Abstract
Objective
To examine associations of patient characteristics and treatment quantity delivered during inpatient spinal cord injury (SCI) rehabilitation with outcomes at 5 years post-injury and compare them to the associations found at 1 year post-injury.
Design
Observational study using Practice-Based Evidence research methodology in which clinicians documented treatment details. Regression modeling was used to predict outcomes.
Setting
Five inpatient SCI rehabilitation centers in the US.
Participants
Participants were 792 SCIRehab participants who were >12 years of age, gave informed consent, and completed both a 1-year and 5-year post-injury interview.
Outcome Measures
Outcome data were derived from Spinal Cord Injury Model Systems (SCIMS) follow-up interviews at 5 years post-injury and, similar to the 1-year SCIMS outcomes, included measures of physical independence, societal participation, life satisfaction, and depressive symptoms, as well as place of residence, school/work attendance, rehospitalization, and presence of pressure ulcers.
Results
Consistent with 1-year findings, patient characteristics continue to be strong predictors of outcomes 5-years post-injury, although several variables add to the prediction of some of the outcomes. More time in physical therapy and therapeutic recreation were positive predictors of 1-year outcomes, which held less true at 5 years. Greater time spent with psychology and social work/case management predicted greater depressive symptomatology 5-years post-injury. Greater clinician experience was a predictor at both 1- and 5 -years, although the related positive outcomes varied across years.
Conclusion
Various outcomes 5-years post-injury were primarily explained by pre-and post-injury characteristics, with little additional variance offered by the quantity of treatment received during inpatient rehabilitation.
Keywords: Spinal cord injury, Rehabilitation, Spinal cord injury model systems, Practice-based evidence
Introduction
This study was an extension of the SCIRehab study, which developed treatment taxonomies for the clinical disciplines involved in SCI rehabilitation care provision.1–8 These taxonomies were used to collect extensive prospective data, including 462,455 rehabilitation interventions, provided by >1,000 clinicians across multiple disciplines, during almost 300,000 sessions for 1,376 individuals who received inpatient rehabilitation at one of six centers from 2007 to 2009. SCIRehab outcomes include functional independence, societal participation, life satisfaction, depression symptoms, place of residence, work/school status, rehospitalization, and occurrence of pressure ulcers.
Patient characteristics and quantity of treatment received during inpatient SCI rehabilitation have been previously compared with SCIRehab outcomes at 1-year post-injury.9–17 Initial 1-year SCIRehab8 findings indicated that participant characteristics were strong predictors of outcomes with treatment duration adding slightly more predictive power.9 Specifically, more time in physical therapy was positively associated with motor FIM™, CHART Physical Independence, Social Integration, and Mobility dimensions, and less likelihood of rehospitalization and pressure ulcers.10 More time in therapeutic recreation had similar positive associations,11 while time spent in other disciplines, such as occupational therapy, speech therapy, psychology, nursing, and social work/case management, had negative relationships with the outcomes.9,12–14,16,17
Few longitudinal studies have examined relationships between patient characteristics, rehabilitation interventions, and outcomes after SCI. However, one outcome that has received significant attention is depression. Prior research has shown that depression symptoms remain relatively stable from 1- to 5-years post-injury, with demographic factors including race, sex, and age being significantly associated with depression symptoms over time.18,19 Longitudinal studies of life satisfaction show demographic factors such as race, sex, marital status, and education and injury-related factors, including neurologic level, age at injury, and functional independence, predict life satisfaction over time.20,21 Pressure ulcers are one of the most common secondary health complications after SCI. In a systematic review, Gélis and colleagues22 found several risk factors for developing pressure ulcers among individuals with chronic SCI, including sociodemographic, medical, and injury characteristics.
With respect to the SCIRehab study, 5 year follow-up data has been used to examine outcomes associated with the intensity of therapeutic recreation during inpatient rehabilitation at 1- and 5-years post-injury.23 More time spent in therapeutic recreation during inpatient rehabilitation was strongly associated with improved community participation, health and function, and recreational participation for individuals with SCI at 1-and 5-years post-injury.23
The purpose of this study was to evaluate associations found between patient and injury characteristics, inpatient rehabilitation treatment, and outcomes at 5-years post-injury for a cohort of SCIRehab patients who were able to be contacted at both 1- and 5-years post-injury. It was hypothesized that the relationships between the quantity of treatment and outcomes found at 1-year post-injury would be maintained at 5-years post-injury. If the same interventions were significantly associated with positive outcomes at both time points, efforts to increase those interventions would likely be associated with positive longer-term outcomes. On the other hand, if fewer or different interventions were significantly associated with 5-year outcomes compared to 1-year outcomes, the potential implications for improving SCI rehabilitation by modifying the dosage of some treatments would be more complex.
Materials and methods
SCIRehab was an observational study using Practice-Based Evidence (PBE) research methodology and did not manipulate treatments or alter routine clinical care.15,24 Instead, it collected data on the process of specialized SCI rehabilitation, relating details of the rehabilitation process to outcomes after controlling for individual injury (FIM at rehabilitation admission,25 injury etiology, days from injury to rehabilitation, Comprehensive Severity Index (CSI®)26) and demographic characteristics (age, sex, race/ethnicity, education, employment status, primary language, payer). The goal was to associate components of routine care with outcomes. The methods and initial results of SCIRehab have been extensively published.9–17
Facilities
The Rocky Mountain Regional Spinal Injury System at Craig Hospital, Englewood, CO led the SCIRehab project. Collaborating SCIRehab facilities that collected 5-year follow-up data included: The Mount Sinai Medical Center, New York, NY; Rehabilitation Institute of Chicago (now Shirley Ryan Ability Lab), Chicago, IL; Shepherd Center, Atlanta, GA; and Carolinas Rehabilitation, Charlotte, NC. There was variation among and across centers (e.g. delivery of care; participants’ clinical and demographic characteristics), all of which may affect outcomes. Each facility obtained Institutional Review Board approval.
Enrollment criteria
Individuals were >12 years of age, gave informed consent, and were admitted to the facilities’ SCI units for initial rehabilitation following traumatic injury. Participants took part in interviews at 1- and 5-years post-injury.
Patient demographic and injury data
Patient demographic and injury data were abstracted from medical records. Neurologic level and completeness of injury were assessed using the International Standards of Neurological Classification of SCI (ISNCSCI) and the American Spinal Injury Association Impairment Scale (AIS).27,28 The FIM™ was used to describe motor and cognitive functional independence at admission to inpatient rehabilitation.25,29 Other injury characteristics included etiology, ventilator use at rehabilitation admission, number of days from SCI to rehabilitation admission, and whether the injury was work-related. Medical severity was assessed with the Comprehensive Severity Index (CSI®),26 which quantifies severity of illness, based on over 2,100 physical findings related to a patient's disease(s). Additional patient characteristics included age at rehabilitation admission, sex, marital status, race, employment status at injury, primary payer, primary language, and body mass index (BMI); categorized as obese (BMI ≥ 30) and not obese (BMI < 30).
Treatment data
Clinical experience index
Clinicians provided information about their education and the number of years worked in SCI rehabilitation. Average clinician experience was computed by weighting the clinician's years of experience by the number of hours of treatment provided.
Time in treatment
Each clinician in each discipline – physical therapy, occupational therapy, speech therapy, psychology, nursing, therapeutic recreation, and social work/case management – used a handheld personal digital assistant to enter data about each treatment delivered. Documentation was based on a taxonomy of each discipline's most important interventions. Clinicians reported duration and other details about the interventions provided. Dosages of discipline-specific interventions were quantified using the number of minutes (in five-minute increments). The training and reliability monitoring used throughout the data collection process has been described previously.30,31 The total number of hours spent with each discipline during rehabilitation was approximated by combining activity minutes.
Outcome data
Data were collected via participant or proxy telephone or in-person interview at 1- and 5-years post-injury. Outcomes reported in this paper include:
Rasch transformed FIM™ motor score.
Societal participation was assessed using the Craig Handicap Assessment and Reporting Technique Short Form (CHART-SF).32,33 Four CHART-SF dimensions were used: Physical Independence, Social Integration, Occupation, and Mobility. Scores on each dimension range from 0 to 100, with 100 indicating performance at a level found in the general population.
Global life satisfaction was measured using the Diener Satisfaction with Life Scale (SWLS).34 Scores range from 7 to 35, with higher scores indicating greater satisfaction. Proxy responses were not allowed on this measure.
Depressive symptoms were assessed using the nine-item Patient Health Questionnaire (PHQ-9).35 Scores range from 0 to 27 with higher scores indicating greater symptomatology; proxy responses were not allowed on this measure.
Place of residence at the time of the anniversary of injury was coded in the same manner as discharge location (i.e. private residence, hospital, nursing home, group living, correctional institute, hotel/motel, deceased, other, homeless, assisted living, and unknown).
Work/school attendance status was assessed using the CHART-SF work and school items. The items were dichotomized to reflect working or being in school versus not.
Rehospitalization from rehabilitation discharge to the 1- and 5-year interview was dichotomized as none vs. one or more hospitalizations.
- The presence of pressure ulcers was dichotomized as pressure ulcer present vs. not present in the 12 months prior to the 1- and 5-year interview.
- o Of note, the duration of time for rehospitalization and pressure ulcer development between rehabilitation discharge and 1-year follow up is shorter than the 365-day interval for 5-year follow-up.
Data processing and analysis
The statistical analyses of this study replicated those conducted for the SCIRehab 1-year outcomes.9 Patient group definition, therapy quantity, and Rasch-converted FIM scores were also defined in the same manner as the SCIRehab 1-year outcomes paper. SPSS version 21 was used to calculate the frequencies and percentages of categorical variables and means and standard deviations for continuous variables. Chi-square and ANOVA tests were conducted to examine differences across the four neurological groups and to examine differences between study participants and non-participants.
Least squares stepwise regression models were used to determine which treatment variables were significantly associated with outcomes after controlling for participant demographic, injury, and additional characteristics listed in Table 1. Regression models were built using SAS version 9.4, with a two-sided P value less than 0.05 considered statistically significant. For each outcome, model selection was performed via linear (for continuous measures) or logistic (for categorical measures) stepwise regression in three distinct blocks. In the first block, potential covariates included participant demographic and injury characteristics shown in Table 1. Those determined to be significant were included in the second block, along with all treatment variables displayed in Table 2. The third block included the rehabilitation center as the main predictor and all significant covariates from the second block.
Table 1.
Participant and injury characteristics by injury group (N = 792).
| Neurologic Injury Group | Difference between Participants and Non-Participants | |||||
|---|---|---|---|---|---|---|
| Characteristic | C1–4 AIS A,B,C (n = 212) |
C5–6 AIS A,B,C (n = 165) |
Paraplegia AIS A,B,C (n = 302) |
AIS D (n = 113) |
Total Analytic Sample (N = 792) |
P |
| Admission motor FIM™, Rasch-transformed, mean (SD) | 6.9 (10.5) | 18.1 (12.0) | 35.9 (6.7) | 29.4 (14.7) | 23.5 (15.8) | 0.17 |
| Admission cognitive FIM™, Rasch-transformed, mean (SD) | 62.4 (18.9) | 66.6 (17.5) | 70.3 (18.8) | 73.9 (19.9) | 68.0 (19.1) | 0.16 |
| Comprehensive Severity Index, mean (SD) | 51.9 (37.4) | 40.1 (28.8) | 32.7 (23.0) | 21.8 (19.6) | 37.8 (30.0) | *0.01 |
| Days from injury to rehabilitation, mean (SD) | 35.5 (25.4) | 32.6 (25.3) | 28.6 (24.5) | 17.9 (13.2) | 29.7 (24.3) | 0.16 |
| Traumatic etiology, count (%) | *0.05 | |||||
| Vehicular | 106 (50.0) | 83 (50.3) | 171 (56.6) | 56 (49.6) | 416 (52.5) | |
| Violence | 13 (6.1) | 14 (8.5) | 36 (11.9) | 2 (1.8) | 65 (8.2) | |
| Sports | 38 (17.9) | 42 (25.5) | 10 (3.3) | 12 (10.6) | 102 (12.9) | |
| Fall or falling object | 50 (23.6) | 25 (15.2) | 69 (22.8) | 33 (29.2) | 177 (22.3) | |
| Medical/Surgery/Other | 5 (2.4) | 1 (0.6) | 16 (5.3) | 10 (8.8) | 32 (4.0) | |
| Age at injury (years), mean (SD) | 38.6 (16.3) | 32.1 (14.5) | 33.1 (13.4) | 47.9 (16.9) | 36.5 (15.8) | *0.01 |
| Sex, count (%) Female | 38 (17.9) | 29 (17.6) | 61 (20.2) | 23 (20.4) | 151 (19.1) | 0.83 |
| Marital status = Not married, count (%) | 120 (56.6) | 114 (69.1) | 186 (61.6) | 62 (54.9) | 482 (60.9) | 0.35 |
| Race/ethnicity, count (%) | ||||||
| White | 166 (78.3) | 136 (82.4) | 217 (71.9) | 87 (77.0) | 606 (76.5) | *< 0.01 |
| Black | 33 (15.6) | 22 (13.9) | 63 (20.9) | 16 (14.2) | 135 (17.0) | |
| Hispanic | 5 (2.4) | 3 (1.8) | 8 (2.6) | 3 (2.7) | 19 (2.4) | |
| Other | 8 (3.8) | 3 (1.8) | 14 (4.6) | 7 (6.2) | 32 (4.0) | |
| Employment status before injury, count (%) | *0.01 | |||||
| Working | 141 (66.5) | 112 (67.9) | 211 (69.9) | 72 (63.7) | 536 (67.7) | |
| Student | 33 (15.6) | 38 (23.0) | 47 (15.6) | 8 (7.1) | 126 (15.9) | |
| Retired | 17 (8.0) | 1 (0.6) | 8 (2.6) | 19 (16.8) | 45 (5.7) | |
| Unemployed/other | 21 (9.9) | 14 (8.5) | 36 (11.9) | 14 (12.4) | 85 (10.7) | |
| Injury work related, count(%) No | 186 (87.7) | 146 (88.8) | 249 (82.5) | 98 (86.7) | 679 (85.7) | 0.39 |
| Body Mass Index at admission, count (%) < 30 | 176 (83.0) | 147 (89.1) | 240 (79.5) | 89 (78.8) | 652 (82.3) | 0.16 |
| Primary language, count (%) English primary language√ | 203 (95.8) | 157 (95.2) | 289 (95.7) | 109 (96.5) | 758 (95.7) | *0.01 |
| Payer, count(%) | ||||||
| Medicare | 8 (3.8) | 3 (1.8) | 13 (4.3) | 19 (16.8) | 43 (5.4) | *< 0.01 |
| Medicaid | 29 (13.7) | 31 (18.8) | 63 (20.9) | 9 (8.0) | 132 (16.7) | |
| Private insurance/pay | 154 (72.6) | 115 (69.7) | 195 (64.6) | 71 (62.8) | 535 (67.6) | |
| Worker's compensation | 21 (9.9) | 16 (9.7) | 31 (10.3) | 14 (12.4) | 82 (10.4) | |
| Education, count (%) | *0.04 | |||||
| < High school diploma | 32 (15.1) | 44 (26.7) | 63 (20.9) | 18 (15.9) | 157 (19.8) | |
| High school diploma or GED | 122 (57.5) | 69 (41.8) | 166 (55.0) | 57 (50.4) | 414 (52.3) | |
| > High school diploma | 58 (27.4) | 51 (30.9) | 72 (23.8) | 35 (31.0) | 216 (27.3) | |
| Other/unknown | 0 (0.0) | 1 (0.6) | 1 (0.3) | 3 (2.7) | 5 (0.6) | |
*Statistically significant differences between participants and non-participants: P < 0.05.
Table 2.
Treatment variables (mean and SD) by injury group and full sample (n = 792).
| Neurologic Injury Group | Difference between Participants and Non-Participants | |||||
|---|---|---|---|---|---|---|
| Characteristic | C1–4 AIS A,B,C (n = 212) |
C5–8 AIS A,B,C (n = 165) |
Paraplegia AIS A,B,C n = (302) |
AIS D (n = 113) |
Total Analytic Sample (N = 792) |
P |
| Length of rehabilitation stay (days) | 73.6 (43.0) | 66.7 (38.2) | 44.3 (22.6) | 36.5 (22.7) | 55.7 (35.7) | 0.79 |
| Clinician experience index | 7.2 (3.4) | 6.8 (3.0) | 6.6 (3.5) | 6.0 (3.8) | 6.7 (3.5) | *<0.01 |
| Occupational therapy total hours | 71.6 (38.5) | 74.0 (39.1) | 36.9 (18.4) | 37.9 (28.2) | 54.0 (35.6) | 0.47 |
| Psychology total hours | 15.3 (11.8) | 13.4 (12.1) | 9.2 (6.5) | 6.6 (6.0) | 11.3 (9.9) | 0.10 |
| Physical therapy total hours | 68.0 (34.6) | 68.3 (35.3) | 51.7 (34.5) | 43.1 (27.8) | 58.3 (35.2) | 0.02 |
| Registered nursing total hours | 45.7 (25.3) | 40.4 (21.1) | 29.8 (15.9) | 22.4 (16.1) | 35.2 (21.6) | 0.64 |
| Speech language pathology total hours | 7.7 (16.1) | 3.2 (6.8) | 2.5 (6.5) | 2.8 (7.1) | 4.1 (10.3) | 0.11 |
| Social work/case management total hours | 12.8 (13.1) | 10.6 (10.5) | 7.1 (7.4) | 5.2 (5.2) | 9.1 (10.1) | 0.73 |
| Therapeutic recreation total hours | 22.1 (15.7) | 28.4 (21.0) | 18.1 (14.3) | 9.7 (10.2) | 20.1 (16.8) | *<0.01 |
*Statistically significant differences between participants and non-participants: P < 0.05.
Adjusted R2 was used in the linear regressions and Maximum Re-scaled R2 (Max R2) was used in the logistic regressions to report the strength of the models.36 The former indicates the percentage of variance in the dependent variable explained collectively by the independent variables. Additionally, for the logistic regressions, the area under the Receiver Operator Curve (c) was used to assess how well the model discriminated between individuals who did and who did not achieve an outcome. In both cases, values closer to 1.0 indicate better models. The tables report adjusted R2, Max R2, and c for all three iterations of the model selection process to facilitate side-by-side comparison of model fit for each outcome. The tables also display parameter estimates with their confidence intervals and P values for the participant characteristics and treatment variables in the final model. For linear regression, the tables display the regression coefficients, and for logistic regression, the tables show the odds ratio estimate.
Results
Participants
Participants (N = 792) were discharged from one of the five SCIRehab centers from fall 2007 through winter 2009 and took part in both 1- and 5-year post-injury interviews. These participants, who were >12 years of age and gave informed consent, were 61% of the original SCIRehab sample. One SCIRehab center could not provide 5-year data and thus, was not included in this study (see Fig. 1). Demographic and injury characteristics and treatment variables are shown by injury group in Tables 1 and 2. The number of participant responses varied from 506 to 786 across the outcome variables.
Figure 1.
Participant flow diagram.
Differences in demographic and injury characteristics between participants and non-participants revealed some differences. The significance levels of the differences between study participants and non-participants are indicated in Tables 1 and 2. Significant differences are reported below in text, with mean differences for continuous variables and standard residuals for categorical variables.
For continuous variables, participants versus non-participants had lower CSI scores (mean difference = 4.77), were younger at the time of injury (mean difference = 2.63 years), received more hours of physical therapy (mean difference = −4.59) and therapeutic recreation (mean difference = −5.41) intervention, and their clinicians had higher clinician experience index scores (mean difference = −1.04). For categorical variables, participants versus non-participant had a lower likelihood of acquiring their injury from violence (standard residual = 2.40), being Black (standard residual = 2.9), being retired (standard residual = 2.3), not speaking English as their primary language (standard residual = 1.90), having Medicare as their primary payer (standard residual = 2.9), and reporting other/unknown education (standard residual = 5.30).
Association of outcomes with patient and treatment variables
Tables 3–13 display only the significant predictors of each outcome at both 1- and 5-years post-injury. Complete tables for 1- and 5-year results for all outcomes are provided as supplemental materials. Table 14 summarizes all of the predictors of Year 1 and Year 5 outcomes by variable type.
Table 3.
Predictors of FIM™ Motor at 1- and 5-years post-injury.
| Outcome: | 1-year FIM™ Motor | 5-year FIM™ Motor | ||||
|---|---|---|---|---|---|---|
| # Observations used | 747 | 506 | ||||
| Step 1: Pt characteristics: adjusted R2 | 0.53 | 0.53 | ||||
| Step 2: Pt characteristics + treatments: adjusted R2 | 0.55 | 0.56 | ||||
| Step 3: Pt characteristics + treatments + center identity: adjusted R2 | 0.56 | 0.57 | ||||
|
Independent variables |
Parameter Estimate |
95% CI |
P |
Parameter Estimate |
95% CI |
P |
| Participant characteristics | ||||||
| Injury group | <0.01 | <0.01 | ||||
| C1–4 ABC | −26.09 | (−30.66, −21.52) | <0.01 | −21.53 | (−26.69, −16.37) | <0.01 |
| C5–8 ABC | −19.56 | (−23.80, −15.32) | <0.01 | −16.21 | (−21.21, −11.22) | <0.01 |
| Para ABC | −15.71 | (−19.39, −12.02) | <0.01 | −15.66 | (−19.79, −11.54) | <0.01 |
| All Ds (Reference) | - | - | - | - | - | - |
| Admission motor FIM™, Rasch-transformed | 0.54 | (0.43, 0.66) | <0.01 | 0.48 | (0.35, 0.60) | <0.01 |
| Comprehensive Severity Index | −0.10 | (−0.15, −0.05) | <0.01 | −0.07 | (−0.12, −0.02) | 0.01 |
| Days from injury to rehabilitation | −0.15 | (−0.20, −0.10) | <0.01 | −0.10 | (−0.15, −0.05) | <0.01 |
| Primary language = English | 6.39 | (0.49, 12.29) | 0.03 | 10.16 | (3.80, 16.51) | <0.01 |
| Treatment variables | ||||||
| Physical therapy total hours | 0.10 | (0.06, 0.15) | <0.01 | 0.14 | (0.10, 0.19) | <0.01 |
| Social work/case management total hours | −0.30 | (−0.45, −0.14) | <0.01 | −0.23 | (−0.37, −0.08) | <0.01 |
Table 4.
Predictors of residence at 1- and 5-years post-injury.
| Outcome: | Reside at home at 1-year anniversary | Reside at home at 5-year anniversary | ||||
|---|---|---|---|---|---|---|
| # Observations used | 786 Yes = 752; No = 34 |
786 Yes = 754; No = 32 |
||||
| Step 1: Pt characteristics: c/Max R2 | 0.67/0.05 | 0.71/0.10 | ||||
| Step 2: Pt characteristics + treatments: c/Max R2 | 0.69/0.10 | 0.71/0.12 | ||||
| Step 3: Pt characteristics + treatments + center identity: c/Max R2 | 0.76/0.15 | 0.75/0.16 | ||||
|
Independent variables |
Odds Ratio Estimate |
95% CI |
P |
Odds Ratio Estimate |
95% CI |
P |
| Participant characteristics | ||||||
| Primary language = English | 3.50 | (1.12, 10.95) | 0.03 | 6.87 | (2.52, 18.77) | <0.01 |
| Treatment variables | ||||||
| Psychology total hours | 0.96 | (0.94, 0.98) | <0.01 | 0.97 | (0.95, 1.00) | 0.03 |
Table 5.
Predictors of working or being in school at 1- and 5-years post-injury.
| Outcome: | Work/School at 1-year anniversary | Work/School at 5-year anniversary | ||||
|---|---|---|---|---|---|---|
| # Observations used | 756: Yes = 239; No = 517 |
756: Yes = 31; No = 725 |
||||
| Step 1: Pt characteristics: c/Max R2 | 0.81/0.35 | 0.70/0.07 | ||||
| Step 2: Pt characteristics + treatments: c/Max R2 | 0.82/0.36 | 0.71/0.08 | ||||
| Step 3: Pt characteristics + treatments + center identity: c/Max R2 | 0.82/0.36 | 0.70/0.07 | ||||
|
Independent variables |
Odds Ratio Estimate |
95% CI |
P |
Odds Ratio Estimate |
95% CI |
P |
| Participant characteristics | ||||||
| Age at injury | 0.97 | (0.96, 0.99) | <0.01 | 0.94 | (0.91, 0.97) | <0.01 |
Table 6.
Predictors of CHART -SF Physical Independence at 1- and 5-years post-injury.
| Outcome: | 1-year CHART-SF: Physical Independence | 5-year CHART-SF: Physical Independence | ||||
|---|---|---|---|---|---|---|
| # Observations used | 758 | 761 | ||||
| Step 1: Pt characteristics: adjusted R2 | 0.41 | 0.34 | ||||
| Step 2: Pt characteristics + treatments: adjusted R2 | 0.45 | 0.36 | ||||
| Step 3: Pt characteristics + treatments + center identity: adjusted R2 | 0.46 | 0.38 | ||||
|
Independent variables |
Parameter Estimate |
95% CI |
P |
Parameter Estimate |
95% CI |
P |
| Participant characteristics | ||||||
| Injury group | <0.01 | <0.01 | ||||
| C1–4 ABC | −27.35 | (−35.62, −19.07) | <0.01 | −33.92 | (−41.83, −26.00) | <0.01 |
| C5–8 ABC | 16.83 | (−24.73, −8.94) | <0.01 | −19.89 | (−27.53, −12.25) | <0.01 |
| Para ABC | −9.11 | (−15.93, −2.29) | 0.01 | −10.91 | (−17.38, −4.44) | <0.01 |
| All Ds (Reference) | - | - | - | - | - | - |
| Admission motor FIM™, Rasch-transformed | 0.73 | (0.53, 0.94) | <0.01 | 0.45 | (0.26, 0.65) | <0.01 |
| Days from injury to rehabilitation | −0.29 | (−0.37, −0.20) | <0.01 | −0.19 | (−0.27, −0.10) | <0.01 |
| Age at injury | −0.18 | (−0.33, −0.02) | 0.03 | −0.15 | (−0.29, −0.01) | 0.03 |
| Payer | <0.01 | <0.01 | ||||
| Medicare | −10.61 | (−20.43, −0.80) | 0.03 | −8.93 | (−18.29, 0.43) | 0.06 |
| Medicaid | −2.26 | (−7.98, 3.45) | 0.44 | −3.66 | (−9.02, 1.70) | 0.18 |
| Worker's compensation | −15.60 | (−22.72, −8.48) | <0.01 | −14.64 | (−21.12, −8.17) | <0.01 |
| Private insurance/pay (Reference) | - | - | - | - | - | - |
| Treatment variables | ||||||
| Physical therapy total hours | 0.24 | (0.16, 0.32) | <0.01 | 0.18 | (0.10, 0.25) | <0.01 |
Table 7.
Predictors of CHART-SF Social Integration at 1- and 5-years post-injury.
| Outcome: | 1-year CHART-SF: Social Integration | 5-year CHART-SF: Social Integration | ||||
|---|---|---|---|---|---|---|
| # Observations used | 742 | 752 | ||||
| Step 1: Pt characteristics: adjusted R2 | 0.18 | 0.14 | ||||
| Step 2: Pt characteristics + treatments: adjusted R2 | 0.20 | 0.17 | ||||
| Step 3: Pt characteristics + treatments + center identity: adjusted R2 | 0.18 | 0.16 | ||||
|
Independent variables |
Parameter Estimate |
95% CI |
P |
Parameter Estimate |
95% CI |
P |
| Participant characteristics | ||||||
| Admission motor FIM™, Rasch-transformed | 0.14 | (0.06, 0.022) | <0.01 | 0.19 | (0.08, 0.30) | <0.01 |
| Age at injury | −0.31 | (−0.42, −0.20) | <0.01 | −0.35 | (−0.49, −0.22) | <0.01 |
| Marital status = Not Married | −10.05 | (−12.94, −7.17) | <0.01 | −9.20 | (−12.69, −5.70) | <0.01 |
| Employment status before injury | <0.01 | <0.01 | ||||
| Working (Reference) | - | - | - | - | - | - |
| Student | −0.16 | (−4.42, 4.11) | 0.94 | 0.89 | (−4.23, 6.00) | 0.73 |
| Retired | 12.09 | (5.74, 18.44) | <0.01 | 11.03 | (3.13, 18.93) | 0.01 |
| Unemployed/other | −4.82 | (−9.04, −0.59) | 0.02 | −5.83 | (−10.81, −0.85) | 0.02 |
| Payer | <0.01 | 0.01 | ||||
| Medicare | −8.19 | (−14.54, −1.85) | 0.01 | −12.77 | (−20.52, −4.97) | <0.01 |
| Medicaid | −7.23 | (−10.74, −3.73) | <0.01 | 0.61 | (−3.55, 4.78) | 0.77 |
| Worker's compensation | −3.49 | (−7.58, 0.60) | 0.09 | −3.86 | (−8.98, 1.27) | 0.14 |
| Private insurance/pay (Reference) | - | - | - | - | - | - |
| Education | 0.02 | <0.01 | ||||
| Less than high school diploma | −5.02 | (−9.11, −0.92) | 0.02 | −10.12 | (−15.05, −5.19) | <0.01 |
| High school diploma or GED | −3.65 | (−6.49, −0.81) | 0.01 | −5.44 | (−8.94, −1.94) | <0.01 |
| More than high school diploma (Reference) | - | - | - | - | - | - |
| Treatment variables | ||||||
| Therapeutic recreation total hours | 0.12 | (0.04, 0.19) | <0.01 | 0.11 | (0.02, 0.20) | 0.02 |
Table 8.
Predictors of CHART-SF Occupation at 1- and 5-years post-injury.
| Outcome: | 1-year CHART-SF: Occupation | 5-year CHART-SF: Occupation | ||||
|---|---|---|---|---|---|---|
| # Observations used | 750 | 771 | ||||
| Step 1: Pt characteristics: adjusted R2 | 0.24 | 0.26 | ||||
| Step 2: Pt characteristics + treatments: adjusted R2 | 0.26 | 0.28 | ||||
| Step 3: Pt characteristics + treatments + center identity: adjusted R2 | 0.27 | 0.30 | ||||
|
Independent variables |
Parameter Estimate |
95% CI |
P |
Parameter Estimate |
95% CI |
P |
| Participant characteristics | ||||||
| Injury group | <0.01 | <0.01 | ||||
| C1–4 ABC | −18.13 | (−27.72, −8.54) | <0.01 | −23.16 | (−31.18, −15.15) | <0.01 |
| C5–8 ABC | −7.97 | (−17.25, 1.30) | 0.09 | −12.02 | (−20.80, −3.24) | 0.01 |
| Para ABC | −6.74 | (−14.57, 1.10) | 0.09 | −0.43 | (−8.12, 7.27) | 0.91 |
| All Ds (Reference) | - | - | - | - | - | - |
| Days from injury to rehabilitation | −0.16 | (−0.25, −0.06) | <0.01 | −0.19 | (−0.29, −0.09) | <0.01 |
| Age at injury | −0.35 | (−0.57, −0.14) | <0.01 | −0.90 | (−1.09, −0.71) | <0.01 |
| Marital status = Not married | −7.25 | (−12.80, −1.71) | 0.01 | −8.79 | (−14.30, −3.28) | <0.01 |
| Education | <0.01 | <0.01 | ||||
| Less than high school diploma | −20.55 | (−28.42, −12.69) | <0.01 | −25.27 | (−32.67, −17.87) | <0.01 |
| High school diploma or GED | −12.68 | (−18.22, −7.13) | <0.01 | −15.19 | (−20.71, −9.66) | <0.01 |
| More than high school diploma (Reference) | - | - | - | - | - | - |
| Treatment variables | ||||||
| Therapeutic recreation total hours | 0.27 | (0.11, 0.42) | <0.01 | 0.19 | (0.04, 0.34) | 0.02 |
Table 9.
Predictors of CHART-SF Mobility at 1- and 5-years post-injury.
| Outcome: | 1-year CHART-SF: Mobility | 5-year CHART-SF: Mobility | ||||
|---|---|---|---|---|---|---|
| # Observations used | 750 | 765 | ||||
| Step 1: Pt characteristics: adjusted R2 | 0.26 | 0.25 | ||||
| Step 2: Pt characteristics + treatments: adjusted R2 | 0.29 | 0.27 | ||||
| Step 3: Pt characteristics + treatments + center identity: adjusted R2 | 0.28 | 0.26 | ||||
|
Independent variables |
Parameter Estimate |
95% CI |
P |
Parameter Estimate |
95% CI |
P |
| Participant characteristics | ||||||
| Injury group | <0.01 | 0.01 | ||||
| C1–4 ABC | −13.44 | (−19.30, −7.59) | <0.01 | −10.24 | (−16.49, −3.99) | <0.01 |
| C5–8 ABC | −9.57 | (−15.20, −3.96) | <0.01 | −5.16 | (−11.16, 0.84) | 0.09 |
| Para ABC | −8.46 | (−13.21, −3.71) | <0.01 | −5.70 | (−10.84, −0.57) | 0.03 |
| All Ds (Reference) | - | - | - | - | - | - |
| Admission motor FIM™, Rasch-transformed | 0.20 | (0.06, 0.34) | 0.01 | 0.22 | (0.07, 0.37) | <0.01 |
| Days from injury to rehabilitation | −0.11 | (−0.17, −0.05) | <0.01 | −0.11 | (−0.18, −0.05) | <0.01 |
| Age at injury | −0.33 | (−0.47, −0.20) | <0.01 | −0.44 | (−0.58, −0.30) | <0.01 |
| Marital status = Not married | −5.56 | (−8.96, −2.16) | <0.01 | −8.33 | (−11.95, −4.71) | <0.01 |
| Race/ethnicity | <0.01 | <0.01 | ||||
| White (Reference) | - | - | - | - | - | - |
| Black | −10.20 | (−14.14, −6.26) | <0.01 | −5.96 | (−10.16, −1.76) | 0.01 |
| Hispanic | 2.52 | (−7.13, 12.17) | 0.61 | −13.32 | (−23.82, −2.82) | 0.01 |
| Other | −3.94 | (−10.98, 3.10) | 0.27 | 2.03 | (−5.70, 9.76) | 0.61 |
| Employment status before injury | <0.01 | <0.01 | ||||
| Working (Reference) | - | - | - | - | - | - |
| Student | 5.90 | (0.90, 10.90) | 0.02 | 7.84 | (2.66, 13.02) | <0.01 |
| Retired | 2.75 | (−4.73, 10.22) | 0.47 | 6.46 | (−0.99, 13.92) | 0.09 |
| Unemployed/other | −6.03 | (−10.92, −1.14) | 0.02 | −3.70 | (−8.82, 1.43) | 0.16 |
| Primary language = English | 15.41 | (7.74, 23.07) | <0.01 | 7.96 | (−0.07, 15.99) | 0.05 |
| Education | <0.01 | <0.01 | ||||
| Less than high school diploma | −8.46 | (−13.36, −3.58) | <0.01 | −16.35 | (−21.48, −11.22) | <0.01 |
| High school diploma or GED | −5.51 | (−8.87, −2.14) | <0.01 | −10.52 | (−14.09, −6.94) | <0.01 |
| More than high school diploma (Reference) | - | - | - | - | - | - |
| Treatment variables | ||||||
| Therapeutic recreation total hours | 0.17 | (0.08, 0.27) | <0.01 | 0.13 | (0.02, 0.23) | 0.01 |
Table 10.
Predictors of depression symptoms (PHQ-9) at 1- and 5-years post-injury.
| Outcome: | 1-year PHQ-9 Total | 5-year PHQ-9 Total | ||||
|---|---|---|---|---|---|---|
| # Observations used | 722 | 708 | ||||
| Step 1: Pt characteristics: adjusted R2 | 0.06 | 0.10 | ||||
| Step 2: Pt characteristics + treatments: adjusted R2 | 0.08 | 0.12 | ||||
| Step 3: Pt characteristics + treatments + center identity: adjusted R2 | 0.07 | 0.10 | ||||
|
Independent variables |
Parameter Estimate |
95% CI |
P |
Parameter Estimate |
95% CI |
P |
| Participant characteristics | ||||||
| Traumatic etiology | 0.02 | 0.01 | ||||
| Vehicular (reference) | - | - | - | - | - | - |
| Violence | 0.39 | (−0.96, 1.73) | 0.57 | 0.82 | (−0.60, 2.23) | 0.26 |
| Sports | −0.93 | (−2.07, 0.22) | 0.11 | −1.25 | (−2.47, −0.03) | 0.04 |
| Fall or falling object | 0.86 | (−0.04, 1.77) | 0.06 | −0.33 | (−1.32, 0.66) | 0.52 |
| Medical/Surgery/Other | 1.92 | (0.10, 3.73) | 0.04 | 2.55 | (0.42, 4.68) | 0.02 |
| Sex = female | 1.14 | (0.21, 2.06) | 0.02 | 1.23 | (0.24, 2.22) | 0.01 |
| Employment status before injury | <0.01 | <0.01 | ||||
| Working (Reference) | - | - | - | - | - | - |
| Student | −0.61 | (−1.71, 0.50) | 0.28 | 0.10 | (−0.97, 1.17) | 0.86 |
| Retired | 0.16 | (−1.44, 1.76) | 0.84 | −1.91 | (−4.06, 0.25) | 0.08 |
| Unemployed/other | 2.38 | (1.17, 3.59) | <0.01 | 3.00 | (1.70, 4.31) | <0.01 |
| Treatment variables | ||||||
| Social work/case management total hours | 0.07 | (0.03, 0.10) | <0.01 | 0.06 | (0.02, 0.10) | <0.01 |
Table 11.
Predictors of global life satisfaction (SWLS) at 1- and 5-years post injury.
| Outcome: | 1-year SWLS Total | 5-year SWLS Total | ||||
|---|---|---|---|---|---|---|
| # Observations used | 672 | 762 | ||||
| Step 1: Pt characteristics: adjusted R2 | 0.09 | 0.11 | ||||
| Step 2: Pt characteristics + treatments: adjusted R2 | 0.10 | 0.12 | ||||
| Step 3: Pt characteristics + treatments + center identity: adjusted R2 | 0.11 | 0.13 | ||||
|
Independent variables |
Parameter Estimate |
95% CI |
P |
Parameter Estimate |
95% CI |
P |
| Participant characteristics | ||||||
| Admission motor FIM™, Rasch-transformed | 0.06 | (0.01, 0.11) | 0.02 | 0.09 | (0.06, 0.13) | <0.01 |
| Age at injury | −0.08 | (−0.12, −0.03) | <0.01 | −0.07 | (−0.11, −0.03) | <0.01 |
| Employment status before injury | 0.01 | <0.01 | ||||
| Working (Reference) | - | - | - | - | - | - |
| Student | 1.69 | (−0.24, 3.62) | 0.09 | −0.26 | (−1.97, 1.45) | 0.77 |
| Retired | 0.98 | (−1.67, 3.62) | 0.47 | 0.79 | (−1.85, 3.43) | 0.56 |
| Unemployed/other | −2.16 | (−4.02, −0.29) | 0.02 | −3.28 | (−5.05, −1.51) | <0.01 |
Table 12.
Predictors of rehospitalization at 1- and 5-years post-injury.
| Outcome: | Rehospitalized between discharge and 1-year anniversary | Rehospitalized in the 12 months preceding 5-year anniversary | ||||
|---|---|---|---|---|---|---|
| # Observations used | 785: Yes = 199; No = 586 | 785: Yes = 231; No = 554 | ||||
| Step 1: Pt characteristics: c/Max R2 | 0.67/0.10 | 0.70/0.14 | ||||
| Step 2: Pt characteristics + treatments: c/Max R2 | 0.73/0.17 | 0.71/0.15 | ||||
| Step 3: Pt characteristics + treatments + center identity: c/Max R2 | 0.73/0.17 | 0.70/0.15 | ||||
| Independent variables | Odds Ratio Estimate | 95% CI | P | Odds Ratio Estimate | 95% CI | P |
| Participant characteristics | ||||||
| Injury group | <0.01 | <0.01 | ||||
| C1–4 ABC | 5.21 | (2.48, 10.96) | <0.01 | 3.78 | (1.95, 7.32) | <0.01 |
| C5–8 ABC | 3.71 | (1.71, 8.03) | 0.08 | 2.12 | (1.04, 4.31) | 0.93 |
| Para ABC | 2.83 | (1.40, 5.69) | 0.79 | 2.37 | (1.23, 4.55) | 0.37 |
| All Ds (Reference) | - | - | - | - | - | - |
Table 13.
Predictors of pressure ulcer at 1- and 5-years post-injury.
| Outcome: | Pressure ulcer at 1-year anniversary | Pressure ulcer at 5-year anniversary | ||||
|---|---|---|---|---|---|---|
| # Observations used | 750: Yes = 82; No = 658 |
750: Yes = 119; No = 631 |
||||
| Step 1: Pt characteristics: c/Max R2 | 0.68/0.08 | 0.66/0.08 | ||||
| Step 2: Pt characteristics + treatments: c/Max R2 | 0.69/0.09 | 0.66/0.08 | ||||
| Step 3: Pt characteristics + treatments + center identity: c/Max R2 | 0.72/0.14 | 0.69/0.10 | ||||
|
Independent variables |
Odds Ratio Estimate |
95% CI |
P |
Odds Ratio Estimate |
95% CI |
P |
| None in common | ||||||
Table 14.
Summary of predictors of Year 1 and Year 5 outcomes by variable type.
| Key | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Year 1 Only | Year 5 Only | Both | |||||||||
| Outcomes | |||||||||||
| Independent variables | FIM Motor | Reside at Home | Return to Work/ School | CHART: Physical Independence | CHART: Social Integration | CHART: Occupation | CHART: Mobility | Rehospitalization | Pressure Sore | PHQ-9 Total | SWLS Total |
| Participant characteristics | ; | ||||||||||
| Injury group | |||||||||||
| Admission motor FIM, Rasch-transformed | |||||||||||
| Admission cognitive FIM, Rasch-transformed | |||||||||||
| Comprehensive Severity Index | |||||||||||
| Days from injury to rehabilitation | |||||||||||
| Traumatic etiology | |||||||||||
| Age at injury | |||||||||||
| Sex is Female | |||||||||||
| Marital status is Not Married | |||||||||||
| Race/ethnicity | |||||||||||
| Employment status before injury | |||||||||||
| Injury work related | |||||||||||
| Body mass index at admission > = 30 | |||||||||||
| Primary language = English | |||||||||||
| Payer | |||||||||||
| Education | |||||||||||
| Treatment variables | |||||||||||
| Length of rehabilitation stay (days) | |||||||||||
| Clinician experience index | |||||||||||
| Occupational therapy total hours | |||||||||||
| Psychology total hours | |||||||||||
| Physical therapy total hours | |||||||||||
| Registered nursing total hours | |||||||||||
| Speech language pathology total hours | |||||||||||
| Social work/case management total hours | |||||||||||
| Therapeutic recreation total hours | |||||||||||
Motor FIM™ scores
Participant characteristics alone were strong predictors of the FIM™ motor scores at 1 and 5-years post-injury (Year 1 R2 = 0.53, Year 5 R2 = 0.53). The addition of treatment variables minimally increased the explained R2 at both 1- and 5-years to 0.55 and 0.56, respectively. Table 3 shows which variables predicted both 1- and 5-year FIM™ motor scores. The common predictors of lower motor FIM™ scores at 1- and 5-years were level and completeness of injury, admission motor FIM™ scores, days from injury to rehabilitation, English as a primary language, and time spent in physical therapy and with social work/case management.
Residence
Most participants resided at home at both 1- (94.9%) and 5-years (95.2%) post-injury. Participant characteristics alone distinguished between participants who resided at home at 5 years versus those who did not at 1 year, although the models had low predictive strength (Year 1 c statistic = 0.67, Max R2 = 0.05; Year 5 c statistic = 0.71, Max R2 = 0.10). The only positive predictors for living at home at both 1- and 5-years were speaking English as one's primary language and hours spent in psychology (Table 4).
Work/school status
Participant characteristics distinguished work/school status at 1 and 5-years post-injury, although the model had lower predictive strength at 5 years (c statistic = 0.70, Max R2 = 0.07) than at 1 year (c statistic = 0.81, Max R2 = 0.35). Treatment covariates added minimal predictive strength to the model (Table 5). The only common predictor of being in work or school at 1- and 5-years post-injury was age at injury.
Societal participation
Tables 6–9 display regression models predicting the four dimensions of the CHART-SF at 1- and 5-years post-injury: Physical Independence (year 1 R2 = 0.45, Year 5 R2 = 0.36), Social Integration (year 1 R2 = 0.20, Year 5 R2 = 0.17), Occupation (year 1 R2 = 0.26, Year 5 R2 = 0.28), and Mobility (year 1 R2 = 0.29, Year 5 R2 = 0.27). Age at injury predicted outcomes on all four dimensions of the CHART-SF at Year 1 and Year 5. The most frequent predictors at both time points were injury group (Physical Independence, Occupation, Mobility), admission motor FIMTM (Physical Independence, Social Integration, Mobility), days from injury to rehabilitation (Physical Independence, Occupation, Mobility), marital status (Social Integration, Occupation, Mobility), education (Social Integration, Occupation, Mobility), and hours spent in therapeutic recreation (Social Integration, Occupation, Mobility).
Mood state and life satisfaction
PHQ-9 items were completed by 92% of participants at Year 1 (M = 4.50, IQR = 1–7) and by 90% of participants at Year 5 (M = 5.00, IQR = 1–7). Participant characteristics and quantity of treatment by specific disciplines were weak predictors of depression symptoms (Year 1 R2 = 0.08; Year 5 R2 = 0.12) (Table 10). Predictors common to both time points were traumatic etiology, female sex, employment status before injury, and hours spent in social work/case management.
SWLS questions were completed by 85% of participants at Year 1 and by 97% at Year 5. The mean SWLS score at Year 1 was 21.00, IQR 15–26; at Year 5 the mean score was 22.35, IQR 17–29. Models predicting SWLS were weak at both time points; the adjusted R2 for patient and treatment variables at Year 1 was 0.10, and 0.12 at Year 5 (Table 11). The addition of rehabilitation center added little explanatory value to the model. Predictors common to both time points were admission motor FIMTM, age at injury, and employment status before injury.
Rehospitalization
Participant characteristics alone distinguished between those who were rehospitalized in the 365 days preceding the first anniversary of their injury and those who were not, although the model had weak predictive strength (c statistic = 0.67, Max R2 = 0.10) (Table 12). Results were similar at the 5-year anniversary (c statistic = 0.70, Max R2 = 0.14). The only common predictor at both time points was injury group.
Pressure ulcer
Patient characteristics alone constituted a fair model fit at both Year 1 (c statistic = 0.68, Max R2 = 0.08) and Year 5 (c statistic = 0.66, Max R2 = 0.08). The addition of the rehabilitation center slightly improved the model fit (Year 1 c statistic = 0.72, Max R2 = 0.14; Year 5 c statistic = 0.69, Max R2 = 0.10). There were no predictors in common at the 1- and 5-year time points (Table 13).
Discussion
Using a subset of SCIRehab data for participants who took part in post-discharge interviews at 1- and 5-years post-injury, associations between injury-related, demographic, rehabilitation discipline treatment time, and a variety of 5-year post-injury outcomes were examined. Particular interest was placed on the relationships between time spent in specific inpatient rehabilitation treatment modalities and 5-year outcomes in comparison to previously published 1-year outcomes. Our hypothesis was partially supported with the observation that some of the relationships found at 1 year continued to have statistical significance at 5 years. Results from the present analyses indicate patient characteristics continue to be strong predictors of outcomes 5-years post-injury, although several therapeutic variables added to the predictability of some of the outcomes. While some associations with 1-year outcomes held at 5 years, other variables were no longer associated and some additional relationships were revealed. A participant's injury group, days from injury to rehabilitation, and age at injury predicted the most outcomes consistently between 1- and 5-years. Of the variables that predicted 1- and 5-year outcomes, all but seven maintained their predictive direction. The most frequent of these seven variables is employment status.
Armed with data from two time points that demonstrate significant associations and durability over time, ideas can be gleaned to improve clinical care delivery and access, surveillance and prevention, prognostic counseling, and research. Although the associations do not represent causality, it is worth considering the potential therapeutic implications of relationships found between treatments during inpatient rehabilitation outcomes and the durability of those relationships. For example, injury group, which is predictive of rehospitalization, may serve as an important marker to guide resource utilization planning, education, and targeted preventive efforts.
The findings may support access to and relevance of services. More time spent in physical therapy and therapeutic recreation were positive predictors of both 1- and 5-year outcomes. Perhaps early learning and adaptation to leisure and recreational possibilities are of great importance to later outcomes. The positive association of greater motor independence with more physical therapy was consistent across the two physical measures and endured at 5 years. Given that total psychology hours positively predicted home residence at Years 1 and 5, access to psychology might be important to whether one lives in the community.
Depression and life satisfaction were weakly predicted by participant and treatment variables. However, finding that more time with psychology and social work/case management predicted greater depressive symptomatology 5-years post-injury is noteworthy. These increased inpatient services may indicate a greater need and symptom severity, warranting additional referral post-discharge. Future studies should look more in-depth at psychosocial needs early post-injury and develop more targeted intervention strategies for long-term sustainable gains. Traumatic etiology, especially among females, may indicate the need for enhanced surveillance or preemptive counseling and education.
Clinician experience was a predictor of both 1- and 5-year outcomes, although the related outcomes varied across years. While the concept of clinician experience has not been studied extensively within SCI rehabilitation, researchers have found that nurses’ experience significantly predicts better outcomes.37,38 Additionally, older, more experienced nurses, physicians, and psychosocial professionals tend to recognize and reflect on their emotions, which can translate to improved outcomes.39 However, a review of 59 studies found more experienced physicians provided lower-quality care with worse outcomes.40 An interdisciplinary study is warranted to investigate further the impact of the multidisciplinary team's experience on rehabilitation outcomes.
Limitations
When interpreting and applying these findings, one should consider that this study was conducted at highly specialized SCI rehabilitation centers. Generalizability to all rehabilitation centers is unknown. The association between clinician experience and outcomes may likely be more pervasive and profound in less specialized centers. While SCIRehab suggests many associations between injury-related and demographic characteristics and rehabilitation discipline treatment time, it is important to remember that these results are correlational and do not imply causality. Not all rehabilitation treatments and disciplines were considered in this study, and therapies were condensed to units of time. For example, the study did not examine medications, other medical treatments, rehabilitation physician experience, different aspects of rehabilitation nursing care, type of therapy delivered, respiratory care, chaplaincy, and rehabilitation engineering, all of which greatly impact outcomes. All treatment variables were limited to inpatient rehabilitation treatments. Thus, the association between outpatient treatments received after discharge and 5-year outcomes is unknown. Errors may exist in the treatment time due to omissions, duplicate reports, and documentation errors. However, these errors were minimized through comparison of therapist reports, billing records, and other available information. Documentation errors are also possible in the data obtained from medical records, such as the demographic and injury information, neurological classification, and the CSI. The payer source is that which was responsible for participants’ inpatient rehabilitation treatment payment. Payer source for post-rehabilitation services may differ from that studied. We also note that there was substantial attrition in this study, with 1,376 participants at 1 year and 792 participants at 5 years, which led to some differences in demographic and injury characteristics.
Conclusion
Associations found between inpatient rehabilitation variables and 1- and 5-year outcomes may play a role in prognostication, future treatment development, and outcome durability. Further research is warranted to determine if attention to and adjustment of the potentially modifiable predictive factors may improve long-term SCI outcomes. In particular, access to clinicians with enhanced experience appears to be an important factor to consider in SCI rehabilitation.
Disclaimer statements
Contributors None.
Funding The data were collected and analyzed under grants from the National Institute on Disability, Independent Living, and Rehabilitation Research (formerly NIDRR), grant numbers H133A060103, H133N060005, 90SI5003 (formerly H133N11000), and 90SI5015 to Craig Hospital; H133N060027 and 90SI5017 to Icahn School of Medicine at Mount Sinai; H133A21943016 to Carolinas Rehabilitation; H133N060009 and 90SI5002 (formerly H133N110005) to Shepherd Center; and H133N060014 and 90SI5009 (formerly H133N110014) to the Rehabilitation Institute of Chicago (now the Shirley Ryan Ability Lab).
Conflicts of interest There are no conflicts of interest to declare.
Data availability statement The data that support the findings of this study are available from the corresponding author, KRM, upon reasonable request.
Supplementary Material
References
- 1.Natale A, Taylor S, LaBarbera J, Mumma S, Bensimon L, McDowell S, et al. SCIRehab: the physical therapy taxonomy. J Spinal Cord Med 2009;32(3):270–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Johnson K, Bailey J, Rundquist J, Dimond P, McDonald C, Reyes IA, et al. SCIRehab: the supplemental nursing taxonomy. J Spinal Cord Med 2009;32(3):328–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ozelie R, Sipple C, Foy T, Cantoni K, Kellogg K, Lookingbill J, et al. SCIRehab: the occupational therapy taxonomy. J Spinal Cord Med 2009;32(3):283–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Wilson C, Huston T, Koval J, Gordon S, Schwebel A, Gassaway J.. SCIRehab: the psychology taxonomy. J Spinal Cord Med 2009;32(3):318–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gordan W, Dale B, Brougham R, Spivack-David D, Georgeadis A, Adornato V, et al. SCIRehab: the speech language pathology taxonomy. J Spinal Cord Med 2009;32(3):306–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cahow C, Skolnick S, Joyce J, Jug J, Dragon C, Gassaway J.. SCIRehab: the therapeutic recreation taxonomy. J Spinal Cord Med 2009;32(3):297–305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Abeyta N, Freeman E, Primack D, Harmon A, Dragon C, Hammond F, et al. SCIRehab: the social work/case management taxonomy. J Spinal Cord Med 2009;32(3):335–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gassaway J, Whiteneck G, Dijkers M.. Clinical taxonomy development and application in spinal cord injury research: the SCIRehab project. J Spinal Cord Med 2009;32(3):260–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Whiteneck G, Gassaway J, Dijkers M, Heinemann A, Kreider S.. Relationships of patient characteristics and rehabilitation services to outcomes following inpatient spinal cord injury: the SCIRehab project. J Spinal Cord Med 2012;35(6):484–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Teeter L, Gassaway J, Taylor S, LaBarbera J, McDowell S, Backus D, et al. Relationship of physical therapy inpatient rehabilitation interventions and patient characteristics to outcomes following spinal cord injury: the SCIRehab project. J Spinal Cord Med 2012;35(6):503–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Cahow C, Gassaway J, Rider C, Joyce J, Bogenshutz A, Edens K, et al. Relationship of therapeutic recreation inpatient rehabilitation interventions and patient characteristics to outcomes following spinal cord injury: the SCIRehab project. J Spinal Cord Med 2012;35(6):547–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Ozelie R, Gassaway J, Foy T, Perritt G, Thimmaiah D, Heisler L, et al. Relationship of occupational therapy inpatient rehabilitation interventions and patient characteristics to outcomes following spinal cord injury: the SCIRehab project. J Spinal Cord Med 2012;35(6):527–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Heinemann AW, Wilson C, Huston T, Koval J, Gordon S, Gassaway J, et al. Relationship of psychology inpatient rehabilitation services and patient characteristics to outcomes following spinal cord injury: the SCIRehab project. J Spinal Cord Med 2012;35(6):578–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gordan W, Gerber D, Spivack-David D, Adornato V, Brougham R, Gassaway J, et al. Relationship of speech language pathology inpatient rehabilitation interventions and patient characteristics to outcomes following spinal cord injury: the SCIRehab project. J Spinal Cord Med 2012;35(6):565–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Whiteneck G, Gassaway J.. The SCIRehab project: what rehabilitation interventions are most strongly associated with positive outcomes after spinal cord injury? J Spinal Cord Med 2012;35(6):482–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hammond F, Gassaway J, Abeyta N, Freeman E, Primack P, Kreider S, et al. Outcomes of social work and case management services during inpatient spinal cord injury rehabilitation: the SCIRehab project. J Spinal Cord Med 2012;35(6):611–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bailey J, Dijkers MP, Gassaway J, Thomas J, Lingefelt P, Kreider SED, et al. Relationship of nursing education and care management inpatient rehabilitation interventions and patient characteristics to outcomes following spinal cord injury: The SCIRehab project. J Spinal Cord Med [Internet] 2012 Nov. [cited 2020 Nov 23];35(6):593–610. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3522899/. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Saunders LL, Krause JS, Focht KL.. A longitudinal study of depression in survivors of spinal cord injury. Spinal Cord [Internet] 2012 Jan. [cited 2020 Nov 23];50(1):72–7. Available from: http://www.nature.com/articles/sc201183. [DOI] [PubMed] [Google Scholar]
- 19.Hoffman JM, Bombardier CH, Graves DE, Kalpakjian CZ, Krause JS.. A longitudinal study of depression from 1 to 5 years after spinal cord injury. Arch Phys Med Rehabil [Internet] 2011;92(3):411–8. Available from: https://www.ncbi.nlm.nih.gov/pubmed/21353823. [DOI] [PubMed] [Google Scholar]
- 20.Pretz CR, Kozlowski AJ, Chen Y, Charlifue S, Heinemann AW.. Trajectories of life satisfaction after spinal cord injury. Arch Phys Med Rehabil [Internet] 2016 Oct 1. [cited 2019 Sep 24];97(10):1706–13.e1. Available from: http://www.sciencedirect.com/science/article/pii/S0003999316301848. [DOI] [PubMed] [Google Scholar]
- 21.van Leeuwen CM, Post MW, Hoekstra T, van der Woude LH, de Groot S, Snoek GJ, et al. Trajectories in the course of life satisfaction after spinal cord injury: identification and predictors. Arch Phys Med Rehabil [Internet] 2011 Feb. [cited 2020 Feb 6];92(2):207–13. Available from: https://linkinghub.elsevier.com/retrieve/pii/S0003999310008178. [DOI] [PubMed] [Google Scholar]
- 22.Gélis A, Dupeyron A, Legros P, Benaïm C, Pelissier J, Fattal C.. Pressure ulcer risk factors in persons with spinal cord injury part 2: the chronic stage. Spinal Cord [Internet] 2009 Sep. [cited 2020 Nov 23];47(9):651–61. Available from: https://proxy.hsl.ucdenver.edu/login?url=https://search.ebscohost.com/login.aspx?direct=true&db=aph&AN=44012337&site=ehost-live. [DOI] [PubMed] [Google Scholar]
- 23.Gassaway J, Sweatman M, Rider C, Edens K, Weber M.. Therapeutic recreation outcomes during inpatient SCI rehabilitation: propensity score analysis of SCIRehab data. Ther Recreation J 2019;53(2):99–116. [Google Scholar]
- 24.Horn S, Gassaway J.. Practice based evidence: incorporating clinical heterogeneity and patient-reported outcomes for comparative effectiveness research. Med Care 2010;48(6):17–22. [DOI] [PubMed] [Google Scholar]
- 25.Fiedler R, Granger C.. Functional Independence measure: a measurement of disability and medical rehabilitation. In: Chino N, Melvin J, (ed.) Functional elevation of stroke patients. Tokyo: Springer-Verlag; 1996. p. 75–92. [Google Scholar]
- 26.Horn SD, Sharkey PD, Buckle JM, Backofen JE, Averill RF, Horn RA.. The relationship between severity of illness and hospital length of stay and mortality. Med Care 1991;29: 305–17. [DOI] [PubMed] [Google Scholar]
- 27.Marino R. Reference manual for the International Standards for neurological classification of spinal cord injury. Chicago (IL: ): American Spinal Injury Association; 2003. [DOI] [PubMed] [Google Scholar]
- 28.National Spinal Cord Injury Statistical Center . Data collection syllabus for the National Spinal cord injury database: 2006–2011. Birmingham (AL: ): University of Alabama; 2011. [Google Scholar]
- 29.Fiedler R, Granger C, Russell C.. UDS(MR)SM: follow-up data on patients discharged in 1994-1996. Am J Phys Med Rehabil 2000;79(2):184–92. [DOI] [PubMed] [Google Scholar]
- 30.Whiteneck G, Gassaway J, Dijkers M, Charlifue S, Backus D, Chen D, et al. Inpatient treatment time across disciplines in spinal cord injury rehabilitation. J Spinal Cord Med 2011;34(11):133–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Whiteneck G, Gassaway J, Dijkers M, Jha A.. New approach to study the contents and outcomes of spinal cord injury rehabilitation: the SCIRehab project. J Spinal Cord Med 2009;32(3):251–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Whiteneck G, Brooks C, Charlifue S, Gerhart K, Mellick D, Overholser D, et al. Guide for use of CHART: Craig Hospital assessment and reporting technique. Englewood (CO: ): Craig Hospital; 1992. [Google Scholar]
- 33.Hall JM, Dijkers M, Whiteneck G, Brooks CA, Krause JS.. The Craig Handicap assessment and reporting technique (CHART): metric properties and scoring. Top Spinal Cord Inj Rehabil 1998;4(1):16–30. [Google Scholar]
- 34.Diener E. Assessing subjective well-being progress and opportunities. Soc Indic Res 1993;31:103–57. [Google Scholar]
- 35.Spitzer RL, Kroenke K, Williams JB.. Validation and utility of a self-report version of PRIME-MD: the PHQ primary care study. primary care evaluation of mental disorders. patient health questionnaire. JAMA 1999;282(18):1737–44. [DOI] [PubMed] [Google Scholar]
- 36.Menard S. Coefficients of determination for multiple logistic regression analysis. Am Stat 2000;54:17–24. [Google Scholar]
- 37.Blegen MA, Vaughn TE, Goode CJ.. Nurse experience and education: effect on quality of care. J Nurs Adm 2001;31(1):33–9. [DOI] [PubMed] [Google Scholar]
- 38.Kendall-Gallagher D, Blegen MA.. Competence and certification of registered nurses and safety of patients in intensive care units. Am J Crit Care 2009;18(2):106–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Martin EB Jr, Mazzola NM, Brandano J, Luff D, Zurakowski D, Meyer EC.. Clinicians’ recognition and management of emotions during difficult healthcare conversations. Patient Educ Couns 2015;98(10):1248–54. [DOI] [PubMed] [Google Scholar]
- 40.Choudhry NK, Fletcher RH, Soumerai SB.. Systematic review: the relationship between clinical experience and quality of health care. Ann Intern Med 2005;142(4):260–73. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.