Classification of long-term care wards and their functional characteristics: analysis of national hospital data in Japan

Abstract

Background

In a rapidly aging society that has promoted extensive reforms of the healthcare system, clarifying functional patterns in long-term care wards is important for developing regional healthcare policies. This study aimed to classify patterns of inpatient characteristics among Japanese long-term care wards and to examine hospital/ward characteristics.

Methods

We analyzed data from 1856 long-term care wards extracted from the 2014 Annual Report for Functions of Medical Institutions in Japan. We classified five clusters of long-term care wards based on inpatients’ medical acuity/activities of daily living using cluster analysis, and compared hospital/ward characteristics across the clusters with a chi-square test or analyses of variance.

Results

Cluster 1 was low medical acuity/high activities of daily living (n = 175); cluster 2, medium medical acuity/high activities of daily living (n = 340); cluster 3, medium medical acuity/low activities of daily living (n = 461); cluster 4, high medical acuity/low activities of daily living (n = 409); and cluster 5, mixed (n = 471). Although clusters 1 and 2 had similar higher proportions of home discharge (48.1% and 34.6%, respectively), there was a difference in length of hospital stay between the clusters (154.6 and 216.6 days, respectively). On the other hand, clusters 3 and 4 experienced a longer length of hospital stay (295.7 and 239.8 days, respectively) and a higher proportion of in-hospital deaths (42.7% and 50.2%, respectively). Characteristics of cluster 5 were not significantly different from the average of overall wards.

Conclusions

There were distinctive differences across hospitals in their use of long-term care wards. Wards with different functions have different support needs; the clusters with high activities of daily living needed support in promoting home discharge, while those with low activities of daily living needed support in providing quality end-of-life care. Our results can be useful for constructing the future regional healthcare system. This study also suggests introducing a standardized patient classification system in long-term care settings.