An input-process-output operational performance model for a long-term care day care center using routinely collected administrative data

Day care centers for long-term care manage staffing, attendance, service use, and costs under reimbursement constraints. Routine administrative records can support monitoring but do not fully capture Balanced Scorecard (BSC) constructs such as staff development, service quality, user experience, or organizational learning. This study examined how these records could be organized and interpreted within a BSC-informed input-process-output operational performance model. This retrospective secondary analysis used 226 operating-day observations from one day care center in Seoul, South Korea, nested within 11 months. Indicators were organized as input indicators, process indicators, utilization output, and financial/accounting outputs. Total pay and Total operating expenses were log-transformed, and model variables were standardized. Direct paths used finite-sample-corrected month-clustered inference and 5,000 month-block bootstrap draws; indirect and total associations used the same bootstrap procedure. The retained model yielded the following conventional in-sample fit indices: chi-square = 7.370, df = 6, p =.288, GFI = 0.998, TLI = 0.988, CFI = 0.997, and RMSEA = 0.032. It accounted for 39.3% of the variance in beneficiary status, 39.3% of the variance in Total pay, and 49.1% of the variance in the allocated daily operating expense proxy. Non-use count was negatively associated with beneficiary status. Service count was positively associated with Total pay, an expected linkage because both measures originated from the claim-record process. Cost-related inference was method-sensitive: the month-block 95% confidence interval for the path from Service count to Total operating expenses included zero despite a statistically significant CR1 cluster-robust test. Routinely collected administrative records can be organized into an input-process-output operational performance framework using the BSC as a broad organizing lens. Because the indicators were administrative proxies rather than full measures of BSC constructs and the data were obtained from a single facility, the findings should be interpreted as exploratory associations rather than evidence of causal effects or cost-efficiency gains.