Effect of Ward-dedicated Physical Therapy Staffing on Outcomes among General Medical Patients in an Acute Hospital: A Difference-in-difference Analysis

Abstract

Objective: The study aimed to examine the effect of dedicated physical therapy (PT) staffing on the outcomes of patients admitted to a general medical ward with acute cholangitis. Methods: This retrospective observational study was conducted in an 865-bed tertiary-care hospital in Japan. Patients with acute cholangitis between September 2015 and August 2017 were enrolled. Patients admitted to a ward with dedicated PT staffing were included in the dedicated group, while those admitted to a ward without dedicated PT staffing were included in the non-dedicated group. Each group was further divided into pre-dedicated and post-dedicated period based on September 1, 2016, at which PT staffing was implemented. The primary outcome was absolute functional gain (AFG), which was defined as the difference between Barthel index at discharge and that at admission. A difference-in-difference analysis was conducted to examine the changes in AFG associated with ward-dedicated PT staffing. Results: We identified 456 patients with acute cholangitis. Complete case analysis was applied, resulting in 252 patients in the final analysis. Patients were assigned to the dedicated group in the pre-dedicated period (n = 66) and post-dedicated period (n = 52), and to the non-dedicated group in the pre-dedicated period (n = 60) and post-dedicated period (n = 74). The adjusted difference-in-difference estimator was 17.1 (95% confidence interval: 5.6 to 28.5, p = 0.003) for AFG. Conclusion: Ward-dedicated PT staffing may improve the AFG of general medical patients in an acute hospital. Ward-dedicated PT staffing should be among the strategies utilized in the acute care process.

Patients in an acute hospital are at risk of disability, and hospitalization-associated disability occurs in approximately one-third of older patients^1,2). Thus, health professionals should evaluate and identify the rehabilitation needs of patients at a high risk for hospitalization-associated disability.

Previous studies have demonstrated the effectiveness of early mobilization programs^3–6) and special care procedures (e.g., comprehensive geriatric assessment^7–10) and acute care for elders^11,12)) as ward-based interventions. In a retrospective pre-post study, Johnson et al.¹³⁾ demonstrated that additional cardiovascular intensive care unit (CVICU)-dedicated physical therapy (PT) staffing results in increased PT treatment and reduced CVICU and post-CVICU length of stay (LOS). In a retrospective observational study, Engel et al.¹⁴⁾ demonstrated that assigning physical therapists to the intensive care unit (ICU) increased the number of patients who received PT, shortened the time required prior to the initiation of PT, and decreased the ICU LOS. Livingstone et al.¹⁵⁾ also reported that PT/occupational therapy staffing is associated with improved outcomes and overall quality of life among long-stay residents. Moreover, in a randomized control trial, Landefeld et al.⁸⁾ reported that comprehensive geriatric assessment can improve basic functioning during activities of daily living (ADLs) and reduce the frequency of discharge to long-term care institutions.

Several studies have investigated the effectiveness and safety of early mobilization programs that involve special care procedures and dedicated PT staffing^{3,5,7,16–19)}. However, the effect of early mobilization programs and dedicated PT staffing on patient outcomes has been mainly investigated among ICU patients^13,14) and nursing home residents¹⁵⁾, while few studies have focused on patients in an acute general hospital. Therefore, the present study aimed to examine the effect of dedicated PT staffing among patients in a general medical ward at an acute hospital.

Methods

Ethical approval

The study protocol was reviewed and approved by Research Ethics Committee, Kameda Medical Center (No. 18-212). Informed consent was obtained in the form of opt-out on the web site because the data were obtained via a retrospective review.

Health insurance reimbursement program for maintaining or improving ADLs

In 2014, the Japanese Ministry of Health, Labor and Welfare established a new health insurance reimbursement program for maintaining or improving ADLs, which involves ward-dedicated rehabilitation therapy staffing. This program allows for preventive interventions for functional decline. The following conditions are required to receive reimbursement:

• The ward needs dedicated rehabilitation therapists.

• The dedicated therapists perform ADLs assessment, provide instructions to maintain or improve ADLs function to aid in the prevention of falls and pressure ulcers, provide information regarding patient status to family, and assist with early discharge planning.

• The dedicated therapists also share information related to the following with medical staff: current ADLs function, post-discharge risks, functional prognosis, activity level, and level of participation preferred by the patient.

In the hospital where the study was conducted, this insurance program was introduced in one general medical ward on September 1, 2016, but not in another general medical ward. Therefore, we had the opportunity to verify the effectiveness of the ward-dedicated PT staffing by using the difference-in-difference method with patients data from before and after the introduction of this insurance program in general medical wards with or without the ward-dedicated PT staffing.

Setting

This retrospective observational study was conducted in an 865-bed tertiary-care hospital in Japan. On September 1, 2016, two full-time, dedicated PT staff were assigned to one of the general medical wards where gastrointestinal and oncology patients were admitted (45-bed capacity) to work for 8 hrs/day and 7 days/week. Thus, one to two therapists were present in this ward each day of the week, including weekends.

We defined the patients in the general medical ward with dedicated PT staffing as the dedicated group and those in another general medical ward without dedicated PT staffing as the non-dedicated group. Patients in the non-dedicated group were gastrointestinal and oncology patients as in the dedicated group. Moreover, we defined the period before September 1, 2016 (the date when the dedicated PT staffing was implemented in our hospital) as the pre-dedicated period (year 1) and the period after September 1, 2016, as the post-dedicated period (year 2).

As a result, patients in the general medical wards where gastrointestinal and oncology patients were admitted were divided into 4 groups: those admitted to the general medical ward with dedicated PT staffing before the start of the staffing (the dedicated group in year 1), those admitted to the general medical ward with dedicated PT staffing after the start of the staffing(the dedicated group in year 2), those admitted to another general medical ward without dedicated PT staffing before the start of the staffing (the non-dedicated group in year 1), and those admitted to another general medical ward without dedicated PT staffing after the start of the staffing (the non-dedicated group in year 2).

The reason why we staffed ward-dedicated PT to the wards where gastrointestinal and oncology patients were admitted was because these patients were mainly older and sometimes orders of rehabilitation were missed despite the need for rehabilitation. Among them, we especially focused on patients with acute cholangitis, who are often older and have a period of rest due to initial treatment. Therefore, we included patients with acute cholangitis as participants.

Participants

We enrolled patients with acute cholangitis who were admitted to the general medical wards from September 2015 to August 2017. Patients with acute cholangitis were diagnosed and assessed for severity in accordance with the Tokyo guidelines²⁰⁾. Data were retrospectively obtained from April 2019 to July 2019 by two physical therapists. This study used the following criteria (Fig. 1):

Fig. 1.

Study inclusion and exclusion criteria

Inclusion criteria

• Patients in a general medical ward with acute cholangitis.

Exclusion criteria

• Transferred from another ward after 14 days of hospitalization.

• Patients admitted in both pre-dedicated period (year 1) and post-dedicated period (year 2).

• Patients who had missing Barthel index (BI) data.

We excluded the patients admitted in both pre-dedicated period (year 1) and post-dedicated period (year 2) because the effect of the periods could not be isolated.

Study groups

The patients were divided into 4 groups: dedicated group in year 1, dedicated group in year 2, non-dedicated group in year 1, and non-dedicated group in year 2. The patients were assigned to either the dedicated or non-dedicated group using the hospital’s bed management system. The decision was made independent of disease severity, etiology, and ADLs dependency.

Regular rehabilitation was conducted in four groups, and dedicated physical therapists were involved only in the dedicated group in year 2. A team of nine physical therapists were involved in the 4 groups, while two dedicated physical therapists were added and involved before regular rehabilitation in the dedicated group in year 2. The details on rehabilitation are described in the “Regular rehabilitation” and “Ward-dedicated PT staffing” sections.

Regular rehabilitation

Regular rehabilitation was delivered in all groups. Patients received regular rehabilitation if the gastroenterologist judged it necessary. The gastroenterologist’s decision was based on ability, including the decline in ADLs. Therapists compared the patient’s current status with the preadmission status and provided rehabilitation to the patients to bridge the gap. As there was no clear protocol, rehabilitation was planned for each individual patient. For example, if the patient was able to walk before admission, the goal of rehabilitation was to restore the ability to walk. If we determined that the patient lacked muscle strength to achieve the goal, we provided strength exercises. Regarding the cancellation criteria for rehabilitation, the recommendations provided by The Japanese Association of Rehabilitation Medicine were used as a guide²¹⁾.

Ward-dedicated PT staffing

The ward-dedicated PT was implemented for the dedicated group in year 2. The dedicated physical therapists evaluated the patients’ social information, grip strength, walking speed, Fried’s frailty phenotype²²⁾, and Short Physical Performance Battery scores²³⁾. The dedicated physical therapists assessed each patient’s abilities and provided instruction on preventive self-exercise to reduce bed rest, maintain and improve ADLs function, and maintain appropriate physical activity. The dedicated physical therapists also delivered the following interventions: walking exercises, stair climbing exercises, muscle strength exercises, and sit-to-stand training for the aim of aerobic and resistance training. In addition, we practiced the requirements of the facility standards as described in the “Health insurance reimbursement program for maintaining or improving ADLs” section.

When the dedicated physical therapists intervention was not enough, the ward-dedicated physical therapists requested additional order for regular rehabilitation to the gastroenterologist. The criteria for requesting a regular rehabilitation were as follows and the decision was made comprehensively: decreased physical function, BI of 85 points or less (criteria for calculation of disuse syndrome rehabilitation fee), history of immobility, possibility of improvement in physical function, and disease trajectory.

Study variables

Data on patient characteristics were obtained from the electronic medical records, which included the following information: principal diagnosis, age, sex, body mass index (BMI), Charlson comorbidity index (CCI), etiology and severity of cholangitis, endoscopic retrograde cholangiopancreatography (ERCP)-related complications, numeric pain rating scale (NPRS) scores at admission, BI values, and the time prior to the initiation of rehabilitation. A gastroenterologist evaluated the BI data at admission and discharge.

Outcome measures and study covariates

The primary outcome was absolute functional gain (AFG), which was defined as the difference between the total BI at discharge and that at admission. The secondary outcomes were LOS and time prior to the initiation of rehabilitation, which was defined as the period between patient admission and the first clinical assessment by a rehabilitation therapist. Factors that were considered potential confounders included age, sex, BMI, CCI, etiology and severity of cholangitis, ERCP-related complications, and NPRS at admission.

Statistical analysis

Data analysis

A complete case analysis was applied, which resulted in 252 patients in the final analysis. Student’s t-test (continuous variables) and Fisher’s exact test (categorical variables) were used to perform pre/post comparisons in each group.

Difference-in-difference approach

We used the difference-in-difference method, which is a quasi-experimental study design^24–27), to investigate the effect of dedicated PT staffing on patient outcomes. This method can distinguish background trends in patient outcomes and the initiation of dedicated PT staffing. The quasi-experimental study design prevents the effects of all confounders on the dedicated and non-dedicated groups without the need to adjust for each confounder. Thus, the difference-in-difference method can reduce potential biases from unmeasured variables.

No specific changes were noted in the dedicated or non-dedicated group within the pre- or post-dedicated period. Hence, we compared effects before and after the implementation of dedicated PT staffing, assuming parallel trends and common shocks. Moreover, we identified the effect of dedicated PT staffing on patient outcomes based on the difference between pre-post time differences in the dedicated group and pre-post time differences in the non-dedicated group. We developed the following model for the difference-in-difference method based on linear regression:

$$y\left(\text{AFG}\right)=\text{\hspace{0.17em}ε\hspace{0.17em}+\hspace{0.17em}β1\hspace{0.17em}dedicated\hspace{0.17em}group\hspace{0.17em}+\hspace{0.17em}β2\hspace{0.17em}post-dedicated\hspace{0.17em}period\hspace{0.17em}+\hspace{0.17em}β3\hspace{0.17em}dedicated\hspace{0.17em}group\hspace{0.17em}*\hspace{0.17em}post-dedicated\hspace{0.17em}period\hspace{0.17em}+\hspace{0.17em}βX}$$

where X refers to age, sex, BMI, CCI, etiology of cholangitis, severity of cholangitis, ERCP-related complications, or NPRS at admission. The β parameters are the regression coefficients associated with X, respectively, and ε is the random error component reflecting the difference between the observed and fitted linear relationship.

The effect of dedicated PT staffing on patient outcomes was denoted as the coefficient of interaction term during the pre- and post-dedicated periods (β3 in the model). The level of statistical significance was set at p <0.05. Anonymized data were analyzed using R version 3.6.0²⁸⁾.

Results

Our analysis of electronic medical records revealed that 456 patients with acute cholangitis were admitted to the general medical wards. Those who were transferred from another ward after 14 days of hospitalization (n = 18), those who were admitted in both pre-dedicated period (year 1) and post-dedicated period (year 2) (n = 5), and those with missing BI data (n = 181) were excluded. Thus, a total of 252 patients were included in the final analysis (Fig. 1). Figure 2 shows the scheme of the study design. The dedicated and non-dedicated groups were divided into pre- and post-dedicated periods (n = 118 [pre: 66, post: 52] and n = 134 [pre: 60, post: 74], respectively). Table 1 shows a comparison of year 1 and year 2 within the dedicated and non-dedicated groups regarding patient demographic and clinical characteristics. There were no significant differences in age, sex, BMI, CCI, severity of cholangitis, etiology of cholangitis, ERCP-related complications, and NPRS between year 1 and year 2 within the dedicated and non-dedicated groups.

Fig. 2.

Scheme of a study design to evaluate the effect of ward-dedicated PT staffing implementation

PT, physical therapy

Table 1.

Patient demographic and clinical characteristics

Variable	Dedicated group (n = 118)			Non-dedicated group (n = 134)
	Year 1 Pre-dedicated period (n = 66)	Year 2 Post-dedicated period (n = 52)	p value	Year 1 Pre-dedicated period (n = 60)	Year 2 Post-dedicated period (n = 74)	p value
ADLs, activities of daily living; BI, Barthel index; BMI, body mass index; CBD, common bile duct; CCI, Charlson comorbidity index; ERCP, endoscopic retrograde cholangiopancreatography; EST, endoscopic sphincterotomy; NPRS, numeric pain rating scale; SD, standard deviation
Age, mean (SD), yr	75.2 (11.8)	77.0 (12.7)	0.42	74.0 (12.1)	75.9 (11.5)	0.56
Sex, male (%)	36 (54.5)	29 (55.8)	1.00	36 (60.0)	45 (60.8)	1.00
BMI, mean (SD), kg/m2	22.2 (4.1)	22.4 (4.4)	0.80	23.2 (4.5)	23.2 (4.8)	0.90
CCI, n (%)			0.35			0.41
Score : 0–1	30 (45.5)	17 (32.7)		16 (26.7)	28 (37.8)
Score : 2–3	20 (30.3)	21 (40.4)		29 (48.3)	31 (41.9)
Score : ≥4	16 (24.2)	14 (26.9)		15 (25.0)	15 (20.3)
Severity of cholangitis, (%)			0.92			0.74
Mild	40 (60.6)	30 (57.7)		39 (65.0)	43 (58.1)
Moderate	22 (33.3)	18 (34.6)		18 (30.0)	26 (35.1)
Severe	4 (6.1)	4 (7.7)		3 (5.0)	5 (6.8)
Etiology of cholangitis, (%)			0.85			0.98
CBD stone	40 (60.6)	35 (67.3)		41 (68.3)	53 (71.6)
Malignant obstruction	13 (19.7)	9 (17.3)		10 (16.7)	11 (14.9)
Benign stricture	2 (3.0)	2 (3.8)		2 (3.3)	2 (2.7)
Other	2 (3.0)	2 (3.8)		6 (10.0)	6 (8.1)
Unknown	9 (13.6)	4 (7.7)		1 (1.7)	2 (2.7)
ERCP-related complications, (%)			0.67			0.9
Bleeding after EST	1 (1.5)	2 (3.8)		2 (3.3)	3 (4.1)
Perforation	0 (0.0)	1 (1.9)		0 (0.0)	1 (1.4)
Post-ERCP pancreatitis	2 (3.0)	1 (1.9)		1 (1.2)	0 (0.0)
Mortality	0 (0.0)	0 (0.0)		0 (0.0)	0 (0.0)
ADLs at admission, mean (SD), BI	80.4 (32.7)	59.9(40.1)	0.003	71.4 (41.8)	68.6 (41.2)	0.69
Pain at admission, mean (SD), NPRS	1.3 (2.2)	1.1 (1.9)	0.72	1.0 (1.8)	1.1 (2.0)	0.8

The adjusted outcome obtained using the difference- in-difference method is shown in Table 2 and Figure 3. Analysis of the interaction revealed that AFG had significantly improved in the Dedicated group (difference-in-differences estimates, 95% confidence interval [CI]: 17.1, 5.6 to 28.5) relative to the non-Dedicated group. However, we observed no significant difference in the time prior to the initiation of rehabilitation (difference-in- differences estimates, 95% CI: −0.4, −2.3 to 1.5) and LOS (difference-in-differences estimates, 95% CI: 1.7, −2.1 to 5.6) in the difference-in-difference analysis.

Table 2.

Unadjusted and adjusted outcomes obtained using the difference-in-difference approach

Variable	Dedicated group		Non-dedicated group		Difference-in- differences estimates (95% CI)*	p value
	Year 1 Pre-dedicated period (n = 66)	Year 2 Post-dedicated period (n = 52)	Year 1 Pre-dedicated period (n = 60)	Year 2 Post-dedicated period (n = 74)
*These adjusted coefficient estimates were estimated from linear regressions that controlled for age, sex, BMI, CCI, severity of cholangitis, etiology of cholangitis, ERCP-related complication, and NPRS
AFG, absolute functional gain; BI, Barthel index; CI, confidence interval; rehab, rehabilitation; SD, standard deviation; LOS, length of stay; BMI, body mass index; CCI, Charlson comorbidity index; ERCP, endoscopic retrograde cholangiopancreatography
AFG mean (SD), BI	0.15 (27.1)	17.1 (28.6)	1.9 (18.8)	3.5 (16.0)	17.1 (5.6 to 28.5)	0.003
Rehab initiation (SD), d	3.6 (3.0)	1.8 (1.2)	4.4 (3.4)	2.9 (2.2)	-0.4 (-2.3 to 1.5)	0.68
LOS mean (SD), d	14.1 (6.3)	16.4 (12.1)	13.3 (7.2)	13.6 (5.9)	1.7 (-2.1 to 5.6)	0.38

Fig. 3.

Changes in outcomes before and after the implementation of ward-dedicated PT staffing in the dedicated and non-dedicated groups

(a) AFG (BI). (b) Time prior to initiation of rehabilitation (days). (c) LOS (days)

PT, physical therapy; AFG, absolute functional gain; BI, Barthel index; LOS, length of stay

Discussion

The present study investigated the effects of ward-dedicated PT staffing on outcomes among general medical patients in an acute hospital, with adjustments for demographic and clinical characteristics. Our findings indicated that AFG increased in patients with acute cholangitis after the implementation of ward-dedicated PT. A subsequent difference-in-difference analysis to remove measured or unmeasured confounders (a limitation of previous studies) also indicated that ward-dedicated PT staffing was associated with an increase in AFG. Moreover, the increase in AFG occurred independent of the time at which rehabilitation was initiated. However, in the BI at the time of admission of the dedicated group, a decrease from 80.4 in the pre-dedicated period to 59.9 in the post-dedicated period was observed. This result indicates that the dedicated group may have been a patient population with more potential for improvement in AFG than the other groups in the second year.

Previous studies have reported that dedicated PT staffing reduces the length of time prior to the initiation of PT, increases PT treatment, and reduces ICU and post-ICU LOS^13,14). Moreover, specific care procedures in the ward may reduce short-term mortality, prevent falls, decrease delirium, and improve physical and cognitive functions^7–12). However, particularly for patients in an acute general hospital, data to clearly show such benefits are lacking. In a retrospective pre-post study, Johnson et al.¹³⁾ demonstrated the effect of increasing the number of physical therapists from two to four in a CVICU setting. Engel et al.¹⁴⁾ also reported a positive effect of a physical therapist-led early mobilization program for ICU patients. However, because of the retrospective nature of the study, effects specific to the implementation of dedicated PT staffing could not be distinguished from the natural temporal trends in patient outcomes. In our study, we were able to exclude the effects of natural temporal trends and those of confounding factors using a quasi-experimental design.

Acute cholangitis is common in older patients, occurring primarily due to stones in the bile ducts and gallbladder^29,30). Factors associated with the development of cholelithiasis include age, obesity, and physical inactivity^31,32). In addition, early decreases in function are likely to occur in patients hospitalized with acute cholangitis due to symptoms (intermittent fever accompanied by chills, right upper quadrant pain, and jaundice) and treatments (antibiotic and biliary drainage). Thus, patients with acute cholangitis may exhibit frailty due to physical inactivity and may be at high risk of hospitalization-associated disability.

We hypothesized that the implementation of ward-dedicated PT staffing would decrease the time required prior to the initiation of rehabilitation and improve AFG. Based on unadjusted results, early rehabilitation assessment and AFG significantly improved following the implementation of ward-dedicated PT staffing. However, when the results were adjusted via the difference-in-difference analysis, we observed no significant difference in the time prior to the initiation of rehabilitation and LOS, whereas differences in AFG remained significant.

Our findings also indicated that ward-dedicated PT staffing significantly improved AFG without decreasing the length of the period prior to rehabilitation. However, implementation of dedicated PT staffing did not decrease the length of this period, which was unexpected. Nonetheless, there may be a clinical explanation for this result. The study hospital encouraged early initiation of rehabilitation for acute medical patients. Physicians performed ERCP in patients with moderate to severe cholangitis within 2 days of hospitalization, and rehabilitation was initiated on the day after ERCP. The time to the initiation of rehabilitation decreased in both the dedicated and non-dedicated groups after the implementation of ward-dedicated staffing. In addition, there was no decrease in the LOS reported in previous studies. In comparison to the CVICU population, the following factors may have affected the LOS in the general medical ward staffing in this study: medical needs (antibiotic therapy duration) and social needs (discharge coordination and introduction of long-term care services). Further studies with strict discharge criteria are needed to evaluate the effect of dedicated PT staffing on LOS.

In a prospective cohort study, Brown et al.³³⁾ noted that low mobility and bed rest are common in hospitalized older patients and are important predictors of functional decline. Zisberg et al.³⁴⁾ demonstrated that in-hospital mobility is an important modifiable factor related to immediate and short-term functional outcomes in older adults. In addition, mobilization programs have been shown to encourage out-of-bed ADLs³⁵⁾ and prevent functional decline³⁶⁾. Lack of health literacy, which is defined as the ability to access, understand, evaluate, and use health information to make reasoned, health-related decisions³⁷⁾, has been associated with physical inactivity and frailty among patients^38–40). In our study, a dedicated physical therapist evaluated all patients before the initiation of regular rehabilitation, and inpatients with no functional decline at admission received instructions regarding self-exercise and appropriate physical activity. These care processes may have contributed to improvements in AFG by increasing health literacy.

Implications

In previous studies, acute inpatient rehabilitation typically focused on early mobilization and discharge planning^{3,4,6,17–19,41–43)}, and the impact of dedicated PT staffing was mainly investigated in ICU or nursing home settings^13–15). Our results suggest that ward-dedicated PT staffing is effective in improving AFG in general medical patients at an acute hospital. However, a 2017 survey revealed that only 2.4% (89 of 3748) of acute hospitals in Japan have ward-dedicated PT staffing. Our findings suggest that ward-dedicated PT staffing should be among the strategies utilized in the acute care process.

Study limitations

This study has several limitations. First, this study included a single-center hospital ward; thus, the generalizability of the methods and findings is limited. Therefore, further multi-center studies are required. Second, the ADLs evaluation was insufficient. BI data were missing, and a difference in BI at admission between each group was observed. In addition, the BI subitems were important in PT but could not be considered due to lack of data. Excluded patients may have had difficulty measuring BI data, and the present results may not be applicable to a population of patients with acute cholangitis for whom BI data would be difficult to measure. Moreover, as the baseline ADLs function (i.e., before admission) was unknown, discriminating between functional decline after hospitalization and baseline functional decline was difficult. However, ADLs function may always be uncertain given that it can be modified by several factors, such as the severity of disease and pain. Third, we could not have the data on daily rehabilitation time or duration of PT in the study because we could not have IDs that can be traced back to electronic medical record of the patients. Therefore, it is difficult to discuss the detail mechanism regarding the results of this study. Further studies are required to establish specific treatment protocols and to verify our findings in a large multi-center prospective cohort.

Conclusion

In the present observational study, we performed a difference-in-difference analysis of medical patients in an acute general hospital to elucidate the effect of ward-dedicated PT staffing. This quasi-experimental design allowed us to exclude the effects of natural temporal trends and confounding factors that limited previous studies. Our findings indicated that ward-dedicated PT staffing can improve AFG independent of the time at which rehabilitation is initiated in general medical patients admitted to an acute hospital.

Acknowledgments

The authors would like to thank all members of the general medical ward. They also thank Hiroki Matsui, MPH, for lending their expertise on the study design and Kimio Yamamoto, PT, for sharing dataset with them. Mihoko Takano, PT, and Yusuke Hirata, PT, performed the PT interventions described in this study and were instrumental to the execution of the ward-dedicated PT staffing. They are also indebted to Nao Hikota, PT, Kazuki Ikeda, PT, Tetsuya Shiraishi, PT, and Kaori Tokeshi, SLT, for their comments and suggestions.

Conflict of Interest

All authors have no competing interests to declare.