The objectives of this study are to develop the methodology for classifying hospitals in the -perspective of hospital functions through the analysis of the variations of casemix among various hospitals in Korea; to classify the hospitals into meaningful groups of optimal size; and to identify the functional characteristics of each hospital group.
51,885 discharge cases in 156 hospitals were obtained from the Korean Medical Insurance Corporation in 1983 and the data were analyzed for these purposes. Smaller hospitals having less than or equal to 120 beds were grouped into 5 units according to the bed size. Finally 52 hospitals or, hospital groups were the study units. Every diagnosis code was transformed into 46 DRGs(Diagnosis Related Groups); DRG proportion in each unit hospital was produced; among 46 DRGs, 25 were selected as having significantly high coefficient of variation; factor analysis was applied to the relevant DRGs;and the variance reductions in medical service cost and length of stay among several combinations were compared.
The major findings of this study are as follows:
1) Based on factor analysis, DRG 4(malignant neoplasm), 10(eye diseases), 11(ear diseases), 12(valvular & congenital heart disease), 34(problems during childbirth) and 36(congenital anomalies) in one group. and DRG 2(pulmonary tuberculosis), 8(psychiatric disorders), 43(internal) and 45(burns), were significantly related with the same factor respectively regardless of the number of factor set.
2) As the number of factor set increased, the variance reduction in the cost increased; the variance reduction in length of stay was maximal where the hospitals were classified into 5 groups. Hence 5 hospital groups were selected as optimal.
3) 5 hospital groups were identified to have the functional characteristics as followings: Group I Hospitals having highly complicated casemix and functioning as central hospital
in the region with the largest service volume
Group 9 Hospitals having cases of longer length of stay with relatively large medical service volume
Group 1ff Hospitals having cases of short length of stay with minimal total´, service volume Group IV Hospitals having mainly mild cases with larger service volume
Group V Hospitals having no overt functional characteristics with minimal unit service volume _
This study, it seems, provides a basis for the functional classification, of hospitals and will
be able to contribute significantly to the reasonable establishment of the national health care system
|