Establishment and Quality Assessment of a Hospital-Associated Disability Database in Japan

BackgroundHospital-Associated Disability (HAD) refers to a condition in which Activities of Daily Living (ADL) decline during hospitalization and is associated with longer hospital stays as well as increased mortality and readmission prevalence among older adults. Many existing registry databases do not adequately capture preadmission ADL information, complicating the accurate identification of HAD. The purpose of this study was to develop and evaluate the quality of a multicenter registry database on HAD that includes detailed information on nutrition, medication use, swallowing, and walking ability. MethodThe research was carried out as a prospective multicenter study involving nine institutions across Japan, with Mie University Hospital designated as the leading institution. The participants were patients aged [&ge;]70 years old who were admitted to a general ward for rehabilitation and were independent in ADL (Katz Index: 6 points) before admission. Data were collected using the REDCap system on admission and discharge. Assessment items included age, gender, Katz Index, Functional Comorbidity Index, Clinical Frailty Scale, malnutrition based on the Global Leadership Initiative on Malnutrition (GLIM) criteria, Food Intake LEVEL Scale (FILS), Functional Ambulation Categories, etc. A decrease in the Katz Index during hospitalization was defined as HAD, and a decrease in the FILS was defined as hospital-associated dysphagia. ResultsA total of 209 patients with a mean age of 80.1 {+/-} 6.5 years were enrolled, and 60.8% were male. The leading causes of hospitalization included malignancies (39.7%), cardiovascular diseases (29.7%), and respiratory diseases (15.3%). HAD occurred in 29.1% of patients, hospital-associated dysphagia was observed in 16.7%, and weight loss during hospitalization occurred in 75.4%. Walking ability declined in 15.9% of patients during their hospital stay. The median length of stay was 19 days [IQR: 13-34]; 82.3% were discharged home, and 2.9% died during hospitalization. Moderate or severe malnutrition on admission according to the GLIM criteria was present in 40.4% of patients. Missing data were few, <5%, except for items such as cancer stage and preadmission weight, and the quality of the data was high. All facilities had a weekend and holiday rehabilitation system, with more than half of the facilities having a dedicated ward dietitian and dental specialist. DiscussionIn this study, we established a multicenter database on HAD in older hospitalized patients and were able to collect high-quality data, including nutritional status, swallowing ability, and rehabilitation information. The incidence of HAD was generally consistent with that reported in previous studies. In the future, this registry is expected to contribute to the identification of risk factors, the development of predictive models for HAD, and the design of early intervention programs for high-risk patients.