TRAUMATIC BRAIN INJURY (TBI) is increasingly conceptualized as a chronic health condition with lifelong and, in some cases, progressive health problems that warrant ongoing monitoring by clinicians with expertise in the field of brain injury medicine.1,2 Individuals with a moderate-severe TBI appear to be at risk for medical and mental health comorbidities for many years after injury,3–6 and this disease burden has been found to have an adverse relationship with functional outcome.7 Factors such as age and premorbid and postinjury health conditions may interact to enhance this risk.8
Readmission to the hospital following an inpatient rehabilitation unit stay due to new TBI may be regarded as a reflection of postinjury health and related healthcare burden.9 Recent work has demonstrated an important demographic shift toward aging of the TBI rehabilitation population.10 One study showed an association of poorer outcomes among older adults hospitalized within 1 year following TBI,11 and another found rehospitalization to be associated with poorer participation in one's own health management.12 Rehospitalization may have many detrimental consequences as it may disrupt community reintegration, cause the loss of a patient's home care services, and lead to secondary medical complications (such as nosocomial infections) that can result in significant morbidity and possibly mortality. Rehospitalization has been increasingly tied to hospital reimbursement, representing an obstacle to value-based care and therefore is a growing financial concern for inpatient rehabilitation units and hospital systems as a whole.
Several studies have examined risk factors for rehospitalization. One study found that in a 3-year follow-up period after TBI, the risk of rehospitalization increased with male gender, older age, greater injury severity, mechanism of injury, and comorbid physical or mental health problems.13 A prospective study found that rehospitalization rates increase with severity of TBI and were highest for those with disorders of consciousness.14 Another study of 510 participants that looked at a 9-month follow-up period found that various medical and surgical reasons were found to contribute to an increased risk of rehospitalization in 28% of the participants.9 A study that looked at outcomes following neurosurgical management of TBI found that those who required hemicraniectomy as opposed to craniotomy had a higher rate of rehospitalization.15 Older adults with multiple comorbidities were found to have an increased rate of rehospitalization at 1 year postinjury.8 Together, these studies demonstrate that more severe TBI, comorbid conditions, and invasive surgical procedures generally pose a higher risk of rehospitalization.
It is clear that rehospitalization is common and poses major healthcare burden for patients recovering from TBI. However, prior studies are limited by brief follow-up time frames8,9,11,16 and used only broad imprecise categories to characterize primary reasons for rehospitalization.16,17 A prior TBI Model Systems (TBIMS) study examined longer-term rates of rehospitalization up to 10 years following moderate-severe TBI and inpatient rehabilitation and found rates were highest in the first year following injury (28%) and mostly attributable to orthopedic causes. At successive time points (2, 5, and 10 years postinjury), rates of rehospitalization are relatively consistent (22.1%-23.4%), with general health reasons being the most common.17 More detailed information about risk profiles and reasons for rehospitalization has since been added to the TBIMS National Database (NDB). The goal of this article is to use this expanded diagnostic coding scheme in the TBIMS NDB to investigate the most common reasons for rehospitalization over 10 years following TBI. We further expand upon prior work that included only baseline predictors9,13,14,16,17 by using information collected following hospital discharge to identify risk factors that are more proximal to rehospitalization.
METHODS
Data source and participants
Data from the current study were collected as part of the TBIMS NDB.18 The TBIMS is a multicenter prospective longitudinal study of individuals with moderate to severe TBI enrolled during inpatient rehabilitation and who are followed prospectively postinjury at 1, 2, 5, and every 5 years thereafter. The data used in the current study come from sites funded by the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR) within the Agency on Community Living, Department of Health and Human Services. TBIMS inclusion criteria include having sustained a TBI meeting at least 1 of following criteria: Glasgow Coma Scale score less than 13 on emergency admission (not because of intubation, sedation, or intoxication), loss of consciousness of 30 minutes or more unrelated to sedation or intoxication, posttraumatic amnesia of more than 24 hours, or trauma-related intracranial abnormality on neuroimaging. In addition, all participants must be 16 years or older, receive their medical care within the TBIMS-affiliated system within 72 hours of injury, and complete inpatient rehabilitation within the system. Participants provide consent or consent can occur via legal proxy. More details on the database, data collected during inpatient rehabilitation (Form I), and at the follow-up assessments (Form II) are available at www.tbindsc.org. Participants in the current study had TBIMS follow-up interviews on or after October 1, 2017. At this date, the rehospitalization variable was updated in the TBIMS NDB (detailed later). Participants were included in the current investigation if they had at least one follow up at 1, 2, 5, or 10 years postinjury and provided information on rehospitalization via self- or proxy report.
Variables
Primary outcome: Rehospitalization
Information is collected from participants with TBI and/or their proxies regarding whether they were ever rehospitalized after discharge from inpatient rehabilitation, and if so, how many rehospitalizations (up to 5), as well as the primary reason for each rehospitalization. The query refers to all types of causes for any inpatient admission greater than 24 hours in any hospital but does not include emergency department or urgent care visits of less than 24 hours. Per the TBIMS follow-up protocol, participants were asked at 1, 2, 5, and 10 years postinjury whether they had been rehospitalized in the past year.
To characterize reasons for rehospitalization, we used the rehospitalization coding system from the Healthcare Cost and Utilization Project (HCUP) National Readmissions Database (NRD), which was expanded in 2015 with the introduction of ICD-10 (International Classification of Diseases, Tenth Revision) diagnostic codes. The NRD can be used to track vital information related to hospital readmissions including cost analysis as it relates to hospital reimbursements. Data collectors across TBIMS centers were trained in coding reasons for each reported rehospitalization into 1 of 18 level 1 HCUP diagnostic codes (expanded from 8 categories in the prior coding system), and these included infectious diseases, neoplasms, endocrine/nutritional and metabolic diseases, diseases of the blood and blood-forming organs, mental illness, diseases of the nervous system and sense organs, diseases of the circulatory system, diseases of the respiratory system, diseases of the digestive system, diseases of the genitourinary system (not including urinary tract infections), complications of pregnancy, diseases of the skin and subcutaneous tissue, diseases of the musculoskeletal system and the connective tissue, congenital anomalies, certain conditions originating in the perinatal period, injury and poisoning (including craniotomy or craniectomy), symptoms/signs of an ill-defined condition, and factors influencing health status and unclassified residual codes. Among each of the 18 level 1 diagnostic codes, data collectors further classified participants' reasons for rehospitalization into a level 2 HCUP diagnosis code.
Demographic variables and covariates
All information was collected per standard TBIMS protocols. Unlike a prior rehospitalization study17 that only used baseline predictors, we wanted to leverage the full breadth of available longitudinal predictors for odds of rehospitalization up to 10 years postinjury. To this end, we characterized the sample at 3 time points: (1) baseline (ie, acute hospitalization and inpatient rehabilitation); (2) in the wave prior to assessment of rehospitalization (except at year 1); and (3) at the same wave of assessment of rehospitalization. The variables collected at these 3 time points were the predictors for rehospitalization at 1, 2, 5, and 10 years postinjury. The included baseline variables were consistent with a TBIMS prior study of rehospitalization,17 and additional variables were selected from clinical experience regarding associations between covariates and TBI outcomes.
The baseline variables included sex, race (White and non-White), education (less than high school, high school or more), inpatient rehabilitation payer source (governmental insurance, other), length of stay during inpatient rehabilitation, Functional Independence Measure (FIM) Motor subscale score at rehabilitation discharge, FIM Cognitive subscale score at rehabilitation discharge, cranial surgery status (craniotomy or craniectomy), pre-index history of TBI (self-reported history of TBI prior to index TBI as determined by the Ohio State University TBI Identification Method).
The variables measured at prior TBIMS follow-up waves included problem of substance use (defined as whether participant in the past month has taken illicit drugs, binge drunk, drink >14 drinks per week [males] or 7 drinks per week [females]), post-index rehospitalization (rehospitalization in the prior wave), Participation Assessment with Recombined Tools—Objective (PART-O) summary (sum of 3 PART-O subscale scores (Out-and-About, Productivity, and Social) divided by 3), and driving independence (participant reports driving vehicle independently).
The variables measured at the same follow-up interview as the rehospitalization variable included age group (16-29, 30-49, 50-64, 65+ years), seizures in the past year, total number of self-reported medical comorbidities (0-1, 2+), and total number of self-reported mental health comorbidities (0-1, 2+). The list of medical conditions considered were hypertension, congestive heart failure, myocardial infarction, other heart conditions, stroke, emphysema, high blood cholesterol, diabetes, pneumonia, liver disease, rheumatoid arthritis, osteoarthritis, sleep disorder, cataracts, chronic pain, dementia, and movement disorder. The list of mental health conditions considered were alcoholism, drug addiction, depression, anxiety, panic attacks, bipolar disorder, attention-deficit disorder, attention-deficit/hyperactive disorder, obsessive-compulsive disorder, and posttraumatic stress disorder.
Data analysis
We compared demographic and clinical characteristics by rehospitalization (yes/no) at 1, 2, 5, and 10 years postinjury using chi-square tests for categorical characteristics, 2-sample t tests for normally distributed continuous characteristics, and Wilcoxon rank-sum tests for markedly skewed continuous characteristics. We calculated rates (per 1000 persons) for each reason for rehospitalization, and the top 10 reasons for rehospitalization at each follow-up time point. We ran a multivariable logistic regression model evaluating predictors associated with the likelihood of rehospitalization at 1, 2, 5, and 10 years postinjury. As a sensitivity analysis, we ran an ordinal logistic regression model evaluating predictors of the likelihood of 0, 1, or 2+ rehospitalizations. All analyses were performed in SAS 9.4 (Cary, North Carolina).
RESULTS
Description of the sample
We investigated 4 cross-sectional samples at 1 year (n = 1203), 2 years (n = 1129), 5 years (n = 1017), and 10 years (n = 894) postinjury. The descriptive characteristics associated with rehospitalization at each follow-up year are documented in Table 1. Overall, characteristics consistently associated with rehospitalization at each time point included older age group, lower FIM Motor score at inpatient rehabilitation discharge, seizures in the past year, and a greater number of medical and mental health comorbidities.
TABLE 1 -
Demographic and clinical variables by
rehospitalization at 1, 2, 5, and 10 years
Variables
n (%) |
Rehospitalization 1 y postinjury |
Rehospitalization 2 y postinjury |
Rehospitalization 5 y postinjury |
Rehospitalization 10 y postinjury |
No
922 (76.6) |
Yes
281 (23.4) |
P
... |
No
891 (78.9) |
Yes
238 (21.1) |
P
... |
No
805 (79.2) |
Yes
212 (20.9) |
P
... |
No
728 (81.4) |
Yes
166 (18.6) |
P
... |
|
Measured at baseline
|
| Sex, n (%) |
|
|
.329 |
|
|
.132 |
|
|
.884 |
|
|
.047a
|
| Female |
222 (24.2) |
75 (27.1) |
|
221 (24.9) |
70 (29.8) |
|
220 (27.3) |
59 (27.8) |
|
174 (23.9) |
52 (31.3) |
|
| Male |
696 (75.8) |
202 (72.9) |
|
665 (75.1) |
165 (70.2) |
|
585 (72.7) |
153 (72.2) |
|
554 (76.1) |
114 (68.7) |
|
| Race, n (%) |
|
|
.275 |
|
|
.273 |
|
|
.735 |
|
|
.567 |
| White |
583 (63.6) |
188 (67.1) |
|
568 (64.0) |
160 (67.8) |
|
549 (68.2) |
142 (67.0) |
|
512 (70.3) |
113 (68.1) |
|
| Non-White |
334 (36.4) |
92 (32.9) |
|
320 (36.0) |
76 (32.2) |
|
256 (31.8) |
70 (33.0) |
|
216 (29.7) |
53 (31.9) |
|
| Education, n (%) |
|
|
.682 |
|
|
.687 |
|
|
.283 |
|
|
.261 |
| Less than HS |
149 (16.5) |
43 (15.4) |
|
156 (17.7) |
44 (18.8) |
|
147 (18.4) |
32 (15.2) |
|
168 (23.1) |
45 (27.3) |
|
| HS+ |
757 (83.6) |
236 (84.6) |
|
727 (82.3) |
190 (81.2) |
|
651 (81.6) |
178 (84.8) |
|
558 (76.9) |
120 (72.7) |
|
| Rehab payer source, n (%) |
|
|
.098 |
|
|
<.001a
|
|
|
.008a
|
|
|
.034a
|
| Government insurance |
320 (34.9) |
111 (40.4) |
|
296 (33.6) |
130 (55.6) |
|
259 (32.2) |
89 (42.0) |
|
202 (27.8) |
60 (36.1) |
|
| Other |
597 (65.1) |
164 (59.6) |
|
585 (66.4) |
104 (44.4) |
|
545 (67.8) |
123 (58.0) |
|
524 (72.2) |
106 (63.9) |
|
| Length of stay during rehabilitation, mean (SD), d |
23.1 (21.3) |
29.2 (28.1) |
<.001a
|
23.2 (20.8) |
27.5 (30.9) |
.112 |
25.2 (26.4) |
30.6 (32.2) |
.002a
|
27.1 (24.3) |
26.8 (26.5) |
.290 |
| FIM Motor score at rehabilitation discharge, mean (SD) |
66.8 (16.8) |
59.7 (20.6) |
<.001a
|
67.5 (16.5) |
61.1 (18.1) |
<.001a
|
66.2 (17.0) |
60.5 (18.8) |
<.001a
|
67.7 (17.9) |
64.6 (18.6) |
.015a
|
| FIM Cognitive score at rehabilitation discharge, mean (SD) |
23.4 (6.5) |
22.5 (7.4) |
.215 |
24.0 (6.3) |
23.0 (7.1) |
.049a
|
23.8 (6.4) |
22.8 (7.4) |
.135 |
23.9 (6.4) |
23.2 (6.3) |
.170 |
| Cranial surgery status, n (%) |
221 (24.1) |
109 (39.6) |
<.001a
|
225 (22.5) |
70 (29.9) |
.177 |
218 (27.2) |
65 (30.7) |
.310 |
196 (27.0) |
39 (23.6) |
.372 |
| Pre-index history of TBI, n (%) |
197 (21.9) |
63 (22.7) |
.764 |
232 (26.3) |
47 (20.1) |
.049a
|
245 (30.6) |
74 (35.1) |
.216 |
197 (27.3) |
36 (21.8) |
.147 |
|
Measured at prior TBIMS follow-up interview
|
| Problem substance use, n (%) |
... |
... |
... |
175 (22.0) |
34 (16.1) |
.059 |
173 (23.1) |
37 (19.4) |
.270 |
164 (26.7) |
35 (24.3) |
.551 |
| Post-index rehospitalization, n (%) |
... |
... |
... |
155 (19.0) |
103 (47.9) |
<.001a
|
118 (15.3) |
72 (36.7) |
<.001a
|
93 (14.4) |
59 (39.2) |
<.001a
|
| PART-O summary, mean (SD) |
... |
... |
... |
1.7 (0.7) |
1.4 (0.7) |
<.001a
|
1.8 (0.7) |
1.6 (0.7) |
<.001a
|
1.8 (0.7) |
1.6 (0.7) |
<.001a
|
| Driving independence, n (%) |
... |
... |
... |
363 (45.8) |
57 (27.3) |
<.001a
|
392 (52.7) |
74 (39.2) |
<.001a
|
376 (59.9) |
83 (56.1) |
.399 |
|
Measured at the same follow-up interview as rehospitalization variable
|
| Age, n (%) |
|
|
.046a
|
|
|
<.001a
|
|
|
<.001a
|
|
|
<.001a
|
| 16-29 y |
271 (29.6) |
73 (26.0) |
|
281 (31.6) |
40 (16.9) |
|
195 (24.2) |
40 (18.9) |
|
103 (14.2) |
8 (4.8) |
|
| 30-49 y |
247 (26.9) |
71 (25.3) |
|
247 (27.8) |
69 (29.1) |
|
280 (35.8) |
53 (25.0) |
|
358 (49.2) |
74 (44.6) |
|
| 50-64 y |
214 (23.3) |
58 (20.6) |
|
189 (21.3) |
58 (24.5) |
|
176 (21.9) |
56 (26.4) |
|
173 (23.8) |
45 (27.1) |
|
| 65+ y |
185 (20.2) |
79 (28.1) |
|
171 (19.3) |
70 (29.5) |
|
154 (19.1) |
63 (29.7) |
|
94 (12.9) |
39 (23.5) |
|
| Seizures in the past year, n (%) |
65 (7.1) |
84 (30.1) |
<.001a
|
78 (8.9) |
54 (23.0) |
<.001a
|
55 (6.8) |
49 (23.6) |
<.001a
|
41 (5.7) |
38 (23.0) |
<.001a
|
| Total number of medical comorbidities, n (%) |
|
|
<.001a
|
|
|
<.001a
|
|
|
<.001a
|
|
|
<.001a
|
| 0-1 |
533 (58.9) |
100 (35.7) |
|
484 (55.3) |
70 (29.5) |
|
465 (58.4) |
73 (34.4) |
|
431 (60.0) |
49 (29.5) |
|
| 2+ |
372 (41.1) |
180 (64.3) |
|
391 (44.7) |
167 (70.5) |
|
331 (41.6) |
139 (65.6) |
|
287 (40.0) |
117 (70.5) |
|
| Total number of mental health comorbidities, n (%) |
|
|
.021a
|
|
|
<.001a
|
|
|
<.001a
|
|
|
.011a
|
| 0-1 |
491 (54.7) |
129 (46.7) |
|
460 (52.9) |
86 (36.9) |
|
430 (54.2) |
82 (39.1) |
|
380 (53.2) |
69 (42.1) |
|
| 2+ |
407 (45.3) |
147 (53.3) |
|
410 (47.1) |
147 (63.1) |
|
363 (45.8) |
128 (60.9) |
|
335 (46.9) |
95 (57.9) |
|
Abbreviations: FIM, Functional Independence Measure; HS, high school; rehab, rehabilitation; PART-O, Participation Assessment with Recombined Tools—Objective; TBI, traumatic brain injury; TBIMS, TBI Model Systems.
aStatistically significant at α = .05.
Rates of rehospitalization
The proportion of the sample rehospitalized at 1, 2, 5, and 10 years postinjury was 23.4%, 21.1%, 20.9%, and 18.6%, respectively.
Reasons for rehospitalization
The top 10 reasons for rehospitalization, and associated rates, at 1, 2, 5, and 10 years postinjury are displayed in Table 2. The top reason for rehospitalization at 1, 2, and 5 years postinjury was “epilepsy; convulsions,” while at year 10 the most common reason was “fractures; treatment of fractures or dislocation” (closely followed by epilepsy). “Intracranial injury; craniotomy or craniectomy procedure” was common at years 1 and 2 and then again at year 10. “Bacterial infection” was common through year 5, and “mood disorders” and “anxiety disorders” made the top 10 in years 5 and 10, respectively. Other top reasons for rehospitalization in the first 10 years postinjury were as follows: “diseases of the heart; heart procedures”; “fractures; treatment of fractures or dislocation”; “diseases of the urinary system; operations on the urinary system”; and “cerebrovascular disease/stroke.” We present the detailed rates of rehospitalization by body system (level 1 HCUP classifications) in Supplemental Digital Content Table 1 (available at: https://links.lww.com/JHTR/A597) and particular diagnoses (level 2 HCUP classifications) in Supplemental Digital Content Table 2 (available at: https://links.lww.com/JHTR/A598).
TABLE 2 -
Top 10 reasons for
rehospitalization at 1, 2, 5, and 10 years post-TBI
a
| Rank |
Year 1 |
Year 2 |
Year 5 |
Year 10 |
| Reason for rehospitalization |
Rate per 1000 persons with TBI |
Reason for rehospitalization |
Rate per 1000 persons with TBI |
Reason for rehospitalization |
Rate per 1000 persons with TBI |
Reason for rehospitalization |
Rate per 1000 persons with TBI |
| 1 |
6.4. Epilepsy; convulsions |
32.4 |
6.4. Epilepsy; convulsions |
25.7 |
6.4. Epilepsy; convulsions |
23.6 |
16.2. Fractures; treatment of fractures or dislocation |
19.0 |
| 2 |
16.4. Intracranial injury; craniotomy or craniectomy procedure |
15.0 |
16.2. Fractures; treatment of fractures or dislocation |
22.1 |
16.2. Fractures; treatment of fractures or dislocation |
16.7 |
6.4. Epilepsy; convulsions |
17.9 |
| 3 |
7.2. Diseases of the heart; heart procedures |
14.1 |
7.2. Diseases of the heart; heart procedures |
15.9 |
7.2. Diseases of the heart; heart procedures |
15.7 |
16.12. Other injuries and conditions due to external causes |
12.3 |
| 4 |
1.1. Bacterial infection |
13.3 |
1.1. Bacterial infection |
15.1 |
10.1. Diseases of the urinary system; operations on the urinary system |
11.8 |
7.2. Diseases of the heart; heart procedures |
7.8 |
| 5 |
16.2. Fractures; treatment of fractures or dislocation |
10.8 |
10.1. Diseases of the urinary system; operations on the urinary system |
15.1 |
1.1. Bacterial infection |
9.8 |
9.6. Lower gastrointestinal disorders; lower gastrointestinal procedures |
7.8 |
| 6 |
17.1. Symptoms; signs; ill-defined conditions |
10.0 |
16.4. Intracranial injury; craniotomy or craniectomy procedure |
11.5 |
8.1. Respiratory infections |
8.8 |
10.1. Diseases of the urinary system; operations on the urinary system |
7.8 |
| 7 |
10.1. Diseases of the urinary system; operations on the urinary system |
9.1 |
16.12. Other injuries and conditions due to external causes |
9.7 |
5.8. Mood disorders; electroconvulsive therapy |
6.9 |
16.1. Joint disorders and dislocations; trauma-related |
6.7 |
| 8 |
13.2. Nontraumatic joint disorders; arthroplasty; other |
9.1 |
17.1. Symptoms; signs; ill-defined conditions |
7.1 |
16.12. Other injuries and conditions due to external causes |
6.9 |
8.1. Respiratory infections |
5.6 |
| 9 |
17.2. Factors influencing healthcare (rehabilitation care, medical examination, other aftercare) |
9.1 |
7.3. Cerebrovascular disease/stroke; endarterectomy; vessel of head and neck |
6.2 |
7.3. Cerebrovascular disease/stroke; endarterectomy; vessel of head and neck |
5.9 |
16.4. Intracranial injury; craniotomy or craniectomy procedure |
5.6 |
| 10 |
7.3. Cerebrovascular disease/stroke; endarterectomy; vessel of head and neck |
7.5 |
8.1. Respiratory infections |
6.2 |
16.1. Joint disorders and dislocations; trauma-related |
5.9 |
5.2. Anxiety disorders |
4.5 |
Abbreviation: TBI, traumatic brain injury.
aFor ties in rate, reasons for rehospitalization are listed in the ascending order of codes.
Predictors of the likelihood of rehospitalization at 1, 2, 5, and 10 years postinjury
Results of the multivariable logistic regression models for each follow-up year are displayed in Table 3. At year 1, the baseline variables significantly associated with rehospitalization were FIM Motor score at rehabilitation discharge (OR = 0.98; 95% CI, 0.97-0.99) and craniotomy or craniectomy (OR = 2.29; 95% CI, 1.64-3.22). Year 1 variables measured at the same cross-sectional time point significantly associated with rehospitalization were seizures in the last year (OR = 5.10; 95% CI, 3.40-7.64) and number of medical comorbidities (OR = 2.89; 95% CI, 2.01-4.15).
TABLE 3 -
Multivariable logistic regression model for predictors of likelihood of
rehospitalization at 1, 2, 5, and 10 years post-TBI
| Variables |
Rehospitalization 1 y postinjury (n = 1113) |
Rehospitalization 2 y postinjury (n = 907) |
Rehospitalization 5 y postinjury (n = 874) |
Rehospitalization 10 y postinjury (n = 709) |
| OR (95% CI) |
P
|
OR (95% CI) |
P
|
OR (95% CI) |
P
|
OR (95% CI) |
P
|
|
Measured at baseline
|
| Sex |
| Male (ref) |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
| Female |
1.19 (0.83-1.69) |
.341 |
0.92 (0.61-1.39) |
.704 |
0.79 (0.53-1.18) |
.254 |
1.28 (0.80-2.05) |
.298 |
| Race |
| White (ref) |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
| Non-White |
0.79 (0.56-1.12) |
.183 |
0.71 (0.45-1.10) |
.125 |
1.30 (0.86-1.96) |
.220 |
0.93 (0.57-1.53) |
.776 |
| Education |
| Less than HS (ref) |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
| HS+ |
0.96 (0.62-1.50) |
.854 |
0.90 (0.54-1.50) |
.682 |
1.48 (0.88-2.49) |
.145 |
0.81 (0.47-1.39) |
.447 |
| Rehab payer source |
| Government insurance (ref) |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
| Other |
1.03 (0.73-1.45) |
.858 |
0.56 (0.38-0.83) |
.004a
|
0.90 (0.61-1.34) |
.609 |
0.94 (0.58-1.51) |
.794 |
| Length of stay during rehabilitation |
1.01 (1.00-1.01) |
.187 |
1.00 (0.99-1.01) |
.968 |
1.00 (0.99-1.01) |
.807 |
1.00 (0.99-1.01) |
.940 |
| FIM Motor score at rehabilitation discharge |
0.98 (0.97-0.99) |
<.001a
|
0.99 (0.97-0.99) |
.027a
|
0.98 (0.97-0.99) |
.009a
|
0.99 (0.98-1.01) |
.235 |
| FIM Cognitive score at rehabilitation discharge |
1.04 (1.01-1.07) |
.011a
|
1.02 (0.98-1.05) |
.366 |
1.02 (0.99-1.06) |
.149 |
0.99 (0.96-1.03) |
.578 |
| Cranial surgery status |
| No cranial surgery (ref) |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
| Craniotomy and/or craniectomy |
2.29 (1.64-3.22) |
<.001 |
0.83 (0.54-1.28) |
.400 |
0.90 (0.59-1.37) |
.616 |
0.89 (0.54-1.45) |
.635 |
| Pre-index history of TBI (yes vs no) |
0.94 (0.64-1.36) |
.726 |
0.53 (0.33-0.83) |
.006a
|
1.20 (0.81-1.76) |
.360 |
0.69 (0.42-1.12) |
.133 |
|
Measured at prior TBIMS follow-up interview
|
| Problem substance use (yes vs no) |
... |
... |
0.69 (0.41-1.17) |
.165 |
0.80 (0.50-1.28) |
.353 |
1.17 (0.70-1.97) |
.551 |
| Post-index rehospitalization (yes vs no) |
... |
... |
3.24 (2.20-4.78) |
<.001a
|
2.24 (1.49-3.38) |
.001a
|
3.25 (2.07-5.10) |
<.001a
|
| PART-O summary |
... |
... |
0.93 (0.66-1.31) |
.679 |
1.01 (0.73-1.40) |
.963 |
0.89 (0.62-1.29) |
.542 |
| Driving independence (yes vs no) |
... |
... |
0.94 (0.59-1.52) |
.814 |
0.89 (0.57-1.38) |
.596 |
1.31 (0.79-2.17) |
.301 |
|
Measured at the same follow-up interview as rehospitalization variable
|
| Age |
| 16-29 y (ref) |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
Ref |
| 30-49 y |
0.87 (0.56-1.34) |
.517 |
2.12 (1.21-3.73) |
.009a
|
0.71 (0.41-1.21) |
.203 |
2.42 (0.97-6.02) |
.057 |
| 50-64 y |
0.83 (0.52-1.32) |
.426 |
1.91 (1.04-3.51) |
.038a
|
1.30 (0.74-2.28) |
.361 |
2.57 (0.97-6.80) |
.057 |
| 65+ y |
1.01 (0.60-1.68) |
.983 |
1.79 (0.92-3.48) |
.085 |
1.59 (0.87-2.89) |
.131 |
4.34 (1.58-11.93) |
.004a
|
| Seizures in the past year (yes vs no) |
5.10 (3.40-7.64) |
<.001a
|
3.00 (1.77-5.06) |
<.001a
|
3.74 (2.22-6.31) |
<.001a
|
3.64 (1.98-6.69) |
<.001a
|
| Total number of medical comorbidities (2+ vs 1 or 0) |
2.89 (2.01-4.15) |
<.001a
|
1.69 (1.01-2.61) |
.017a
|
1.61 (1.06-2.43) |
.025a
|
2.10 (1.31-3.39) |
.002a
|
| Total number of mental health comorbidities (1+ vs 0) |
1.19 (0.86-1.64) |
.293 |
1.67 (1.13-2.47) |
.010a
|
1.57 (1.06-2.30) |
.023a
|
1.13 (0.72-1.78) |
.595 |
| Total model AUC |
0.761 |
|
0.782 |
|
0.739 |
|
0.764 |
|
Abbreviations: AUC, area under the curve; FIM, Functional Independence Measure; HS, high school; rehab, rehabilitation; PART-O, Participation Assessment with Recombined Tools—Objective; TBI, traumatic brain injury; TBIMS, TBI Model Systems.
aStatistically significant at α = .05.
At year 2, the baseline variables significantly associated with rehospitalization were nongovernmental insurance (OR = 0.56; 95% CI, 0.54-0.83), FIM Motor score at rehabilitation discharge (OR = 0.99; 95% CI, 0.97-0.99), and pre-index history of TBI (OR = 0.53; 95% CI, 0.33-0.83). The prior wave (year 1) variables associated with rehospitalization were rehospitalization at year 1 (OR = 3.24; 95% CI, 2.20-4.78). Year 2 concurrently measured variables that were significantly associated with rehospitalization were as follows: age strata 30 to 49 years (OR = 2.12; 95% CI, 1.21-3.73) and 50 to 64 years (OR = 1.91; 95% CI, 1.04-3.51) relative to age 16 to 29 years, seizures in the past year (OR = 3.00; 95% CI, 1.77-5.06), total number of medical comorbidities (OR = 1.69; 95% CI, 1.01-2.61), and number of mental health comorbidities (OR = 1.67; 95% CI, 1.13-2.47).
At year 5, the only baseline variables associated with rehospitalization was FIM Motor score at rehabilitation discharge (OR = 0.98; 95% CI, 0.97-0.99). Prior wave rehospitalization at year 2 was significantly associated with year 5 rehospitalization (OR = 2.24; 95% CI, 1.49-3.38). The variables collected at year 5 time point that were significantly associated with rehospitalization were as follows: seizures in the past year (OR = 3.74; 95% CI, 2.22-6.31) and greater number of medical (OR = 1.61; 95% CI, 1.06-2.43) and mental health comorbidities (OR = 1.57; 95% CI, 1.06-2.30).
At year 10, no baseline variables were associated with rehospitalization. Prior wave (year 5) rehospitalization was associated with higher odds of year 10 rehospitalization (OR = 3.25; 95% CI, 2.07-5.10). Year 10 variables associated with rehospitalization were as follows: age 65+ years (OR = 4.34; 95% CI, 1.58-11.93), seizures in the past year (OR = 3.64; 95% CI, 1.98-6.69), and a greater number of medical comorbidities (OR = 2.10; 95% CI, 1.31-3.39).
The model area under the curve for the 1-, 2-, 5-, and 10-year models were all greater than 0.7 (range, 0.739-0.782), indicating a moderately strong discrimination of rehospitalization based on the variables included.
The results of our sensitivity analysis that evaluated predictors of 0, 1, and 2+ rehospitalizations for each follow-up year were mostly similar to the primary binary logistic regression models (see Supplemental Digital Content Table 3, available at: https://links.lww.com/JHTR/A599).
DISCUSSION
Rehospitalization following discharge from acute inpatient rehabilitation is an indicator of long-term health and well-being that also reflects some of the healthcare system and economic impacts of TBI. Here, we expand upon prior TBIMS investigations of rehospitalization16,17 by using newly collected and highly detailed rehospitalization data from the TBIMS NDB to characterize the most common reasons for rehospitalization up to 10 years postinjury and investigated baseline, prior visit, and concurrent predictors of past-year rehospitalization at each follow-up interval.
Rates of rehospitalization decreased slightly over time, from 23.4% to 18.6% by year 10. There was some variability in the most common reasons for rehospitalization across the follow-up periods. However, seizure disorder was the single most common cause for rehospitalization at years 1, 2, and 5 and the second most common at year 10. Although it is possible that a seizure disorder may have predated (or even caused) the index TBI,19 these hospitalizations likely reflect complications of posttraumatic epilepsy (PTE), defined as recurrent late-onset seizures after TBI that affects up to 25% of individuals with moderate-severe TBI.20–22
Some of the rehospitalizations in the first year after injury likely reflect planned procedures relating to the injury itself, such as invasive neurosurgical procedures (craniotomy and craniectomy) for deferred or elective cranioplasty, complications such as hydrocephalus or infection, or shunt revisions. Similarly, treatment of fractures may reflect surgical revision or correction of injuries sustained concurrently with the head trauma.
However, the fact that these same rehospitalization reasons are so common 5 and 10 years postinjury (at which point “fractures/treatment of fractures/dislocation” is the number one reason for rehospitalization) is also noteworthy. Together with the prevalence of other injuries related to “external causes” at years 2 and 5, and “trauma-related” joint disorders and dislocations at years 5 and 10, these findings may reflect risk for reinjury among those with TBI, which has been extensively documented.23,24 Given the low likelihood of craniotomy or craniectomy relating to the index injury is taking place 10 years later, the hospitalizations for “intracranial injury” at year 10 may represent new TBIs. Furthermore, and consistent with our prior work, rates of psychiatric hospitalization increased over time.17
Cerebrovascular disease/stroke was among the top causes for rehospitalization at years 1, 2, and 5; this is consistent with prior evidence of increased risk for acute ischemic stroke following TBI25 and also the longer-term risk for ischemic and hemorrhagic stroke.26,27 The causes for this association are unknown but may, in part, be related to known or occult traumatic cervical vascular injury concurrent with their TBI. Other common causes of rehospitalization may be less closely associated with the initial injury. Bacterial infections and cardiac conditions are common reasons for hospitalization in the general population28; that said, there is now more known on cardiac manifestations in patients with TBI, including theories on the neurocardiac axis.29 Also, those with TBI are more likely to develop chronic diseases such as heart disease.30 Diseases or operations involving the urinary system were also common. Although TBI generally does not cause true neurogenic bladder as with spinal cord lesions, many patients have difficulty with higher-level urinary control leading to incontinence and/or retention31 along with co-occurring spinal cord injury. Admissions may be from complicated urinary tract infections or complications from remote histories of urinary tract infections or prolonged use of catheterization.
Our investigation into predictors of rehospitalization indicated that the most stable baseline predictor of past-year rehospitalization across follow-up visits at years 1, 2, and 5 was lower FIM Motor score at rehabilitation discharge, which replicates prior work in the TBIMS NDB that exclusively investigated baseline predictiors.17 Notably, the finding that no baseline predictors were associated with rehospitalization risk by 10 years underscores the importance of including prior-visit and concurrent time-varying covariates in models predicting remote TBI outcomes. When incorporating information collected at the prior and concurrent study visits, we found that prior rehospitalization, past-year seizure, and a greater burden of medical and/or psychiatric comorbidities were consistent predictors of rehospitalization. There was some evidence that middle-aged participants were more likely to be rehospitalized than their younger counterparts (ie, at year 2) in that older age (ie, age 65+ years) was significantly associated with an increased risk for rehospitalization at the 10-year follow-up. This finding is consistent with recent TBIMS NDB research32 that found that the negative effects of injury chronicity are most pronounced among older adults with TBI.
There are limitations to this project that warrant mention. The TBIMS NDB includes individuals who receive care at inpatient rehabilitation facilities with specialized TBI units and therefore findings may not generalize to all patients recovering from TBI. However, this limitation is mitigated by previous research demonstrating that the TBIMS NDB is representative of all patients receiving inpatient rehabilitation for TBI in the United States.33 The present study used a recently introduced rehospitalization variable from 2017, and with our current sample, we were limited in our ability to examine subgroups of participants who may be differentially impacted by TBI and access to care including payer source34 (eg, workers' compensation) and race/ethnicity35 (eg, Hispanic). Future studies with larger samples should use more granular categorizations of predictors of rehospitalization to facilitate more meaningful interpretations. Information on rehospitalization was collected via self-report or proxy report and therefore may be subject to underreporting or even overreporting. In addition, there may be selection bias based on those having more medical problems and hospitalizations being more prone to complete their follow-up visits.36 We also do not have data in the TBIMS on whether rehospitalizations were planned or unplanned. In addition, coding reasons for rehospitalization were done by the study staff based on structured protocols but there could be some miscategorization despite rigorous training and oversight. Future research would benefit from studying population cohorts that include individuals with and without TBI to better understand the specific medical diagnoses more prevalent among people with TBI. Also, we advise more research that uses objective healthcare claims data containing supplemental information from that available in self-report interviews, such as costs, duration of subsequent hospital stays, and surgical and medical interventions received. Important strengths of the current study, which extends findings from previously published articles on rehospitalization following moderate-severe TBI,17,37,38 include long length of follow-up, detailed categorization of reasons for rehospitalization, and inclusion of proximal (prior visit) and concurrent (same visit) predictors in all models rather than relying exclusively on baseline predictors of long-term rehospitalization risk.
Taken together, these findings reinforce the need for ongoing care and surveillance, at least among a subset of long-term survivors of moderate-severe TBI; these needs are not limited to older adults. Rehospitalizations may result from medical or mental health deterioration, undertreatment, ineffective treatment, and/or nonadherence to prescribed medications and care regimens. Current findings reiterate the need for regular and ongoing surveillance and long-term care, including general and specialized care to manage seizures and other late complications of TBI. Thoughtfully deployed preventive measures can reduce avoidable rehospitalizations relating to seizure, new injuries, infection, and stroke. Current PTE practice strategies, which include post–acute antiepileptic medication only if there is a clinically expressed seizure after the first 24 hours.39,40 New cases of PTE may develop years after injury, suggesting careful screening at long-term follow-up may be warranted, particularly those known to be at greatest risk.41 Similarly, those at an elevated risk for reinjury23,24,42 may require closer monitoring and physical therapy if new balance or motor impairments emerge.43,44
Although empirical research informing specific care guidelines for those with chronic TBI is currently lacking, the results reported herein reinforce the need for and may help inform the content of a chronic disease management model designed to improve long-term health for individuals with TBI. Understanding the risk factors associated with rehospitalization in a patient recovering from TBI can inform clinical decision-making and screening/surveillance practices to improve medical and functional outcomes and decrease the healthcare burden caused by rehospitalization.
CONCLUSION
Severe and moderate TBI is gaining acceptance as a chronic health condition, and there is an increased appreciation of the increased risk of potential medical and surgical complications years after a TBI. Thus, it is important to inform clinicians, families, and other stakeholders which patients might be the most vulnerable and at an increased risk of developing such medical complications requiring rehospitalization. Our findings in the present study can inform improvements in the chronic care of patients recovering from moderate-severe TBI, including medical care, surveillance, prevention, lifestyle, and healthy behaviors to potentially reduce rehospitalization over the life span after the TBI.
REFERENCES
1. Corrigan JD, Hammond FM.
Traumatic brain injury as a
chronic health condition. Arch Phys Med Rehabil. 2013;94(6):1199–1201. doi:10.1016/j.apmr.2013.01.023
2. Masel BE, DeWitt DS.
Traumatic brain injury: a disease process, not an event. J Neurotrauma. 2010;27(8):1529–1540. doi:10.1089/neu.2010.1358
3. Hammond FM, Corrigan JD, Ketchum JM, et al. Prevalence of medical and psychiatric comorbidities following
traumatic brain injury. J Head Trauma Rehabil. 2019;34(4):E1–E10. doi:10.1097/HTR.0000000000000465
4. Izzy S, Chen PM, Tahir Z, et al. Association of
traumatic brain injury with the risk of developing chronic cardiovascular, endocrine, neurological, and psychiatric disorders. JAMA Netw Open. 2022;5(4):e229478. doi:10.1001/jamanetworkopen.2022.9478
5. Kumar RG, Juengst SB, Wang Z, Dams-O'Connor K, Dikmen SS, O'Neil-Pirozzi TM. Epidemiology of comorbid conditions among adults 50 years and older with
traumatic brain injury. J Head Trauma Rehabil. 2018;33(1):15–24. doi:10.1097/HTR.0000000000000273
6. Kumar RG, Ketchum JM, Hammond FM, et al. Health and cognition among adults with and without
traumatic brain injury: a matched case-control study. Brain Inj. 2022;36(3):415–423. doi:10.1080/02699052.2022.2034190
7. Malec JF, Ketchum JM, Hammond FM, et al. Longitudinal effects of medical comorbidities on functional outcome and life satisfaction after
traumatic brain injury: an individual growth curve analysis of NIDILRR
Traumatic Brain Injury Model System data. J Head Trauma Rehabil. 2019;34(5):E24–E35. doi:10.1097/HTR.0000000000000459
8. Kumar RG, Olsen J, Juengst SB, et al. Comorbid conditions among adults 50 years and older with
traumatic brain injury: examining associations with demographics, healthcare utilization, institutionalization, and 1-year outcomes. J Head Trauma Rehabil. 2019;34(4):224–232. doi:10.1097/HTR.0000000000000470
9. Hammond FM, Horn SD, Smout RJ, et al.
Rehospitalization during 9 months after inpatient rehabilitation for
traumatic brain injury. Arch Phys Med Rehabil. 2015;96(8)(suppl):S330–S339.e334. doi:10.1016/j.apmr.2014.09.041
10. Lamm AG, Goldstein R, Giacino JT, Niewczyk P, Schneider JC, Zafonte R. Changes in patient demographics and outcomes in the inpatient rehabilitation facility
traumatic brain injury population from 2002 to 2016: implications for patient care and clinical trials. J Neurotrauma. 2019;36(17):2513–2520. doi:10.1089/neu.2018.6014
11. Hammond FM, Baker-Sparr CA, Dahdah MN, et al. Predictors of 1-year global outcomes after
traumatic brain injury among older adults: a NIDILRR
Traumatic Brain Injury Model Systems Study. J Aging Health. 2019;31(10)(suppl):68S–96S. doi:10.1177/0898264318819197
12. Erler KS, Juengst SB, Whiteneck GG, et al. The association of
rehospitalization with participation 5 years after
traumatic brain injury. J Head Trauma Rehabil. 2018;33(6):E77–E84. doi:10.1097/HTR.0000000000000386
13. Saverino C, Swaine B, Jaglal S, et al.
Rehospitalization after
traumatic brain injury: a population-based study. Arch Phys Med Rehabil. 2016;97(2)(suppl):S19–S25. doi:10.1016/j.apmr.2015.04.016
14. Nakase-Richardson R, Tran J, Cifu D, et al. Do
rehospitalization rates differ among injury severity levels in the NIDRR
Traumatic Brain Injury Model Systems program? Arch Phys Med Rehabil. 2013;94(10):1884–1890. doi:10.1016/j.apmr.2012.11.054
15. Kelly ML, Shammassian B, Roach MJ, Thomas C, Wagner AK. Craniectomy and craniotomy in
traumatic brain injury: a propensity-matched analysis of long-term functional and quality of life outcomes. World Neurosurg. 2018;118:e974–e981. doi:10.1016/j.wneu.2018.07.124
16. Tran J, Hammond F, Dams-O'Connor K, et al.
Rehospitalization in the first year following veteran and service member TBI: A VA TBI Model Systems study. J Head Trauma Rehabil. 2017;32(4):264–270. doi:10.1097/HTR.0000000000000296
17. Dams-O'Connor K, Mellick D, Dreer LE, et al.
Rehospitalization over 10 years among survivors of TBI: a National Institute on Disability, Independent Living, and Rehabilitation Research
Traumatic Brain Injury Model Systems study. J Head Trauma Rehabil. 2017;32(3):147–157. doi:10.1097/HTR.0000000000000263
18. Dijkers MP, Marwitz JH, Harrison-Felix C. Thirty years of National Institute on Disability, Independent Living, and Rehabilitation Research
Traumatic Brain Injury Model Systems Center Research—an update. J Head Trauma Rehabil. 2018;33(6):363–374. doi:10.1097/HTR.0000000000000454
19. Zwimpfer TJ, Brown J, Sullivan I, Moulton RJ. Head injuries due to falls caused by seizures: a group at high risk for traumatic intracranial hematomas. J Neurosurg. 1997;86(3):433–437. doi:10.3171/jns.1997.86.3.0433
20. Englander J, Bushnik T, Duong TT, et al. Analyzing risk factors for late posttraumatic seizures: a prospective, multicenter investigation. Arch Phys Med Rehabil. 2003;84(3):365–373. doi:10.1053/apmr.2003.50022
21. Ferguson PL, Smith GM, Wannamaker BB, Thurman DJ, Pickelsimer EE, Selassie AW. A population-based study of risk of epilepsy after hospitalization for
traumatic brain injury. Epilepsia. 2010;51(5):891–898. doi:10.1111/j.1528-1167.2009.02384.x
22. Ritter AC, Wagner AK, Fabio A, et al. Incidence and risk factors of posttraumatic seizures following
traumatic brain injury: a
Traumatic Brain Injury Model Systems study. Epilepsia. 2016;57(12):1968–1977. doi:10.1111/epi.13582
23. Bannon SM, Kumar RG, Bogner J, et al. Reinjury after moderate to severe TBI: rates and risk factors in the NIDILRR
Traumatic Brain Injury Model Systems. J Head Trauma Rehabil. 2021;36(1):E50–E60. doi:10.1097/HTR.0000000000000586
24. Dams-O'Connor K, Gibbons LE, Bowen JD, McCurry SM, Larson EB, Crane PK. Risk for late-life re-injury, dementia and death among individuals with
traumatic brain injury: a population-based study. J Neurol Neurosurg Psychiatry. 2013;84(2):177–182. doi:10.1136/jnnp-2012-303938
25. Kowalski RG, Haarbauer-Krupa JK, Bell JM, et al. Acute ischemic stroke after moderate to severe
traumatic brain injury: incidence and impact on outcome. Stroke. 2017;48(7):1802–1809. doi:10.1161/STROKEAHA.117.017327
26. Chen GS, Liao KH, Bien MY, Peng GS, Wang JY. Increased risk of posttrauma stroke after
traumatic brain injury-induced acute respiratory distress syndrome. J Neurotrauma. 2016;33(13):1263–1269. doi:10.1089/neu.2015.4063
27. Liao CC, Chou YC, Yeh CC, Hu CJ, Chiu WT, Chen TL. Stroke risk and outcomes in patients with
traumatic brain injury: 2 nationwide studies. Mayo Clinic Proc. 2014;89(2):163–172. doi:10.1016/j.mayocp.2013.09.019
28. Agency for Healthcare Research and Quality. HCUP Fast Stats. Healthcare Cost and Utilization Project (HCUP). Updated April 21, 2021. Accessed May 30, 2021.
https://hcup-us.ahrq.gov/faststats/NationalDiagnosesServlet
29. El-Menyar A, Goyal A, Latifi R, Al-Thani H, Frishman W. Brain-heart interactions in
traumatic brain injury. Cardiol Rev. 2017;25(6):279–288. doi:10.1097/CRD.0000000000000167
30. Selassie AW, Cao Y, Church EC, Saunders LL, Krause J. Accelerated death rate in population-based cohort of persons with
traumatic brain injury. J Head Trauma Rehabil. 2014;29(3):E8–E19. doi:10.1097/HTR.0b013e3182976ad3
31. Kadow BT, Tyagi P, Chermansky CJ. Neurogenic causes of detrusor underactivity. Curr Bladder Dysfunct Rep. 2015;10(4):325–331. doi:10.1007/s11884-015-0331-6
32. Rabinowitz AR, Kumar RG, Sima A, et al. Aging with
traumatic brain injury: deleterious effects of injury chronicity are most pronounced in later life. J Neurotrauma. 2021;38(19):2706–2713. doi:10.1089/neu.2021.0038
33. Corrigan JD, Cuthbert JP, Whiteneck GG, et al. Representativeness of the
Traumatic Brain Injury Model Systems National Database. J Head Trauma Rehabil. 2012;27(6):391–403. doi:10.1097/HTR.0b013e3182238cdd
34. Lequerica AH, Sander AM, Pappadis MR, et al. The association between payer source and
traumatic brain injury rehabilitation outcomes: a TBI Model Systems study. J Head Trauma Rehabil. Published online April 21, 2022. doi:10.1097/HTR.0000000000000781
35. Lequerica AH, Botticello A, O'Neill J, et al. Relationship between Hispanic nativity, residential environment, and productive activity among individuals with
traumatic brain injury: a TBI Model Systems study. J Head Trauma Rehabil. 2019;34(1):E46–E54. doi:10.1097/HTR.0000000000000398
36. Krellman JW, Kolakowsky-Hayner SA, Spielman L, et al. Predictors of follow-up completeness in longitudinal research on
traumatic brain injury: findings from the National Institute on Disability and Rehabilitation Research
Traumatic Brain Injury Model Systems program. Arch Phys Med Rehabil. 2014;95(4):633–641. doi:10.1016/j.apmr.2013.10.016
37. Moore L, Stelfox HT, Turgeon AF, et al. Rates, patterns, and determinants of unplanned readmission after traumatic injury: a multicenter cohort study. Ann Surg. 2014;259(2):374–380. doi:10.1097/SLA.0b013e31828b0fae
38. Piccenna L, Lannin NA, Gruen R, Pattuwage L, Bragge P. The experience of discharge for patients with an acquired brain injury from the inpatient to the community setting: a qualitative review. Brain Inj. 2016;30(3):241–251. doi:10.3109/02699052.2015.1113569
39. Temkin NR, Dikmen SS, Wilensky AJ, Keihm J, Chabal S, Winn HR. A randomized, double-blind study of phenytoin for the prevention of posttraumatic seizures. N Engl J Med. 1990;323(8):497–502. doi:10.1056/NEJM199008233230801
40. Haltiner AM, Temkin NR, Dikmen SS. Risk of seizure recurrence after the first late posttraumatic seizure. Arch Phys Med Rehabil. 1997;78(8):835–840. doi:10.1016/s0003-9993(97)90196-9
41. Temkin NR. Risk factors for posttraumatic seizures in adults. Epilepsia. 2003;44(s10):18–20. doi:10.1046/j.1528-1157.44.s10.6.x
42. Albrecht JS, Al Kibria GM, Greene CR, Dischinger P, Ryb GE. Postdischarge mortality of older adults with
traumatic brain injury or other trauma. J Am Geriatr Soc. 2019;67(11):2382–2386. doi:10.1111/jgs.16098
43. Gardner RC, Dams-O'Connor K, Morrissey MR, Manley GT. Geriatric
traumatic brain injury: epidemiology, outcomes, knowledge gaps, and future directions. J Neurotrauma. 2018;35(7):889–906. doi:10.1089/neu.2017.5371
44. Kornblith ES, Langa KM, Yaffe K, Gardner RC. Physical and functional impairment among older adults with a history of
traumatic brain injury. J Head Trauma Rehabil. 2020;35(4):E320–E329. doi:10.1097/HTR.0000000000000552
