Hospital length of stay (LOS) is an essential indicator of the efficiency of health service delivery. Optimal discharge timing can enhance patient flow, freeing up beds and health care staff time [1]. Shortening LOS is often associated with improved patient outcomes. A previous study in the United States found that reduced hospital LOS did not increase readmissions or mortality, demonstrating that efficiency improvements can maintain or enhance patient care quality [2].

Delayed discharge, sometimes called bed blocking, has been a serious problem in health service delivery for the past half-century [3]. Older patients are particularly at risk of delayed hospital discharge because of their complex health and social needs [4]. When a patient stays in a hospital bed longer than necessary, limited health care resources are consumed, leading to the cancellation of elective surgeries, extending waiting lists, and hindering the admission of emergency cases [5]. Delayed discharge incurs additional costs. A previous systematic review estimated the mean cost per delayed discharge at US $7020, including only hospitalization costs, after adjusting for purchasing power parity [6].

Developing a discharge plan and coordinating the care services that patients will use after discharge from the early phase of hospitalization is essential for facilitating timely discharge. Previous studies have identified causes of delayed discharge such as complexity in arranging care services after discharge, waiting for admission to institutions such as community hospitals, and challenges in nursing home care and long-term care (LTC) facilities [4,7-9]. Adequate discharge planning, including organizing discharge and postdischarge services, reduces LOS [10].

The Japanese government launched the long-term care insurance (LTCI) system in 2000 [11]. As part of Japan’s universal health coverage framework, the LTCI system allows all individuals aged ≥65 years to access publicly funded LTC services, including home-based care (eg, home visit nursing and care), community-based services (eg, adult day services), and institutional care (eg, nursing homes and LTC facilities). To use these services, older adults must undergo a formal care needs certification process administered by municipal governments. This assessment evaluates physical and cognitive function to determine eligibility and assign an appropriate level of care [11,12].

As noted previously, it is important for effective transitional care that discharge planning and coordination of postdischarge services begin during hospitalization. In this context, initiating the care needs certification process while the patient is still in the hospital is considered ideal as it may facilitate a smoother transition to LTC services. However, it remains unclear whether the timing of LTC certification applications—particularly those submitted during hospitalization—is associated with LOS. To generate hypotheses about this potential relationship, this study conducted an ecological analysis using prefectural-level data from Japan. These results are expected to provide a foundation for future research using national databases that include individual-level data.

This ecological cross-sectional study was conducted at the prefectural level to explore the association between the timing of LTC certification applications—specifically those submitted during hospitalization—and LOS. All data used in this study were aggregated at the level of Japan’s 47 prefectures and obtained from publicly available government databases. These datasets are accessible online, and detailed information, including URLs, is provided in the Data Availability section of this paper. The analysis was conducted in June 2024 using the most recent datasets available at that time in which the exposure and outcome data were from the same reference year.

This manuscript adheres to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for reporting cross-sectional studies.

This study was determined to be nonapplicable to ethics review by the ethics committee of the National Center for Geriatrics and Gerontology (approval 1772) in accordance with Article 6, Section 1, of the Ethical Guidelines for Medical and Biological Research Involving Human Subjects in Japan. This study used only aggregated, publicly available secondary data at the prefectural level; therefore, informed consent was not required. No direct human participation took place, and no individual-level or identifiable data were used.

All variables were summarized using descriptive statistics, including the mean and median, SD, IQR, and ratio of minimum to maximum. A linear regression model was used to analyze the data. The dependent variable in the model was the average LOS among the population aged ≥65 years in each prefecture during FY 2020. A logarithmic transformation was applied because of the right-skewed distribution of the average LOS. The independent variables included the proportion of LTC certification applications submitted during hospitalization and other covariates. The Akaike information criterion and Bayesian information criterion were used to assess the goodness of fit for each model.

The outcome variable from the 2020 Patient Survey provides the average LOS among the population aged ≥65 years. However, there are several limitations. First, the data reflect the average LOS of patients discharged from hospitals during a specific period (September 1-30, 2020). Second, hospitalization patterns and LOS in 2020 were likely influenced by the COVID-19 pandemic. To address these limitations, we conducted sensitivity analyses using data from the 2018 and 2020 Hospital Report and the 2017 Patient Survey. Although the Hospital Report does not include the average LOS among specific age groups, it covers every hospital nationwide, with data collected monthly throughout the year. The average LOS of curative care beds is available, excluding psychiatric, infectious disease, tuberculosis, and LTC beds. This study used the average LOS of curative care beds to reduce bias derived from the differences in bed functions and focus on the most general types of beds.

A significance level of 5% was used for all tests (2-tailed). All statistical analyses were conducted using R (version 4.3.1; R Foundation for Statistical Computing).

The number of LTC certification applications submitted during hospitalization, their proportion among all new LTC certification applications, and the average LOS among individuals aged ≥65 years are summarized in Table 1.

A higher proportion of LTC certification applications submitted during hospitalization was significantly associated with longer hospital stay (β=0.04, 95% CI 0.01–0.06; P=.003).

The multivariate association between the exposure and outcome variables is shown in Table 2. A higher proportion of LTC certification applications submitted during hospitalization was significantly associated with a longer average LOS among individuals aged ≥65 years after adjusting for covariates.

The results of the sensitivity analyses using data from the Hospital Report (2018 and 2020) and the Patient Survey (2017) are provided in Tables S1-S3 in Multimedia Appendix 1, supporting the findings of the primary analysis.

Our results indicate that a higher proportion of LTC certification applications submitted during hospitalization was associated with a longer average LOS at the prefecture level. The primary analysis used average LOS data from the 2020 Patient Survey as the outcome variable. However, relying solely on this data source may introduce bias due to its data collection methods and period.

To address this potential bias, we conducted sensitivity analyses using average LOS data from multiple sources, including the Hospital Report (2018 and 2020) and the 2017 Patient Survey. These sensitivity analyses supported the primary results and enhanced the robustness of our findings, reinforcing the observed association between the timing of certification applications and average LOS.

Discharge planning and coordination of postdischarge services starting during hospitalization are essential for effective transitional care. A meta-analysis has shown that such planning, including the arrangement of postdischarge services, can reduce LOS [10]. However, contrary to this expectation, our findings showed that prefectures with a higher proportion of in-hospital LTC certification applications had longer average LOS.

One possible explanation for our findings is that this reflects delays in the LTC certification process itself. In recent years, significant delays have been reported in many municipalities [27]. Several news sources have highlighted that these delays are driven by a rapid increase in the number of applications, which has strained administrative capacity. Indeed, Japan’s increasing aging population continues to drive up the demand for LTC certification [28]. The care needs level for each applicant is determined by a municipal committee based on multiple sources, including face-to-face assessments of physical and cognitive function [11,12,29]. These assessments are time and labor intensive, and as the volume of applications increases, so does the workload for investigators, which may further prolong the certification process. Therefore, delays in certification may hinder the timely arrangement of postdischarge LTC services, potentially contributing to extended hospital stays. Previous studies have noted that discharge delays are often related to challenges in arranging postacute care services [9]. Therefore, although initiating the LTC certification process during hospitalization is generally considered a good practice to support a smooth transition to community-based care, this intended benefit may not be realized when systemic delays are present.

Although our model accounted for several relevant covariates, some unmeasured or residual confounding likely remains, and causal inference is limited by the study’s ecological design. There are 3 alternative explanations and considerations.

First, the number of in-hospital certification applications may simply reflect the presence of patients who are expected to have longer hospital stays. For example, those with multiple chronic conditions, functional dependence, or cognitive impairment or those living alone are more likely to need LTC services after discharge [7-9,19-21]. Because of their longer stays, these patients may have more opportunities to initiate the certification process while still hospitalized. Given the ecological nature of our study, reverse causality cannot be ruled out. Therefore, further investigation using individual-level data is essential to clarify the relationship between the timing of LTC certification application and hospital LOS.

Second, the inclusion of rehabilitation wards in the hospital dataset used to calculate average LOS and in-hospital LTC certification application rates may help explain the observed association. In Japan, these facilities typically provide longer inpatient care than acute care hospitals. For example, patients with stroke may stay for up to 150 days, with an average LOS of 67.5 days compared to 29.5 days in acute care hospitals [30]. In such settings, patients are generally in stable medical condition, and the extended LOS may increase opportunities to initiate LTC certification during hospitalization. Thus, prefectures with a higher concentration of rehabilitation wards may tend to have both longer LOSs and a higher proportion of in-hospital LTC certification applications.

Third, prefectures with longer LOSs may also have fewer available institutional care beds. In such areas, LTC certification applications may be submitted during hospitalization out of necessity, but discharge may be delayed due to a lack of available placements. Waiting for admission to a nursing home has been identified as a key factor contributing to prolonged hospital stays in previous studies [7-9,22]. Japan faces serious shortages in institutional LTC capacity. A document in 2022 reported that 275,000 people were on waiting lists for public nursing homes [31]. Therefore, variation in LTC resource availability across prefectures may help explain the association observed in this study.

This study has some limitations. As an ecological study, confounder adjustment may not have been sufficient because of unmeasured confounders, particularly at the individual level. Consequently, the findings cannot be generalized to individual-level associations because of the potential for ecological fallacy. For example, patients with complex care needs may have longer hospital stays and may also be more likely to submit LTC certification applications during hospitalization. Therefore, these results should be interpreted with caution, and further analyses using individual-level data are strongly warranted.

To our knowledge, this is the first study to examine the association between LTC certification application during hospitalization and the average length of hospital stay, offering important insights for policymakers aiming to strengthen the continuity of transitional care in aging societies such as Japan. Our findings indicate that initiating LTC certification during hospitalization is associated with longer hospital stays, which may reflect systemic delays that hinder discharge coordination and lead to inefficient use of hospital resources. Therefore, prefectures with particularly high proportions of in-hospital applications may benefit from targeted interventions to accelerate the certification process.

Although we used the most recently available data at the time of analysis (June 2024), there was an inevitable time lag between data collection and public release—for example, the 2020 open LTCI claims data were published in November 2023, and the 2020 Patient Survey was published in June 2022. Importantly, at a June 2025 meeting of the Ministry of Health, Labour, and Welfare’s LTCI subcommittee, hospital representatives reported that delays in the LTC certification process may be contributing to unnecessary extensions of hospital stays [32]. This contemporary policy discussion underscores the continued relevance of our findings despite the temporal gap.

Several digital transformation initiatives are already underway in Japan to address these challenges [33,34]. These include artificial intelligence–assisted assessment tools; digitized application workflows; remote operation of certification review meetings; and the development of shared, interoperable information platforms among hospitals, insurers (municipalities), LTC service providers, and care recipients. Such innovations have the potential not only to improve the efficiency and responsiveness of the LTC system but also to enhance health care delivery by facilitating timelier discharge planning and reducing unnecessary hospital stays.

Our study suggests a possible association between LTC certification application during hospitalization and hospital LOS at the prefectural level in Japan. These findings may reflect underlying system-level factors related to discharge coordination and administrative processes. Further research using individual-level data is needed to clarify the mechanisms and examine how LTC certification application timing may influence discharge outcomes. Policymakers in aging societies may benefit from considering these findings as they work toward improving the efficiency and continuity of transitional care.