Stroke is known as one of the leading causes of disability worldwide (1–7). Moreover, it poses a great socioeconomic burden to individuals and society (6, 8). Rehabilitation after acute stroke treatment has been demonstrated to reduce the impairment of body function and limitation of activity, thereby facilitating an early return to home (7, 9, 10).

The guidelines for healthcare professionals from the American Heart Association/American Stroke Association emphasize the need for sustained and coordinated efforts for stroke rehabilitation (11). In many countries, specialized rehabilitation units or hospital systems have been developed to provide stroke survivors with rehabilitation treatment in the post-acute period (12, 13). These rehabilitation hospitals dedicated to the post-acute period are often subject to reimbursement tiers, which are based on a case-mix classification determined by disease group, severity, age, comorbidities, and other variables. For example, the prospective payment system in the United States utilizes prepayment tiers mainly determined by functional independence measures and age (14, 15). Japan's convalescent rehabilitation ward system sets an upper limit on the length of hospitalization that is dependent on disease groups. Additionally, post-acute rehabilitation facilities are evaluated by government or private agencies for indicators such as functional gains and returning home rate (16).

In South Korea, a post-acute rehabilitation hospital system was not introduced until 2019 (17), and reimbursement by the National Health Insurance Service (NHIS) was made in the form of fee-for-service for individual physiotherapy and occupational therapy (18). The payment level made to long-term care hospitals (LTCHs) without deduction did not differ from that of acute care hospitals. However, no criteria were established for the target population for medical rehabilitation or the duration of provision. When patients were hospitalized for rehabilitation treatment, there was also no evaluation system for providers, and payment by the NHIS for rehabilitation treatment was guaranteed for up to 2 years after the stroke (19).

South Korea's post-acute rehabilitation system was launched in 2020 (18). Medical institutions adopting the system allow patients to be admitted within 3 months after stroke and provide inpatient rehabilitation programs for up to 6 months. During this period, providing rehabilitation treatment for up to 4 h daily is possible. The rehabilitation physicians decide the types of treatment to be applied (e.g., gait training and swallowing training). Changes in body function, activity level during hospitalization, and return home rate after discharge are monitored (19).

To evaluate the effects of this system in the future, it is necessary to analyze the use of medical resources related to rehabilitation services to set a baseline before its introduction. No previous studies have described the use of rehabilitation medical resources before the introduction of the post-acute rehabilitation system. Hence, the methodology of this study can serve as a useful reference for countries planning to implement a post-acute rehabilitation system in the future to evaluate the effects after the introduction of this system.

This study aimed to investigate the length of hospital stay (LOS) and related factors for rehabilitation treatment after an ischemic stroke in 2017 before the introduction of the post-acute rehabilitation system.

Among the 5,188 patients linked to the NHIS-NHID, 868 died during hospitalization or within 2 years after discharge from the RCCVCs. Overall, 800 patients were readmitted to a tertiary hospital for purposes other than inpatient rehabilitation, 315 of whom experienced medical complications (Supplementary Table 1). A total of 240 patients had stroke recurrence, and 245 patients had planned readmissions (Supplementary Table 2). Hence, 1,668 patients were excluded from the analysis.

Analyses were performed for the remaining 3,520 patients (Table 1). Among them, 1,335 patients returned home after discharge from the RCCVCs without inpatient rehabilitation, and 2,185 patients were transferred to another hospital for inpatient rehabilitation. Among the 2,581 patients with minor strokes, 1,455 (56.4%) were readmitted to another hospital for inpatient rehabilitation. Meanwhile, of the 939 patients with moderate severity or above, 209 patients (22.3%) returned home after discharge from the RCCVCs without inpatient rehabilitation.

The median LOS in the RCCVCs was 7 (IQR: 6, 12) days for all 3,520 patients, 6 (IQR: 5, 9) days for the “not receiving inpatient rehabilitation” group, and 8 (IQR: 6, 14) days for the “receiving inpatient rehabilitation” group.

The patients in the inpatient rehabilitation group were older and had a higher severity when measured by the NIHSS than those who did not receive inpatient rehabilitation (P < 0.001 and P < 0.001, respectively). The proportion of women was higher in the inpatient rehabilitation group (P < 0.001).

The median LOS from ER admission to home discharge in the inpatient rehabilitation group was 47 (IQR, 17–196) days. During this period, these patients received rehabilitation treatment while being hospitalized at three hospitals sequentially at a median value (IQR: 2, 4), rather than being admitted to a single hospital (Table 2). Patients with severe stroke had a longer LOS and a higher number of hospital admissions (P < 0.001). Female (P < 0.001 and P = 0.006, respectively) and elderly (P < 0.001 and P = 0.001, respectively) patients also had a longer LOS and a higher number of hospital admissions. The lowest income level group (Q1) had a longer LOS than that of the other income level groups (P = 0.050).

Table 3 describes the multivariate regression model between LOS and the explanatory variables. In this model, LOS was longer with older age and increasing severity. In addition, the LOS was longer for women and Q1, the lowest income group (P = 0.002 and P = 0.020, respectively).

Although there are no standard guidelines on the strategies for rehabilitation treatment according to the severity of the stroke, patients with minor stroke (NIHSS 0–4) may generally undergo outpatient-based rather than intensive inpatient rehabilitation. By contrast, patients with moderate or higher stroke severity often require inpatient rehabilitation during the post-acute period (25). This study investigated the rehabilitation healthcare delivery system before the introduction of the post-acute rehabilitation hospital system and identified that rehabilitation after stroke was not properly performed based on severity (9). Before 2000, Japan had no dedicated system in inpatient facilities to provide early and intensive rehabilitation (26). Prior to 2000, Japan faced the same situation as South Korea.

In the US, inpatient rehabilitation facilities are institutions that provide intensive inpatient rehabilitation treatment after an acute period of 13 diseases, including stroke. Patients can undergo intensive multidisciplinary rehabilitation programs. They receive 3 h of treatment for 5 of 7 consecutive days (18, 27). Patients who have completed functional improvement are subject to an_active return-to-society program, and those whose condition does not improve discontinue rehabilitation treatment and are transferred to an LTCH (18). The Japanese national insurance system introduced kaifukuki (convalescent) rehabilitation wards (KRW) in 2000 (26). Rehabilitation treatment was limited to 3 h per day, and the maximum length of stay for patients with stroke in the KRW was limited to 150 days. The timing of discharge is set when patients reach a plateau in activities of daily living according to an interactive evaluation, which has facilitated the home discharge of patients with stroke with severe disability (16).

In our study, among 939 patients with stroke of moderate severity or above, 209 (22.3%) returned home after discharge from the RCCVCs without inpatient rehabilitation (Table 1). In South Korea, rehabilitation services were reimbursed based on the fee-for-service system even before the post-acute rehabilitation hospital system was introduced. Therefore, rehabilitation treatment may not be implemented in the post-acute period in these groups, not for economic reasons, but because the referral system between medical institutions did not work. Since the median LOS in RCCVCs was 6 days or less and outpatient-based rehabilitation treatment is rarely performed in South Korea, rehabilitation services could be undersupplied for these patients (26).

Meanwhile, among 2,581 patients with minor stroke with an NIHSS score of ≤4, 1,455 (56.4%) were readmitted to another hospital for inpatient rehabilitation after discharge from RCCVCs (Table 1). The average (median) LOS was 118.0 (29) days (Table 2). In this group, rehabilitation treatment could be oversupplied. Therefore, this study shows that rehabilitation services could be over- or under-supplied to patients with stroke without systematic provision of medical rehabilitation in the post-acute period (9).

Another characteristic phenomenon demonstrated in this study was that patients with stroke in the inpatient rehabilitation group visited several hospitals after being discharged from the RCCVCs, resulting in prolonged LOS (9, 28). For example, patients with moderate severity (NIHSS 5–15) who were discharged from RCCVCs and transferred to another hospital to receive rehabilitation services were admitted to two additional hospitals before returning home. The median net LOS was 128 days. This phenomenon may be related to the following issues. First, in South Korea, rehabilitation treatment is mainly provided on an inpatient basis, and outpatient-based rehabilitation treatment is rarely performed after discharge to the home. When receiving rehabilitation treatment on an outpatient basis, the out-of-pocket ratio is higher than that of inpatient rehabilitation. In addition, because of hospitalization fees that can be billed to the NHIS, providers prefer inpatient to outpatient rehabilitation. Regardless of severity, patients with stroke who want rehabilitation treatment after discharge from RCCVCs might have no option other than hospitalization. Second, there were no criteria for determining whether inpatient rehabilitation was required in South Korean NHIS in 2017. Instead, if the hospitalization period exceeds 2 weeks, hospitalization reimbursement is deducted, and after 1 month, it is further deducted (26). However, when patients are admitted to another hospital post-discharge, the deduction rate is not applied cumulatively but rather is reset. Therefore, providers might be concerned with the discharge of patients with stroke but are not interested in whether they are discharged home or admitted to another hospital. Third, among the rehabilitation treatment items based on the fee-for-service reimbursement system, patient/family education and counseling for return to home performed by the rehabilitation team were not included. The absence of these services may have delayed the return to home after the stroke.

Several factors influencing LOS may be related to the specific situation in South Korea rather than being universal. First, the long LOS for women may be because women in South Korea have a higher share of household work and fewer people at home to take care of them (29, 30). Considering these environmental factors, women may be more likely to have difficulty returning home than men after a stroke. Second, the longer LOS in the lowest income group (Q1) can be at least partially explained by the fact that 10–30% of co-payments for medical expenses are not applied to the group (31).

This study has some limitations. First, the NHIS-NHID does not contain information on family members or support. Therefore, information on family support could not be obtained, although this is a known factor that affects LOS (32, 33). Second, the RCCVC system in South Korea performs cardio-cerebrovascular thrombolysis, an emergency procedure. Therefore, patients with other types of stroke, such as cerebral hemorrhage, were excluded from this study. Third, only hospitalization for rehabilitation treatment was investigated and presented, and the rehabilitation treatment itself was not quantitatively analyzed. Because rehabilitation treatment was based on the fee-for-service system as of 2017, it was too complicated to analyze the actual rehabilitation contents and amount of treatment. Fourth, the income level was classified according to the level of NHIS payment. In South Korea, when calculating the NHIS co-payment, the patient's income is determined by salary or tax return, and tax exemption is not included in the calculation. Therefore, patients' incomes in this study may have been underestimated.

In future studies, the use of rehabilitation medical resources and return to home in patients with stroke that occurred after the introduction of the post-acute rehabilitation hospital system in 2020 should be investigated, and the results should be compared with those of this study.

In conclusion, before the introduction of the post-acute rehabilitation system in South Korea, rehabilitation treatment after stroke was both over- and under-supplied, and return to home was delayed after discharge from acute care hospitals. Returning home after a stroke was the most delayed in the lowest-income group, high-severity group, and female patients. These results support the need for a post-acute rehabilitation system that specifies the subject, duration, and intensity of rehabilitation treatment. The data reported in this study may serve as basic evidence for planning new systems in countries where a post-acute rehabilitation system has not yet been established.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by the Institutional Review Board of Seoul National University Hospital. Written informed consent from the participants' legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.

SB conceived and planned the study, performed the analysis, wrote the original draft with input from all authors, visualized the results, validated the study, and contributed to the interpretation of the results. JK conceived and planned the study, performed the analysis, wrote the original draft with input from all authors, and contributed to the interpretation of the results. H-IS conceived, planned, validated the study, contributed to the interpretation of the results, and supervised the entire process. All the authors have read and approved the final manuscript.