Acute Care Rehabilitation Services Utilization and Post-Acute Discharge Destination among Adults with Traumatic Brain Injury: The Moderating Effect of Functional and Physical Performance at Discharge

Abstract

Objective: To investigate whether the relationships between acute care occupational therapy (OT) and physical therapy (PT) utilization and community discharge are moderated by functional or physical performance at discharge among individuals hospitalized with traumatic brain injury (TBI). Setting: 14 acute care hospitals in the state of Colorado. Participants: We studied 5599 adults hospitalized with TBI between June 2018 and April 2021. Design: In a secondary analysis of de-identified electronic health record (EHR) data, multivariable moderation logistic regression models were performed to calculate odds ratios (ORs) for the likelihood of community discharge among patients who utilized OT/PT services. Main Measures: Functional (activities of daily living [ADL]) and physical (mobility) performance at discharge, OT and PT utilization, and community discharge status. Results: Overall, 67% of patients discharged to the community. The mean age of the sample was 55 years (SD = 20 years). Most participants were male (64%) and non-Hispanic White (72%). Mean hospital length of stay was 6 days (SD = 6 days). Both OT and PT utilization (OT: OR = 1.21, 95% CI [1.11, 1.33]; PT: OR = 1.22, 95% CI [1.14, 1.30]) and discharge ADL and mobility scores (ADL: OR = 1.34, 95% CI [1.30, 1.39]; mobility: OR = 1.38, 95% CI [1.33, 1.42]) were significantly and positively associated with community discharge. The OT and PT utilization-by-discharge ADL and mobility interaction terms yielded slightly negative, but statistically significant moderation effects in both models (ORs = 0.99, 95% CIs [0.98, 1.00]); indicating the magnitude of the OT and PT utilization effect diminished as ADL and mobility scores increased. Several sociodemographic characteristics and clinical factors were also independently associated with community discharge in both models (p-values < 0.001–0.04). Conclusions: Greater OT and PT utilization was associated with increased odds of community discharge. Similarly, higher ADL and mobility scores at discharge were associated with increased odds of community discharge. The small, but statistically significant negative interaction terms in both models indicated that the magnitude of the OT and PT utilization effect diminished as ADL and mobility scores increased. This study’s findings can guide occupational and physical therapists in their efforts to facilitate a safe transition to the community for patients with TBI.

1. Introduction

Traumatic brain injury (TBI) is a pressing public health concern in the United States, with approximately 2.5 million TBI-related emergency department visits; 288,000 TBI-related hospitalizations; and 61,000 TBI-related deaths reported each year [1]. TBI is linked to significant disability, often resulting in challenges with achieving basic activities of daily living (ADLs) like upper and lower body dressing, grooming and hygiene, or feeding and meal preparation, as well as difficulties in physical tasks such as ambulation, or transfers [2,3,4].

According to the Department of Health and Human Services, a safe discharge to the community following an acute care stay reflects the quality of the healthcare services delivered [5]. Community discharge has been associated with several positive impacts on patients, such as higher levels of functional independence [6,7,8], fewer cognitive or behavioral issues [7,9], lower healthcare cost [8,10], and enhanced quality of life [9,11]. Higher rates of community discharge were associated with rehabilitation services delivered during acute care stays [7,12,13,14,15].

In acute care settings, occupational and physical therapists assess patients’ functional and physical performance limitations, and evaluate their living conditions, to customize treatment plans aimed at ensuring a safe discharge to the community [16]. However, there is a gap in research providing empirical evidence on the factors that elucidate the relationship between the utilization of acute care rehabilitation services (e.g., OT and PT) and discharge location.

A key goal of acute care rehabilitation services is to ameliorate patients’ functional and physical performance [17]. Research has shown that rehabilitation services are associated with better functional and physical performance in patients with TBI [18,19]. Moreover, existing literature indicates that a higher rehabilitation services frequency leads to greater functional and physical improvements at the time of discharge from acute care for patients with TBI [12,19].

Patients with TBI who achieve higher functional and physical scores are more likely to be discharged to the community, as improved functional and physical performance during acute care rehabilitation ideally enables them to safely perform daily activities and manage basic mobility within their home environment [20]. Numerous studies have investigated the relationship between functional and physical status and discharge location for patients with TBI, revealing that TBI survivors who achieved greater functional and physical improvements were more likely to be discharged to the community [21,22,23]. To date, no research has investigated whether discharge functional and physical performance influences the relationship between the utilization of acute care rehabilitation services (e.g., OT and PT) and community discharge. Exploring how OT and PT utilization relates to community discharge could help improve the type and intensity of acute care OT and PT services provided to individuals with TBI, aiming to maximize the likelihood of a safe community discharge.

The aim of this study was to examine whether discharge functional or physical performance influences the relationship between the utilization of acute care OT and PT services and community discharge. We hypothesized that patients with greater OT and PT utilization would be more likely to have a community discharge and that this relationship would differ depending on patients’ functional or physical performances at discharge (e.g., patients with lower ability level would be less likely to have a community discharge).

2. Methods

2.1. Participants and Procedure

We conducted a retrospective cross-sectional study of de-identified electronic health record (EHR) data for patients admitted to 14 trauma centers, levels I to IV, within a single large health system in the state of Colorado. This study included EHR data from 401,350 patients admitted and discharged between June 2018 and April 2021. Inclusion criteria were an adult (aged ≥ 18 years), admitted to the hospital with a TBI diagnosis based on International Classification of Diseases (ICD-10) codes for admission, and had at least one OT/PT evaluation/treatment session. We excluded patients who did not survive the hospitalization or were missing data on key variables of interest. These data were validated, de-identified, organized, and supported by the Health Data Compass Data Warehouse project (healthdatacompass.org). After applying the inclusion criteria, the final sample consisted of 5599 adults (Figure 1). A signed data-use agreement was in place, and this study was approved by the Colorado State University Institutional Review Board. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines were applied to this study [24]. See (Appendix A,

Table A1. STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies.

	Item No	Recommendation	Addressed
Title and abstract	1	(a) Indicate the study’s design with a commonly used term in the title or the abstract	Abstract “Design” section
		(b) Provide in the abstract an informative and balanced summary of what was done and what was found	Abstract “Design” and “Results” sections
Introduction
Background/rationale	2	Explain the scientific background and rationale for the investigation being reported	Introduction paragraphs 1–5
Objectives	3	State specific objectives, including any prespecified hypotheses	Introduction final paragraph
Methods
Study design	4	Present key elements of study design early in the paper	Methods “Participation and Procedure” paragraph
Setting	5	Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, follow-up, and data collection	Methods “Participation and Procedure” paragraph
Participants	6	(a) Give the eligibility criteria, and the sources and methods of selection of participants	Methods “Participation and Procedure” paragraph
Variables	7	Clearly define all outcomes, exposures, predictors, potential confounders, and effect modifiers. Give diagnostic criteria, if applicable	Methods “Measures” section
Data sources/measurement	8 *	For each variable of interest, give sources of data and details of methods of assessment (measurement). Describe comparability of assessment methods if there is more than one group	Methods “Measures” section
Bias	9	Describe any efforts to address potential sources of bias	N/A
Study size	10	Explain how the study size was arrived at	N/A
Quantitative variables	11	Explain how quantitative variables were handled in the analyses. If applicable, describe which groupings were chosen and why	Methods “Measures” section
Statistical methods	12	(a) Describe all statistical methods, including those used to control for confounding	Methods “Data Analysis” section
		(b) Describe any methods used to examine subgroups and interactions	Methods “Data Analysis” section
		(c) Explain how missing data were addressed	Cases with missing data were excluded
		(d) If applicable, describe analytical methods taking account of sampling strategy	N/A
		(e) Describe any sensitivity analyses	N/A
Results
Participants	13 *	(a) Report numbers of individuals at each stage of study—eg numbers potentially eligible, examined for eligibility, confirmed eligible, included in the study, completing follow-up, and analysed	N/A
		(b) Give reasons for non-participation at each stage	N/A
		(c) Consider use of a flow diagram	Figure 1
Descriptive data	14 *	(a) Give characteristics of study participants (eg demographic, clinical, social) and information on exposures and potential confounders	Table 1
		(b) Indicate number of participants with missing data for each variable of interest	N/A
Outcome data	15 *	Report numbers of outcome events or summary measures	Table 1
Main results	16	(a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates and their precision (eg, 95% confidence interval). Make clear which confounders were adjusted for and why they were included	Table 1, Table 2 and Table 3
		(b) Report category boundaries when continuous variables were categorized
		(c) If relevant, consider translating estimates of relative risk into absolute risk for a meaningful time period	N/A
Other analyses	17	Report other analyses done—eg analyses of subgroups and interactions, and sensitivity analyses	N/A
Discussion
Key results	18	Summarise key results with reference to study objectives	Discussion paragraphs 1–2
Limitations	19	Discuss limitations of the study, taking into account sources of potential bias or imprecision. Discuss both direction and magnitude of any potential bias	Study Limitations section
Interpretation	20	Give a cautious overall interpretation of results considering objectives, limitations, multiplicity of analyses, results from similar studies, and other relevant evidence	Discussion paragraphs 3–8
Generalisability	21	Discuss the generalisability (external validity) of the study results	Conclusions section
Other information
Funding	22	Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based	N/A

* Give information separately for exposed and unexposed groups. An Explanation and Elaboration article discusses each checklist item and gives methodological background and published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely available) at “https://www.strobe-statement.org/checklists/ (accessed on 10 April 2024)”.

) for more information about STROBE statements.

2.2. Measures

2.2.1. OT/PT Utilization

OT/PT utilization was measured using service units billed for each OT or PT encounter (e.g., 1 unit =≥ 8 min through 22 min) during the patient’s acute care stay. See (Appendix A,

Table A2. Counting Minutes for Service Units.

Number of Service Units	Number of Minutes
1	(≥8 min through 22 min)
2	(≥23 min through 37 min)
3	(≥38 min through 52 min)
4	(≥53 min through 67 min)
5	(≥68 min through 82 min)
6	(≥83 min through 97 min)
7	(≥98 min through 112 min)
8	(≥113 min through 127 min)

Note: If only one service is provided in a day, providers should not bill for services lasting less than 8 min. For any single timed CPT code measured in 15-min units, providers should bill for one 15-min unit if the treatment time is between 8 and 22 min. If a single modality or procedure lasts between 23 and 37 min, 2 units should be billed [43].

) for more details about the service units and their equivalent billed minutes.

2.2.2. Functional/Physical Performance at Discharge

Patients functional (i.e., basic self-care) or physical (i.e., basic mobility) performance at discharge was measured using the Activity Measure for Post-Acute Care (AM-PAC^TM) “6-Clicks” daily activity inpatient short form [16]. The AM-PAC “6-Clicks” daily activity inpatient short form includes six activities of daily living (ADL) items: upper body dressing, lower body dressing, bathing, toileting, grooming, and eating [16]. The AM-PAC “6-Clicks” Basic Mobility short form includes six activities of physical performance: bed mobility, sit to stand, supine to sit, seated transfers, ambulation, and ascending stairs [16]. Each item is scored according to the level of assistance needed from a 1 indicating that total assistance is needed to complete the item to a 4 indicating that no assistance is needed to complete the item. The ADL and physical functioning items result in a total score ranging from 6 (i.e., indicating that the patient needs total assistance) to 24 (i.e., indicating that the patient requires no assistance) for each short form [16]. OT typically completes the ADL items and PT completes the physical performance items. The AM-PAC “6-Clicks” has very high internal consistency among OT and PT practitioners: Cronbach’s alpha = 0.91 and 0.96 for ADL and mobility assessments, respectively [16]. The AM-PAC “6-Clicks” scores are standardized [16], with higher z-scores indicating greater ADL performance (range = 17.07–57.54), as well as greater basic mobility abilities (range = 23.55–61.14).

2.2.3. Community Discharge

We constructed a binary indicator of community discharge (yes/no). Community discharge included home and other supported living facilities (e.g., senior living facility/assisted living facility). A non-community discharge includes all institutional discharges including long-term care, mental health facility, nursing facility, rehabilitation facility, short term hospital, and skilled nursing facility.

2.2.4. Covariates

We included person-level factors such as age (years); sex (female/male); race/ethnicity (White, Black, Hispanic, Multiple race, and Other [e.g., Asian, American Indian, Alaska native, native Hawaiian, and Pacific Islander]); presence of a significant other (yes/no); insurance type (e.g., Medicare, Medicaid, VA, Others, and Private); length of stay (days); comorbidity burden (using Functionally relevant TBI Comorbidity Index [Fx-TBI-CI] [25]), and TBI severity (e.g., mild, moderate, and severe [26]) as covariates. The Fx-TBI-CI is a method used to categorize patients’ comorbidities based on the International Classification of Diseases (ICD-10) diagnosis codes obtained from hospital administrative data [25]. The Fx-TBI-CI was calibrated based on the function of individuals with TBI receiving inpatient care [25]. We constructed a weighted summary index according to Kumar et al. to evaluate comorbidity burden [25]. We used ICD-10 diagnosis codes obtained from the Defense and Veterans Brain Injury Center (DVBIC) to classify TBI severity as mild, moderate, or severe [26].

2.3. Data Analysis

Descriptive statistics (e.g., means, standard deviations, percentages) were calculated, both for the total sample and stratified by community discharge. We used bivariate analyses to examine unadjusted differences in patient characteristics between community discharge groups (yes/no). None of the seven continuous variables met the normality assumption for parametric testing so we used Mann–Whitney U Tests for those variables. Chi-square tests were used for categorical variables. Multivariable moderation logistic regression models were used to assess the community discharge status as the dependent variable and rehabilitation services utilization (e.g., OT, and PT) and patients functional (i.e., ADL) or physical (i.e., mobility) performance at discharge as the main predictors of interest. In the OT model, we computed the main effect of OT utilization on community discharge, the main effect of functional performance (ADL) scores at discharge on community discharge, and the moderating effect of ADL scores on the relationship between OT utilization and community discharge by including an OT utilization-by-discharge ADL score interaction term. Similarly, in the PT model, we computed the main effect of PT utilization on community discharge, the main effect of physical performance (mobility) scores at discharge on community discharge, and the moderating effect of mobility scores on the relationship between PT utilization and community discharge by including a PT utilization-by-discharge mobility score interaction term. Estimates (e.g., odds ratios, confidence intervals) were adjusted for age, sex, race/ethnicity, presence/absence of a significant other, insurance type, length of stay, comorbidity burden, and TBI severity levels. Statistical significance was evaluated at α = 0.05 for all parameter estimates. All analyses were performed using R (Version 4.3.1, R Foundation for Statistical Computing, Vienna, Austria) [27].

3. Results

Among the 5599 patients included in this study, 67% were discharged to the community. The sample had a mean age of 54.7 years (SD = 20.2), with the majority being male (64%), White (72%), and having no significant other (58%). The most common insurance type was Medicare (39%), followed by Medicaid (26%). Most patients experienced moderate TBI severity (83%) (

Table 1. Sample characteristics with descriptive summaries of community discharge.

		Community Discharge, n (%)
	N (%)	No	Yes	p-Value
Total	5599	1825 (32.6%)	3774 (67.4%)
OT utilization, in units, mean (SD) *	3.9 (4.1)	4.9 (4.1)	3.4 (4.0)	<0.001
Discharge ADL score, in AM-PAC, mean (SD) *	19.8 (4.3)	17.1 (4.7)	21.3 (3.1)	<0.001
PT utilization, in units, mean (SD) *	5.1 (5.7)	6.4 (5.4)	4.4 (5.7)	<0.001
Discharge Mobility score, in AM-PAC, mean (SD) *	20.0 (4.3)	17.2 (4.8)	21.5 (3.1)	<0.001
Age in years, mean (SD)	54.7 (20.2)	61.7 (19.5)	51.3 (19.6)	<0.001
Sex				<0.001
Female	2002 (35.8%)	733 (36.6%)	1269 (63.4%)
Male	3597 (64.2%)	1092 (30.1%)	2505 (69.9%)
Race/Ethnicity				<0.001
White	4010 (71.6%)	1381 (34.4%)	2629 (65.6%)
Black	341 (6.1%)	95 (27.9%)	246 (72.1%)
Hispanic	877 (15.7%)	212 (24.2%)	665 (75.8%)
Multiple race	103 (1.8%)	42 (40.8%)	61 (59.2%)
Other	268 (4.8%)	95 (35.4%)	173 (64.6%)
Significant other				0.006
Yes	2337 (41.7%)	714 (30.6%)	1623 (69.4%)
No	3262 (58.3%)	1111 (34.1%)	2151 (65.9%)
Insurance Type				<0.001
Medicare	2185 (39.0%)	998 (45.7%)	1187 (54.3%)
Medicaid	1439 (25.7%)	358 (24.9%)	1081 (75.1%)
VA	155 (2.8%)	33 (21.3%)	122 (78.7%)
Other	396 (7.1%)	66 (16.7%)	330 (83.3%)
Private	1424 (25.4%)	370 (25.9%)	1054 (74.1%)
TBI Severity				<0.001
Mild	835 (14.9%)	207 (24.8%)	628 (75.2%)
Moderate	4629 (82.7%)	1536 (33.2%)	3093 (66.8%)
Severe	135 (2.4%)	82 (60.7%)	53 (39.3%)
Length of stay in day, mean (SD)	5.8 (5.6)	8.7 (6.8)	4.4 (4.3)	<0.001
Community Density				0.504
Urban	5112 (91.3%)	1673 (32.7%)	3439 (67.3%)
Rural	487 (8.7%)	152 (31.2%)	335 (68.8%)
FX-TBI-CI, mean (SD)	1.0 (2.4)	1.9 (3.0)	0.6 (1.9)	<0.001

SD = standard deviation; OT = occupational therapy; PT = physical therapy; ADL = activity of daily living; AM-PAC = Activity Measure for Post-Acute Care; FX-TBI-CI = functionally relevant TBI comorbidity index; * OT/PT utilization units and discharge ADL/Mobility scores only include the subset of patients who received those services n = 4585 and 4769, respectively; p-value for the bivariate analyses: Chi-square, t-test, and ANOVA; bold value = significant p-value < 0.05.

). There were statistically significant differences between OT and PT utilization, and discharge ADL and mobility scores on community discharge (p-values ≤ 0.001). The mean OT utilization units were 3.9 (SD = 4.1) and the average discharge ADL score was 19.8 (SD = 4.3) (scores closer to 24 indicated that the patient requires no assistance). The mean PT utilization units were 5.1 (SD = 5.7) and the average discharge mobility score was 20.0 (SD = 4.3) (scores closer to 24 indicate that the patient requires no assistance). There were statistically significant differences in community discharge for patients’ sex (p < 0.001), race/ethnicity (p-value < 0.001), significant other status (p-value = 0.006), insurance type (p < 0.001), and TBI severity groups (p < 0.001). All five continuous covariates also differed significantly between groups (p-value < 0.001). See Table 1 for detailed summary statistics.

There was sufficient evidence to suggest that OT and PT utilization, and ADL and mobility score at discharge were good predictors of community discharge (p-values < 0.001) (

Table 2. Results of moderation logistic regression models for community discharge.

	OT Moderation Model			PT Moderation Model
	Beta Coefficients	Standard Error (SE)	p-Value	Beta Coefficients	Standard Error (SE)	p-Value
Intercept	−2.89	0.39	<0.001	−3.29	0.39	<0.001
OT utilization units	0.19	0.05	<0.001	--	--	--
Discharge ADL score (AM-PAC)	0.30	0.02	<0.001	--	--	--
OT utilization * Discharge ADL score	−0.01	0.002	<0.001	--	--	--
PT utilization units	--	--	--	0.20	0.03	<0.001
Discharge Mobility score (AM-PAC)	--	--	--	0.32	0.02	<0.001
PT utilization * Discharge Mobility score	--	--	--	−0.01	0.002	<0.001
Age (years)	−0.02	0.002	<0.001	−0.02	0.002	<0.001
Sex (ref. = Female)	0.03	0.08	0.740	−0.03	0.08	0.686
Race/Ethnicity (ref. = White)
Black	0.53	0.19	0.007	0.48	0.19	0.007
Hispanic	0.39	0.12	<0.001	0.26	0.12	0.026
Multiple race	0.05	0.3	0.883	−0.28	0.29	0.334
Other	0.19	0.19	0.324	0.28	0.19	0.143
Significant other (ref. = No)	0.48	0.09	<0.001	0.44	0.08	<0.001
Insurance Type (ref. = Private)
Medicare	−0.42	0.12	<0.001	−0.39	0.12	<0.001
Medicaid	0.22	0.13	0.084	0.27	0.13	0.026
VA	−0.22	0.28	0.443	−0.19	0.30	0.526
Other	0.54	0.21	0.01	0.56	0.21	0.01
TBI severity (ref. = Mild)
Moderate	−0.23	0.12	0.04	−0.26	0.12	0.027
Severe	−0.83	0.31	<0.001	−0.68	0.29	0.018
Length of stay (days)	−0.15	0.01	<0.001	−0.15	0.01	<0.001
Residence (ref. = Rural)
Urban	−0.21	0.14	0.154	−0.23	0.14	0.113
FX-TBI-CI	−0.08	0.002	<0.001	−0.10	0.02	<0.001

OT = occupational therapy; PT = physical therapy; ADL = activities of daily living; AM-PAC = Activity Measure for Post-Acute Care; * = OT and PT interaction terms (i.e., OT utilization-by-discharge ADL score and PT utilization-by-discharge mobility score); Ref. = reference category; VA = Veterans Affairs; FX-TBI-CI = functionally relevant TBI comorbidity index; -- = not applicable; bold value = significant p-value < 0.05.

). OT and PT utilization illustrated statistically significant positive associations with the log odds of community discharge for patients who received OT ($\widehat{\beta }$ = 0.19; SE = 0.05; p < 0.001), and PT ($\widehat{\beta }$ = 0.20; SE = 0.03; p < 0.001). Patients’ ADL and mobility scores at discharge demonstrated statistically significant positive associations with the log odds of community discharge for patients who received OT ($\widehat{\beta }$ = 0.30; SE = 0.02; p < 0.001), and PT ($\widehat{\beta }$ = 0.32; SE = 0.02; p < 0.001). Both OT and PT interaction terms (i.e., OT utilization-by-discharge ADL score and PT utilization-by-discharge mobility score) showed a small but significant negative association with the log odds of community discharge ($\widehat{\beta }$ = −0.01; SE = 0.002; p < 0.001). Age demonstrated statistically significant negative associations with the log odds of community discharge for patients who received OT and PT ($\widehat{\beta }$ = −0.02; SE = 0.002; p < 0.001). Black and Hispanic patients were significantly and positively associated with the log odds of community discharge for patients who received OT (Black: $\widehat{\beta }$ = 0.53; SE = 0.19; p-value = 0.01; Hispanic: $\widehat{\beta }$ = 0.39; SE = 0.12; p < 0.001), and PT (Black: $\widehat{\beta }$ = 0.48; SE = 0.19; p-value = 0.01; Hispanic: $\widehat{\beta }$ = 0.26; SE = 0.12; p-value = 0.03) relative to their non-Hispanic White counterparts.

The presence of a significant other was significantly and positively associated with the log odds of community discharge for patients who received OT ($\widehat{\beta }$ = 0.48; SE = 0.09; p < 0.001), and PT ($\widehat{\beta }$ = 0.44; SE = 0.08; p < 0.001) relative to patients without a significant other. Compared to patients with private insurance, patients with Medicare insurance were significantly and negatively associated with the log odds of community discharge for patients who received OT ($\widehat{\beta }$ = −0.42; SE = 0.12; p < 0.001), and PT ($\widehat{\beta }$ = −0.39; SE = 0.12; p < 0.001). In comparison to patients with mild TBI, patients with severe TBI were significantly and negatively associated with the log odds of community discharge for patients who received OT ($\widehat{\beta }$ = −0.83; SE = 0.31; p < 0.001), and PT ($\widehat{\beta }$ = −0.68; SE = 0.29; p-value = 0.02). Comorbidity burden score was significantly and negatively associated with the log odds of community discharge for patients who received OT ($\widehat{\beta }$ = −0.08; SE = 0.002; p < 0.001), and PT ($\widehat{\beta }$ = −0.10; SE = 0.02; p < 0.001). Longer lengths of stay in acute care was significantly and negatively associated with the log odds of discharge to the community for patients who received OT and PT ($\widehat{\beta }$ = −0.15; SE = 0.01; p < 0.001). There was not sufficient evidence to suggest that the other covariates (i.e., sex, and residence location) were significantly associated with community discharge (p-values = 0.11–0.74). Table 2 presents the coefficient estimates from the logistic regression analysis.

The moderating effect of those variables on the relationship between OT and PT utilization and community discharge was negative and statistically significant (ORs = 0.99, 95% CIs [0.98, 1.00]). This indicates that the magnitude of the OT and PT utilization effect diminished as ADL and mobility scores increased. The main effect of OT and PT utilization was significantly positively associated with community discharge: (OR = 1.21, 95% CI [1.11, 1.33]); (OR = 1.22, 95% CI [1.14, 1.30]), respectively (

Table 3. Results of pairwise comparisons (odds ratios) for community discharge.

	Community Discharged
	OT Moderation Model			PT Moderation Model
	OR	SE	95% CI	OR	SE	95% CI
Intercept	0.06	0.02	[0.02, 0.12]	0.04	0.01	[0.02, 0.08]
OT utilization units	1.21	0.05	[1.11, 1.33]	--	--	--
Discharge ADL score (AM-PAC)	1.34	0.02	[1.30, 1.39]	--	--	--
OT utilization * Discharge ADL score	0.99	0.002	[0.98, 1.00]	--	--	--
PT utilization units	--	--	--	1.22	0.04	[1.14, 1.30]
Discharge Mobility score (AM-PAC)	--	--	--	1.38	0.02	[1.33, 1.42]
PT utilization * Discharge Mobility score	--	--	--	0.99	0.002	[0.98, 1.00]
Race/Ethnicity
White/Other	0.83	0.15	[0.57, 1.21]	0.75	0.14	[0.52, 1.10]
White/Multiple race	0.96	0.28	[0.53, 1.73]	1.33	0.39	[0.75, 2.35]
White/Hispanic	0.67	0.08	[0.53, 0.87]	0.76	0.09	[0.61, 0.98]
White/Black	0.59	0.11	[0.41, 0.86]	0.62	0.12	[0.43, 0.90]
Other/Multiple race	1.15	0.40	[0.58, 2.30]	1.75	0.59	[0.90, 3.41]
Other/Hispanic	0.82	0.18	[0.54, 1.26]	1.01	0.22	[0.67, 1.54]
Other/Black	0.71	0.19	[0.43, 1.19]	0.82	0.21	[0.49, 1.35]
Multiple race/Hispanic	0.71	0.19	[0.38, 1.33]	0.58	0.18	[0.32, 1.06]
Multiple race/Black	0.62	0.18	[0.31, 1.22]	0.47	0.16	[0.24, 0.90]
Hispanic/Black	0.87	0.20	[0.57, 1.32]	0.81	0.17	[0.53, 1.22]
Age (year)	0.98	0.002	[0.97, 0.98]	0.98	0.002	[0.97, 0.99]
Significant other
No/Yes	0.62	0.05	[0.53, 0.74]	0.64	0.05	[0.55, 0.76]
Insurance Type
Private/Medicare	1.53	0.19	[1.20, 1.95]	1.48	0.18	[1.17, 1.89]
Private/Medicaid	0.80	0.10	[0.63, 1.03]	0.76	0.09	[0.59, 0.97]
Private/VA	1.25	0.35	[0.72, 2.18]	1.21	0.36	[0.67, 2.17]
Private/Other	0.58	0.12	[0.39, 0.87]	0.57	0.12	[0.38, 0.86]
Medicare/Medicaid	0.53	0.07	[0.40, 0.69]	0.51	0.07	[0.39, 0.67]
Medicare/VA	0.82	0.24	[0.46, 1.45]	0.81	0.25	[0.44, 1.49]
Medicare/Other	0.38	0.08	[0.25, 0.58]	0.39	0.08	[0.25, 0.59]
Medicaid/VA	1.55	0.81	[0.89, 2.72]	1.59	0.48	[0.88, 2.87]
Medicaid/Other	0.72	0.15	[0.48, 1.09]	0.75	0.16	[0.50, 1.14]
VA/Other	0.47	0.15	[0.24, 0.89]	0.47	0.16	[0.24, 0.93]
TBI Severity
Mild/Moderate	1.26	0.15	[1.00, 1.58]	1.30	0.15	[1.03, 1.63]
Mild/Severe	2.28	0.72	[1.23, 4.23]	1.97	0.58	[1.11, 3.50]
Moderate/Severe	1.81	0.54	[1.01, 3.26]	1.52	0.42	[0.89, 2.61]
Length of stay (days)	0.86	0.01	[0.85, 0.88]	0.86	0.01	[0.84, 0.88]
FX-TBI-CI	0.93	0.02	[0.90, 0.96]	0.91	0.02	[0.88, 0.94]

OR = odds ratio; SE = standard error; CI = confidence intervals; bold values = p < 0.05; -- = not applicable; OT = occupational therapy; PT = physical therapy; * = OT and PT interaction terms (i.e., OT utilization-by-discharge ADL score and PT utilization-by-discharge mobility score).

). Specifically, patients with greater OT/PT utilization were 1.21 to 1.22 times more likely to be discharged to the community. The main effect of functional performance (ADL) and physical performance (mobility) scores at discharge demonstrated statistically significant positive associations with community discharge: (OR = 1.34, 95% CI [1.30, 1.39]); (OR = 1.38, 95% CI [1.33, 1.42]), respectively. This means that patients with higher ADL and mobility performance scores at discharge were 1.34 to 1.38 times more likely to be discharged to the community.

Several covariates were significantly associated with community discharge. White individuals demonstrated statistically significant negative associations with community discharge relative to their Hispanic and Black counterparts in both models (OT model White vs. Hispanic (OR = 0.67, 95% CI [0.53, 0.87]), White vs. Black (OR = 0.59, 95% CI [0.41, 0.86]); PT model White vs. Hispanic (OR = 0.76, 95% CI [0.61, 0.98]), White vs. Black (OR = 0.62, 95% CI [0.43, 0.90]). This implies that non-Hispanic White patients were 0.67 to 0.76 times less likely to be discharged to the community compared to their Hispanic counterparts in both OT and PT models. Similarly, non-Hispanic White patients 0.59 to 0.62 times less likely to be discharged to the community compared to their Black counterparts in both OT and PT models. Age demonstrated statistically significant associations with community discharge (ORs = 0.98, 95% CIs [0.97, 0.99]) in both OT and PT models. Specifically, for each one-year increase in age, older patients were 0.98 times less likely to be discharged to the community relative to younger patients. Compared to patients with significant others, patients with no significant other were less likely to experience community discharge (OR = 0.62, 95% CI [0.53, 0.74]); (OR = 0.64, 95% CI [0.55, 0.76]), for OT and PT models, respectively. Specifically, individuals without a significant other were 0.62 to 0.64 times less likely to experience community discharge. Using patients with Medicare insurance as the reference group, patients with private insurance were 1.48 to 1.53 times more likely to be discharged to the community (OT: OR = 1.53, 95% CI [1.20, 1.95]); (PT: OR = 1.48, 95% CI [1.20, 1.95]). Using severe or moderate TBIs as reference categories, patients with mild TBI were 1.26 to 2.28 times more likely to be discharged to the community (OT model: Mild vs. moderate TBI OR = 1.26, 95% CI [1.00, 1.58]); mild vs. severe TBI OR = 2.28, 95% CI [1.23, 4.23]), PT model: Mild vs. moderate TBI OR = 1.30, 95% CI [1.03, 1.63]); mild vs. severe TBI OR = 1.97, 95% CI [1.11, 3.50]). Length of stay was associated with community discharge in both OT and PT models (ORs = 0.86, 95% CIs [0.84, 0.88]). Particularly, patients with longer lengths of stay were 0.86 times less likely to be discharged to the community relative to patients with shorter lengths of stay. Patients with greater comorbidity burden (i.e., functional comorbidity index [FX-TBI-CI] scores) were 0.91 to 0.93 times less likely to be discharged to the community (OR = 0.93, 95% CI [0.90, 0.96]) in OT model and (OR = 0.91, 95% CI [0.88, 0.94]) in PT model. On the other hand, sex, and community density variables did not show significant associations with community discharge in either model (p-values = 0.11–0. 0.69). See Table 3. Refer to Figure 2 and Figure 3 for illustrations of the moderation models.

4. Discussion

This study represents the first investigation on whether ADL and mobility performance scores at discharge moderate the relationships between acute care OT and PT utilization and community discharge among adults with TBI. Greater OT and PT utilization was associated with 1.21 to 1.22 times increased odds of being discharged to the community, respectively. Additionally, a higher ADL and mobility performance score at discharge was independently associated with 1.34 to 1.38 times greater odds of community discharge, respectively. Subsequently, the small, but statistically significant negative interaction terms in both models indicated that the magnitude of the OT and PT utilization effect diminished as ADL and mobility scores increased (ORs = 0.99, 95% CIs 0.98, 1.00]). Results may inform efforts to enhance the type or amount of acute care OT or PT services delivered to individuals with TBI to maximize safe community discharge. Results will also offer valuable insights for future research, particularly in exploring the interaction between the amount of acute care OT and PT services utilization, discharge ADL and mobility performance, and their impact on community discharge.

Our findings support our hypothesis that greater OT or PT utilization would be more likely to have a community discharge and that this relationship would differ depending on patients’ functional or physical performances at discharge. Interestingly, both OT and PT utilization were negatively associated with community discharge in the preliminary bivariate analyses (Table 1). This suggests confounding by indication. Therapy intensity is based on need. Thus, patients with lower functional status and who are, accordingly, less likely to discharge to the community typically receive more acute therapy. However, when controlling for differences in patient characteristics (including functional status) in the multivariable model, both OT and PT utilization were positively associated with community discharge. These results are consistent with previous research indicating that greater OT and PT utilization is associated with an increased likelihood of community discharge [13,14,15]. Our findings also align with prior studies reporting that a higher ADL and mobility performance score at discharge is associated with an increased likelihood of community discharge [21,22]. However, we also found that the relationship between service utilization and discharge setting varies across discharge ADL and mobility scores. Meaning that the magnitude of the OT and PT utilization effect diminished as ADL and mobility scores increased. In our study, the observed negative statistically significant influence of ADL and mobility scores at discharge between rehabilitation services utilization and community discharge may be attributed to patients’ cognitive, affective, or behavioral sequalae of TBI, which are often more disabling than objective ADL and physical performance limitations [28]. In our sample, patients with TBI may show a high ADL and mobility performance score at discharge while exhibiting some personality changes such as impulsivity, irritability, or apathy. Thus, patients with high ADL and mobility scores may require an institutional discharge to address impairments not measured by the ‘6-Clicks’ Basic Mobility or Daily Activity short forms. There is a ‘6-Clicks’ Applied Cognitive Short Form; however, these data were not yet entered into EHR data. Future research should investigate the relationship between acute care OT and PT utilization and community discharge, while accounting for patients’ cognitive abilities, to better address individuals with TBI prognosis following an acute care setting.

Our findings showed that non-Hispanic White patients with TBI were less likely to be discharged to the community relative to both Hispanic and Black patients (OT model White vs. Hispanic (OR = 0.67, 95% CI [0.53, 0.87]), White vs. Black (OR = 0.59, 95% CI [0.41, 0.86]); PT model White vs. Hispanic (OR = 0.76, 95% CI [0.61, 0.98]), White vs. Black (OR = 0.62, 95% CI [0.43, 0.90]). This finding is consistent with prior research indicating that non-Hispanic White patients were more likely to experience an institutional discharge (e.g., inpatient rehabilitation facilities) compared to non-ethnic minority patients [29,30]. The effect of ethnic minority status on community discharge may be due to socioeconomic barriers (e.g., lack of health insurance, lack of transportation, or geographic region) [31,32]. Thus, future research should investigate the impact of ethnic minority status on discharge disposition to better understand how could we support the implementation of the health equity framework in acute care.

Our findings revealed a negative association between a patient’s age and their likelihood of community discharge (ORs = 0.98, 95% CIs [0.97, 0.99]). This finding aligns with previous studies indicating that increased age was associated with institutional discharge in patients with TBI [19]. Older patients in our sample may represent patients with serious premorbid conditions who have demonstrated poor functional/physical outcomes at the time of discharge [33]. As a result, these patients may require further inpatient services to continue improving their independence in everyday activities and mobility prior to discharge to the community. In our study, older patients may have not fully recovered during their acute care stay, making institutional discharge the most medically appropriate plan. Also, in our study, older patients may represent patients whose prior living situation was not in the community. More studies on the influence of age on discharge destination are necessary to understand potential disparities in post-acute discharge disposition while accounting for prior living status.

Our findings emphasized a significant association between the presence of a significant other and the likelihood of community discharge. Compared to patients with significant others, patients with no significant other were less likely to experience community discharge: (OR = 0.62, 95% CI [0.53, 0.74]); (OR = 0.64, 95% CI [0.55, 0.76]), for OT and PT models, respectively. This finding aligns with previous studies indicating that patients who have caregivers were more likely to be discharged home [34]. The influence of having a significant other on discharge disposition could be related to patients’ goals and preferences for post-acute care and treatment [5]. More research is needed to accurately describe the effect of significant other status on healthcare utilization across settings and discharge disposition.

Using patients with Medicaid insurance as the reference group, we observed that patients with private insurance had lower odds of being discharged to the community following PT (OR = 0.76, 95% CI [0.59, 0.97]). This finding aligns with a recent study indicating that patients with public insurance were more likely to discharge to the community relative to patients with private insurance [35]. The influence of insurance type on discharge disposition could be related to patients’ socioeconomic status which impacts discharge disposition [36]. Hence, future studies are warranted to explore the impact of financial barriers, including insurance types and coverages, on safe discharge planning across various settings and systems.

Our findings elucidated significant differences in community discharge based on TBI severity and comorbidity burdens. Our results indicated that compared to patients with severe TBI and greater comorbidity burdens, patients with mild TBI and lower comorbidity burdens were substantially more likely to discharge to the community ((OT model: mild vs. severe TBI OR = 2.28, 95% CI [1.23, 4.23]); (comorbidity burden OR = 0.93, 95% CI [0.90, 0.96])); ((PT model: mild vs. severe TBI OR = 1.97, 95% CI [1.11, 3.50]); (comorbidity burden OR = 0.91, 95% CI [0.88, 0.94])). These findings are consistent with prior research indicating that patients with greater TBI injury severity and comorbidity burden were less likely to be discharged to the community relative to patients with lesser TBI injury severity and comorbidity burden [37,38]. In our sample, patients with TBI who have a greater comorbidity burden may be more likely to have medical complications such as paroxysmal sympathetic hyperactivity that may pose a barrier to safe community discharge [39,40,41]. Future research should examine the influence of TBI severity level and comorbidity burden on community discharge across different medical settings.

Our study revealed a significant negative relationship between length of stay in acute care and the likelihood of community discharge (ORs = 0.86, 95% CIs [0.84, 0.88]). This finding aligns with previous literature indicating that patients with longer lengths of stay were less likely to be discharged to the community relative to patients with shorter stays [21,42]. The influence of length of stay on discharge disposition could be related to patients’ functional and physical status, impacting community discharge. Patients with longer lengths of stay may have greater medical complications that may pose a barrier to community discharge [40,41]. Thus, future studies should examine the influence of patients’ lengths of stay and accompanied medical complications on the likelihood of community discharge.

5. Study Limitations

This study is subject to a few limitations. Firstly, the reliance on EHR data introduces the possibility of missing information, misclassifications, and coding and reporting errors. Additionally, our investigation was confined to 14 hospitals within a single large health system, which may limit the generalizability of our findings to other health systems. It is crucial for future research to replicate our findings in diverse health systems to assess the applicability of our results to geographically varied patients with TBI. Our dataset had no information about prior living situation; thus, future research should explore the influence of patients’ prior living situation on discharge location. Additionally, we did not include cognitive assessment data in our analyses. Standardized cognitive assessments are often not administered to all patients in acute care settings or are not recorded in standard electronic flow sheets. Future studies should examine the influence of patients’ cognitive status, if available, on patients’ discharge disposition. Our dataset also had no information about prior functional status, current neurological deficits, home health OT/PT services, or whether a patient underwent surgery during their acute care stay. Future research should explore the influence of patients’ prior functional status, the presences of neurological deficits, home health OT/PT services, and surgeries on discharge ADL/mobility score and specific discharge locations. Our study did not examine specific discharge destinations, such as Inpatient Rehabilitation Facility (IRF) versus skilled nursing facility; instead, discharge disposition was categorized broadly into community versus institutional. Future research should explore the association between OT/PT utilization and specific discharge destinations to provide a more detailed understanding of the determinants of discharge location among patients hospitalized with TBI in acute care settings.

6. Conclusions

We examined whether ADL and mobility performance score at discharge moderated the relationship between acute care OT and PT utilization and discharge to the community among adults with TBI. Both ADL and mobility performance scores at discharge moderated the relationship between OT and PT utilization and community discharge. Higher OT and PT utilization was associated with increased odds of being discharged to the community and higher ADL and mobility performance score at discharge was independently associated with greater odds of community discharge. The observed positive statistically significant effects of OT and PT utilization simply decreased with increasing discharge ADL and mobility scores. Further research is warranted to determine the generalizability of our findings, evaluate the impact of patients’ prior health and cognitive status, and other sociodemographic factors (e.g., age, ethnicity, insurance type, and significant other status) on the relationship between OT and PT utilization and community discharge. Findings of this study can inform occupational and physical therapists’ efforts aimed at enhancing the type or amount of acute care OT and PT services delivered to individuals with TBI to maximize safe community discharge.