desirable features in that it is nonparametric approach and does not requires rather strict assumption regarding distributive functions of the parameter. Our DEA method follows the input-orientation (cost minimization) and two alternative assumptions are adopted for sake of comparison, VRS (Variable returns to scale and CRS (constant returns to scale)
The case studies under the coverage of this research include 710 hospital units under supervision of the Ministry of Public Health that are classified under 3 groups, namely, regional hospitals (23 units), provincial hospitals (58 units), and community hospitals (62 units). Three sources of information are pooled into our database, namely, a) the financial data based of the yearly accrual accounting report of individual hospitals with details of spendings by categories such as personnel expenses, drug and medication cost, utilities and others—all figures refer to fiscal year of 2006; b) the provision of health cares by individual hospitals that are broadly grouped into 2 categories, viz., in-patient day and out-patient cases and the relative weight of DRG (diseases related grouping to reflect degree of severity of illness; c)the basic information of public hospitals that are inclusive of the number of bed, numbers of medical staffs and support staffs, where the hospitals are located, and others.
Findings: The average efficiency scores in case of regional hospitals ranged from 94-97%, in comparison to the provincial hospitals whose averages ranged from 86-89% and the community hospitals whose average values ranged from 75-81%. These estimates imply that there are room for further improvement and , if so, -would reduce hospital costs by 3-7% in the case of regional hospitals, 11-14% in the provincial hospitals, and 19-25% for community hospitals.
Another topic of study is to investigate whether the public hospitals are operating under constant-cost, or decreasing-cost, or increasing-cost situation. Based on economic theory, the decreasing cost situation implies there is a possibility to reduce the unit cost by expanding the scale of operation, and this suggests that the particular hospitals are that those hospitals may have operated beyond the full capacity. The research team performed regression analysis of cost function by different types of public hospitals-the results indicate that: a) the regional hospitals tended to operate at constant-cost that implies a full capacity utilization; b) in the cases of provincial- and community-hospitals, our regression estimates indicate that they were operating at decreasing cost and that implies a less-than full capacity situation.
Discussion: The results from the DEA efficiency and the regression estimate of cost function reinforce each other and confirm to our prior expectation that the regional hospitals were almost always operated at full capacity and their efficiency score ranked highly. Remind that the efficiency score reflects the excessive input, for instance, the cost efficiency score of 90% implies there is a scope for cost reduction of approximately 10%. The total cost of regional hospitals were, on average, 1,330 million baht, in comparison to 499 million baht and 65 million baht for provincial- and community-hospitals respectively. Taken into account the magnitude of the cost of production, the scope for cost reduction in the case of regional hospitals is high and there might have motivated hospital managers to either reduce the excess inputs or to increase the scale of production whenever the utilization rate fell below the standard norms—in addition, there might be other qualitative factors such as location at the city-centered and the completeness of hospital care services provided by regional hospitals that raised confidence of patients, compared with the provincial-and community-hospitals. These do not mean to belittle the management of the provincial-and community hospitals, but they might have operated in different situations from the regional hospitals. To our understanding, some of community hospitals are located far from the town-centered, in island, in the rural and the remoted area with sparse density population-these situations made it difficult for those hospitals to operate at full capacity, And taken into consideration that demand for health cares are subjected to daily and seasonal fluctuations in illness—despite of these limitation and the nature of contingent demand, those community hospital are still necessary from the public management perspective, and accordingly, they tended to operate at below full capacity and, thus, higher unit cost and they fared lower in terms of efficiency scores.
Similarly to other researches, there are limitations to the present study which is considered to be at a “pilot” stage, we are of opinion that there are scope for much further improvement in our model and database. We would like to express our gratitude to the Research Steering Committee (chaired by Professor Ammar Siamwalla) and the financial support of the Health Insurance System Research Office for valuable comments and advices over the past year that have led to a marked improvement in our model. Following their advices the research team had conducted the hospital visit and held 3 focus group meetings in the north, central and southern provinces from which the medical doctors, the financial analysts and support staffs of hospitals were invited to comment and to suggest ideas for improvement in the next stages.
