Abstract
Context/Objective
Community integration (CI) is a crucial rehabilitation goal after spinal cord injury (SCI). There is a pressing need to enhance our understanding of the factors associated with CI for individuals with traumatic or non-traumatic etiologies, with the latter being notably understudied. Accordingly, our research explores the associations and potential mediators influencing CI across these populations.
Setting
Specialized neurological rehabilitation center.
Participants
Community-dwelling individuals who were admitted as inpatients within 3 months post-injury (n = 431, 51.9% traumatic, 48.1% non-traumatic), assessed in relation to community integration within 1–3 years after discharge.
Outcome Measure
Community Integration Questionnaire (CIQ). Covariates: American Spinal Injury Association Impairment Scale (AIS), Functional Independence Measure (FIM) and Hospital Anxiety and Depression Scale (HADS).
Results
Multiple linear regression yielded age, B AIS grade, educational level (< 6 years and <12 years), time since injury to admission, length of stay, HADS-depression at discharge, total FIM at discharge and three social work interventions (support in financial, legal and transportation services) as significant predictors of total CIQ score (Adjusted R2 = 41.4). Multiple logistic regression identified age, traumatic etiology, educational level (< 6 years and <12 years), length of stay, HADS depression at discharge, total FIM at discharge and one social work intervention (transportation support) as significant predictors of good community integration, AUC (95% CI): 0.82 (0.75-0.89), Sensitivity:0.76, Specificity:0.73. We identified motor FIM at discharge and motor FIM efficiency as causal mediators of total CIQ.
Conclusions
We identified modifiable factors during rehabilitation—functional independence, depression, and social work interventions—that are associated with CI.
Keywords: Community integration, Spinal cord injury, Traumatic, Non-traumatic, Causal
Introduction
Globally, between 250,000 and 500,000 individuals experience a spinal cord injury (SCI) each year, with an estimated two to three million people worldwide living with an SCI-related disability (1). Inpatient physical rehabilitation following injury often results in improved motor task performance and independence in daily living activities (2). However, over the past recent decades the rehabilitation paradigm has shifted to a biopsychosocial model in which the medical, psychological and social aspects of the injury are simultaneously addressed (3). As a result, community integration is now a core goal of rehabilitation (4). Integration into the community is a multi-faceted issue that refers to participation in society, including engagement in occupation, recreation, social activities and relationships with others (5). Although community integration is a key goal of SCI rehabilitation, it is recognized as a challenge to obtain through rehabilitation (6). A previous study demonstrated that within the first year following SCI, 55% of the participants struggled with social participation, though data was collected during 2010–2012 in Sidney involving 90% traumatic etiologies (7).
As the lengths of hospital stays and inpatient rehabilitation after SCI are globally decreasing (8, 9) and particularly in Europe (10), there is less time to help people with SCI prepare for community integration. Yet, the quantity of inpatient rehabilitation services received was previously reported to be weakly associated with participation early post-SCI in a study conducted in the United States involving traumatic etiologies (11). There is a pressing need to increase the understanding of factors that affect successful integration and inclusion, and to develop strategies to facilitate meaningful participation.
Compared with traumatic SCI, there are relatively few publications on non-traumatic SCI (12). The incidence of non-traumatic SCI is increasing as it occurs more commonly in older age groups. Given global ageing, the incidence of non-traumatic SCI is predicted to surpass their traumatic counterparts in the near future. Besides age, patients with non-traumatic etiologies often exhibit incomplete lesions associated with paraplegia, and there is a higher proportion of female and retired individuals in this group (13). Additionally, these patients frequently suffer from co-existing conditions such as diabetes, cardiovascular and pulmonary diseases, along with impaired memory and retention capabilities (13). These factors collectively contribute to distinct community integration challenges for these individuals. Specifically, older age may impact the availability and type of inner circle family support. Older patients often rely more on spousal support, which may be limited due to the spouse’s own health issues or decreased physical capabilities. Conversely, younger patients might access more diverse family support from parents or siblings, enhancing their community integration prospects. Additionally, the presence of multiple health issues can necessitate specialized healthcare services and social support systems, thus affecting their overall integration into the community.
We hypothesized that specific inpatient demographics, clinical baseline measures, rehabilitation interventions and outcomes are associated with community integration at statistically significant levels. Given that the linkage between detailed rehabilitation outcomes and community reintegration is scarcely studied, our study seeks to fill this gap. Therefore the objectives of this study were to analyze post-acute patients admitted to rehabilitation within three months after SCI (in a neurological rehabilitation center in Spain), and assessed in relation to their community integration situation within 1–3 years after discharge, in order to (1) Compare individuals with traumatic and non-traumatic etiologies at baseline in relation to medical, psychological and social aspects (2) investigate predictive factors of community integration (eventually modifiable during inpatient rehabilitation, such as depression, anxiety, independence in ADLs or social work interventions) and (3) identify causal mediators of good community integration.
Methods
Study design
We conducted a retrospective observational cohort study analyzing patients treated at the institut Guttmann hospital in Catalonia Spain. This study conforms to the STROBE Guidelines (“Strengthening the Reporting of Observational Studies in Epidemiology”) (14). The data that support the findings of this work are available from the corresponding author upon reasonable request.
Participants and setting
This study was carried out in the Institut Guttmann, a specialized care, university Spanish hospital serving (mainly) the population from Catalonia, Spain. This is a major referral neurological center, accredited by the Joint Commission International. All the patients admitted within three months after injury at the Spinal Cord Injury Unit due to traumatic and non-traumatic causes between May 2007 and June 2021 with a complete community integration assessment performed within 1–3 years after discharge were initially included. Cases where we were not able to confirm date of injury or date of birth, cases with a lack of neurological assessment using the American Spinal Injury Association (ASIA) Impairment Scale (AIS) or neurological level of injury, cases with no complete functional independence or psychological assessments, patients with a rehabilitation length of stay of fewer than 3 days, were excluded from the study.
Clinical included variables
International Standards for the neurological classification of spinal cord injuries according to the AIS grades were used to assess the affected neurological level and the extent of the SCI. The three groups established were: C1–C8 AIS A,B,C; T1–S5 AIS A,B,C and AIS D (15). The International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) was used to report the neurological level of injury.
The Functional Independence Measure (FIM) (16) was used for measuring independence in ADLs, it has been extensively applied in rehabilitation settings for individuals with SCI (17). The total FIM score ranges from 18 to 126, covering, higher scores represent greater functional independence.
In our results we analyze gain and efficiency for the FIM, calculated as follows (18): Gain = FIM score at discharge – FIM score at admission and Efficiency = Gain / length of stay.
In relation to community integration we followed the approach introduced by Willer and colleagues (5). They defined community integration as “effective role performance in community settings’ in three different contexts: home, social and productivity” and introduced the Community Integration Questionnaire (CIQ). The CIQ is a 15-item scale including 3 main subscales (home, social and productivity). The CIQ is reported to have adequate internal consistency (∝ ≥ 0.7), high interrater reliability (r ≥ 0.9, p < 0.01), and high construct validity (19). It is a recognized measure for examining community integration following SCI (20) and is extensively used in SCI research (19, 21,22). The higher the total CIQ score, the higher the community integration (Range 0-25). Low integration was defined as total CIQ < 15.1 (23).
Anxiety and depression were assessed with the Hospital Anxiety and Depression Scale (HADS) (24) with extensive applications in SCI (25). The severity of anxiety and depression is assessed using two subscales, each consisting of seven items that are rated on a four-point Likert-type scale. The following cutoff scores have been used previously to denote probable anxiety or depression and were also applied in this study: 0-7: No case; 8-10: Borderline abnormal; 11-21: Case (24).
Social work interventions in clinical rehabilitation
Social workers’ role in healthcare organizations (especially during inpatient rehabilitation after SCI) supporting patients’ social needs is increasingly being recognized (26). Social workers in healthcare settings are uniquely qualified to provide interventions to address patients’ social needs (27). At Hospital Institut Guttmann, the categorization of interventions has been systematically organized using standardized categories consistent with established practices found in related research (28). These categories have remained unchanged throughout the entire period under study:
Financial (I1): Assisting with securing funding for medications, treatments, and necessary healthcare services to eliminate financial barriers to care.
Supportive Counseling (I2): Providing emotional and mental health support, including e.g. services for addiction, to enhance psychological resilience and well-being.
Care Coordination (I3): Managing case files, arranging home visits, and coordinating care across various healthcare settings to ensure continuity and comprehensive support.
Education (I4): Informing about health and wellness programs and crisis intervention, facilitating better self-management of health conditions.
Community Service (I5): Connecting patients to community and peer advocacy resources; facilitating access to food, clothing, and translation services.
Applications and Reporting (I6): Handling the filing of applications and mandatory reports for agencies such as the Department of Child Services and Adult Protective Services.
Housing (I7): Assisting in securing stable housing arrangements, from emergency shelters to long-term housing solutions.
Transportation (I8): Arranging transportation fares and passes and linking with transportation services to facilitate access to medical and community resources.
Durable Medical Equipment (I9): Managing the provision, maintenance, and replacement of essential medical/orthopedic equipment to support physical health and mobility.
Legal (I10): Facilitating connections to legal aid, including attorneys and law enforcement, to address legal concerns related to healthcare and rehabilitation.
The cohort of included participants was identified through a thorough review of the hospital's electronic health records (EHRs), which encompassed the entire period under study. This approach ensured a comprehensive inclusion of patients admitted to our rehabilitation center within the specified timeframe.
As for the outcome measurements and covariates (FIM, HADS, AIS, CIQ), these were performed and collected by trained professionals adhering strictly to the hospital’s established protocols. These protocols are designed to ensure consistency and accuracy in data collection during the whole period under study, allowing for reliable assessments of patient outcomes over time.
Statistical analyses
Statistical analyses were performed using R-v3.5.1 (64 bits) (29). The significance level was set at P < 0.05. Absolute and relative frequencies were reported for categorical variables. Non-parametric statistics with median and interquartile range (IQR) were used for non-normally distributed continuous variables. Shapiro–Wilk test was used to assess normality. Differences between groups were analyzed with Student t-test, Mann–Whitney U-test and Chi2 test (χ2).
Multiple linear regression analysis was used to investigate total CIQ predictors. In the multiple regression models, multicollinearity of independent variables was tested by the variance inflation factor (VIF) a VIF above 2 indicates a multicollinearity problem (30).
Potential predictors of low community integration were analyzed using logistic regression with dependent variable total CIQ dichotomized using the 15.1 cutoff (23). To test the models’ goodness of fit and accuracy, AUC-ROC curves were used. AUC was calculated by means of 10-fold cross validation in a test set, we independently partitioned initial data in training set (70%) and test set (30%), as in related research using the caret R package (31).
We used a model-based inference approach (32) for causal mediation analyses, applying the mediation R package (33). We analyzed modifiable factors (FIM, HADS and social work interventions) as possible mediators and constructed individual single mediator models for each of such hypothesized mediators(34). For each of the single mediator models, we estimated the average causal mediation effect (ACME), average direct effect (ADE), the total effect and the proportion mediated (33). As the mediator-outcome path may be confounded, we used a directed acyclic graph (DAG) to describe causal assumptions required to avoid confounding and selection bias, as in related research (35) (Fig. 1).
Figure 1.
Participants’ selection flowchart.
Ethical considerations
The study was approved by the Institut Guttmann Ethics Committee of Clinical Research. The participants were anonymized and non-identifiable. Participants provided informed written consent to be included in research studies addressed by hospital Institut Guttmann.
Results
The source population (n = 491) was all the patients admitted within three months after injury at the Spinal Cord Injury Unit of institut Guttmann hospital due to traumatic or non-traumatic causes between 2007 and 2021 with at least one CIQ assessment performed within 1–3 years after discharge. Cases where we were not able to confirm date of injury or date of birth (n = 2), cases with a lack of neurological assessment using the American Spinal Injury Association (ASIA) Impairment Scale (AIS) or neurological level of injury (n = 6), cases with no complete FIM assessment within a week after admission or discharge (n = 29) or HADS assessments (n = 16), patients with a rehabilitation length of stay of fewer than 3 days (n = 7), were excluded from the study, leaving (n = 431) the final sample (Fig. 1). None of the enrolled participants died during hospitalization, either before CIQ assessment. None of the participants had a change in AIS grade from admission to discharge.
The number of analyzed participants by year (at the moment of CIQ assessment, i.e. between 2009 and 2023) are presented in Supplementary material Figure SM1. The mean time since discharge to CIQ assessment was 2.3 (0.6) years.
Detailed etiologies are presented in Table SM1, most frequent traumatic causes were traffic accidents (39.3%), falls (25.0%) and work-related accidents (17.0%). Meanwhile, most frequent non-traumatic causes were degenerative (22.2%), vascular (21.7%), tumors (20.3%) and iatrogenesis (12.6%).
Comparing socio-demographics, clinical outcomes and social work interventions for patients with traumatic vs non-traumatic etiologies
Table 1 presents demographic and clinical variables. Mean age at injury was significantly older for non-traumatic injuries compared to traumatic injuries (59 vs 41 years). The proportion of males and females was significantly different being close to 50–50 for non-traumatic and 80% males in traumatic causes. Complete injuries (AIS A) were significantly lower for non-traumatic patients (15.0% vs 45.1%) with significantly higher FIM scores (73 vs 64 points in total FIM) and shorter LOS (75.1 vs 92.9 days).
Table 1.
Demographics and clinical characteristics for traumatic vs non-traumatic included participants (n = 431).
| Total (N = 431) |
Non-traumatic (N = 207) |
Traumatic (N = 224) |
P | |
|---|---|---|---|---|
| Age at injury | 50.8 (35.9–63.9) | 59.4 (46.6–66.8) | 41.4 (28.7–57.4) | <0.001 |
| Sex Male | 295 (68.4%) | 114 (55.1%) | 181 (80.8%) | <0.001 |
| AIS | <0.001 | |||
| A | 132 (30.6%) | 31 (15.0%) | 101 (45.1%) | |
| B | 56 (13.0%) | 22 (10.6%) | 34 (15.2%) | |
| C | 92 (21.3%) | 55 (26.6%) | 37 (16.5%) | |
| D | 151 (35.0%) | 99 (47.8%) | 52 (23.2%) | |
| Level | 0.002 | |||
| Paraplegia | 269 (62.4%) | 145 (70.0%) | 124 (55.4%) | |
| Tetraplegia | 162 (37.6%) | 62 (30.0%) | 100 (44.6%) | |
| Neurological level | <0.001 | |||
| C1-C4-ABC | 53 (12.3%) | 19 (9.2%) | 34 (15.2%) | |
| C5-C8-ABC | 42 (9.7%) | 11 (5.3%) | 31 (13.8%) | |
| T1-S5-ABC | 185 (42.9%) | 78 (37.7%) | 107 (47.8%) | |
| All D | 151 (35.0%) | 99 (47.8%) | 52 (23.2%) | |
| TSI | 47.5 (30.5–68.4) | 38.4 (28.4–56.5) | 56.1 (36.3–73.5) | <0.001 |
| TFIM at admission | 67.0 (53.0–89.0) | 73.0 (58.0–96.0) | 64.0 (50.0–81.0) | < 0.001 |
| MFIM at admission | 32.0 (18.5–54.0) | 38.0 (25.0–62.5) | 29.0 (16.0–46.0) | < 0.001 |
| LOS | 85.1 (56.0–115.0) | 75.1 (47.5–111.0) | 92.9 (67.2–117.0) | 0.002 |
| TFIM at discharge | 107.0 (85.0–114.0) | 109.0 (87.5–116.0) | 106.0 (77.7–113.0) | 0.039 |
| MFIM at discharge | 73.0 (51.0–79.0) | 74.0 (52.5–81.0) | 71.0 (43.0–78.0) | 0.035 |
| MFIM gain | 23.0 (9.0–37.0) | 19.0 (8.0–34.0) | 25.0 (12.0–41.0) | 0.008 |
| MFIM efficiency | 27.2 (10.6–45.7) | 24.3 (10.0–45.4) | 31.8 (12.6–46.4) | 0.237 |
| HADS Anxiety (discharge) | 3.0 (2.0–6.0) | 3.0 (2.0–7.0) | 3.0 (2.0–6.0) | 0.773 |
| HADS Depression (discharge) | 3.0 (1.0–5.0) | 3.0 (1.0–5.0) | 3.0 (1.0–5.0) | 0.784 |
| HADS Depression (discharge) categories | 0.249 | |||
| Normal 0–7 | 365 (86.5%) | 168 (83.6%) | 197 (89.1%) | |
| Borderline 8–10 | 31 (7.3%) | 18 (9.0%) | 13 (5.9%) | |
| Abnormal | 26 (6.2%) | 15 (7.5%) | 11 (5.0%) | |
| Educational level | 0.469 | |||
| Illiterate | 2 (0.5%) | 2 (1.0%) | 0 (0.0%) | |
| Read and write | 21 (4.9%) | 12 (5.8%) | 9 (4.0%) | |
| Primary | 241 (55.9%) | 113 (54.6%) | 128 (57.1%) | |
| Secondary | 110 (25.5%) | 55 (26.6%) | 55 (24.6%) | |
| University | 57 (13.2%) | 25 (12.1%) | 32 (14.3%) |
Notes: TSI: time since injury to inpatient rehabilitation admission; TFIM: total Functional Independence Measure; MFIM: motor Functional Independence Measure; HADS: Hospital Anxiety and Depression Scale; AIS: American Spinal Injury Association Impairment Scale; LOS: length of stay. NOTE: Categorical variables are represented as number of individuals and percentage n (%) and continuous variables as median (Q1–Q3).
Table 2 presents variables assessed by social workers, the proportion of patients living at discharge with their sentimental couple was significantly higher for patients with non-traumatic etiologies (75.4% vs 48.7%) as opposed to those living with their parents (7.7% vs 27.2%). The proportion of mobility and transport intervention (I8) was larger in patients with traumatic etiologies (28% vs 39.7%) as well as legal services intervention (I10) (8.7% vs 42.0%).
Table 2.
Social characteristics for traumatic vs non-traumatic included participants (n = 431).
| Total (N = 431) |
Non-traumatic (N = 207) |
Traumatic (N = 224) |
P | |
|---|---|---|---|---|
| Living situation at discharge | <0.001 | |||
| Sentimental Couple | 265 (61.5%) | 156 (75.4%) | 109 (48.7%) | |
| Parents | 77 (17.9%) | 16 (7.7%) | 61 (27.2%) | |
| Living alone | 39 (9.0%) | 15 (7.2%) | 24 (10.7%) | |
| Others | 50 (11.6%) | 20 (9.7%) | 30 (13.4%) | |
| Social Work intervention | ||||
| I1 | 1 (0.2%) | 0 (0.0%) | 1 (0.4%) | 0.336 |
| I2 | 417 (96.8%) | 200 (96.6%) | 217 (96.9%) | 0.881 |
| I3 | 290 (67.3%) | 143 (69.1%) | 147 (65.6%) | 0.445 |
| I4 | 411 (95.4%) | 198 (95.7%) | 213 (95.1%) | 0.781 |
| I5 | 430 (99.8%) | 207 (100.0%) | 223 (99.6%) | 0.336 |
| I6 | 430 (99.8%) | 207 (100.0%) | 223 (99.6%) | 0.336 |
| I8 | 147 (34.1%) | 58 (28.0%) | 89 (39.7%) | 0.010 |
| I9 | 51 (11.8%) | 21 (10.1%) | 30 (13.4%) | 0.297 |
| I10 | 112 (26.0%) | 18 (8.7%) | 94 (42.0%) | < 0.001 |
| CIQ home (0–10) | 5.7 (3.0) | 5.5 (2.8) | 5.9 (3.1) | 0.075 |
| CIQ social (0–10) | 7.1 (2.4) | 5.7 (3.0) | 7.5 (2.4) | <0.001 |
| CIQ productivity (0–5) | 0.9 (1.5) | 0.7 (1.4) | 1.0 (1.6) | 0.006 |
| CIQ total (0–25) | 13.8 (5.3) | 12.9 (4.9) | 14.6 (5.5) | <0.001 |
Notes: CIQ: Community Integration Questionnaire; Social work interventions: financial (I1), emotional and mental health support services (I2), coordination of care continuity (I3), education (I4), community service (I5), applications and reporting (I6), housing (I7), mobility and transport (I8), provision, repair, maintenance, and/or replacement of medical/orthopedic equipment (I9) and legal services, link to attorneys and law enforcement (I10). Social work intervention I7 was not included in the analysis because it was applied to 100% of patients. Categorical variables are represented as number of individuals and percentage n (%) and continuous variables as mean (SD)
Table 2 additionally presents mean CIQ values at 2.3(0.6) years post-discharge. CIQ score was significantly higher for traumatic patients than for the non-traumatic, this is further illustrated in Fig. 3 showing total CIQ boxplots every two years. In the supplementary material, Figures SM2 to SM5 detail the evolution of mean CIQ values from 2009 to 2023 by natural year, covering total CIQ and its components—home, social, and productivity.
Figure 3.
Total CIQ boxplots by year for all included participants (n = 431).
Predictors of community integration
Multiple linear regression analysis identified age, B AIS grade, educational level, time since injury to admission, I1, I8, I10, length of stay, HADS depression and total FIM at discharge as significant predictors of total CIQ score (Table 3). No multicollinearity was observed between the variables.
Table 3.
Multiple linear regression analysis of total CIQ score at 1–3 years after discharge (n = 431).
| Variables | Unadjusted coefficients | Standard error | Adjusted coefficients | 95% CI | VIF | P |
|---|---|---|---|---|---|---|
| Intercept | 17.605 | 6.375 | (5.071, 30.139) | 0.006 | ||
| Sex (Male) | 0.426 | 0.465 | 0.037 | (−0.488, 1.340) | 1.183 | 0.360 |
| Age | −0.082 | 0.016 | −0.267 | (−0.114, −0.050) | 1.950 | <0.001 |
| Etiology (Traumatic) | 0.910 | 0.501 | 0.085 | (−0.075, 1.896) | 1.584 | 0.070 |
| Level (Tetraplegia) | −0.309 | 0.551 | −0.028 | (−1.394, 0.774) | 1.795 | 0.574 |
| AIS-A | 0.911 | 0.697 | 0.078 | (−0.459, 2.283) | 1.608 | 0.192 |
| AIS-B | 1.627 | 0.773 | 0.101 | (0.105, 3.148) | 1.687 | 0.036 |
| AIS-C | 0.091 | 0.620 | 0.006 | (−1.127, 1.309) | 1.628 | 0.883 |
| Educational level (low) | −3.275 | 0.614 | −0.299 | (−4.482, −2.067) | 1.267 | <0.001 |
| Educational level (mid) | −1.965 | 0.706 | −0.160 | (−3.354, −0.576) | 1.400 | 0.005 |
| Living with couple | −0.809 | 0.490 | −0.073 | (−1.773, 0.155) | 1.436 | 0.099 |
| TSI | −0.018 | 0.008 | −0.082 | (−0.036, −0.001) | 1.111 | 0.035 |
| I1 | −8.439 | 4.225 | −0.076 | (−16.747, −0.132) | 1.065 | 0.046 |
| I2 | −1.149 | 1.194 | −0.038 | (−3.498, 1.199) | 1.156 | 0.336 |
| I3 | −0.354 | 0.458 | −0.031 | (−1.255, 0.546) | 1.168 | 0.439 |
| I4 | −0.459 | 1.018 | −0.017 | (−2.461, 1.542) | 1.069 | 0.652 |
| I5 | 0.281 | 4.322 | 0.002 | (−8.215, 8.777) | 1.115 | 0.948 |
| I6 | −1.072 | 4.170 | −0.009 | (−9.270, 7.126) | 1.038 | 0.797 |
| I8 | 1.045 | 0.451 | 0.093 | (0.157, 1.933) | 1.165 | 0.021 |
| I9 | 1.200 | 0.643 | 0.073 | (−0.064, 2.465) | 1.110 | 0.062 |
| I10 | −1.189 | 0.540 | −0.098 | (−2.251, −0.128) | 1.427 | 0.028 |
| LOS | −0.014 | 0.005 | −0.105 | (−0.024, −0.003) | 1.268 | 0.012 |
| HADS Depression | −0.293 | 0.058 | −0.190 | (−0.408, −0.177) | 1.049 | <0.001 |
| TFIM at discharge | 0.080 | 0.012 | 0.336 | (0.056, 0.104) | 1.836 | <0.001 |
| Adjusted R2 = 41.4; R2 = 44.6 | ||||||
Notes: CIQ: Community Integration Questionnaire; AIS: American Spinal Injury Association Impairment Scale; TFIM: Total Functional Independence Measure; HADS: Hospital Anxiety and Depression Scale; TSI: time since injury to inpatient rehabilitation admission; LOS: length of stay; VIF variance inflation factor. Social work interventions: financial (I1), emotional and mental health support services (I2), coordination of care continuity (I3), education (I4), community service (I5), applications and reporting (I6), housing (I7), mobility and transport (I8), provision, repair, maintenance, and/or replacement of medical/orthopedic equipment (I9) and legal services, link to attorneys and law enforcement (I10). Reference variables: Sex (female), Etiology (non-traumatic), Level (paraplegia), AIS (D), Educational level (university degree).
In the supplementary material, Table SM2 presents a power analysis demonstrating that our sample size achieves a medium effect size with an alpha error of 0.05 and a power of 0.95.
Multiple logistic regression identified age, traumatic etiology, educational level, I8, length of stay, HADS depression and total FIM at discharge as significant predictors of CIQ score > 15.1 points (Table 4). Using this cutoff, participants were categorized into two distinct groups: 40% having high community integration (CIQ score > 15.1), while 60% fell into the low community integration category (CIQ score ≤ 15.1). Our model yielded an AUC (95% CI): 0.823 (0.7516-0.8961). ROC curve is presented in Supplementary material Figure SM7.
Table 4.
Multiple logistic regression analysis of total CIQ score > 15.1 points (n = 431).
| OR (95%CI) | P | |
|---|---|---|
| Sex (Male) | 0.924 (0.540, 1.578) | 0.773 |
| Age | 0.969 (0.952, 0.986) | <0.001 |
| Etiology (Traumatic) | 1.757 (1.003, 3.104) | 0.049 |
| AIS-A | 1.204 (0.664, 2.179) | 0.538 |
| Educational level (low) | 0.195 (0.089, 0.409) | <0.001 |
| Educational level (mid) | 0.334 (0.139, 0.768) | 0.011 |
| Living with couple | 0.683 (0.391, 1.190) | 0.179 |
| I2 | 0.336 (0.067, 1.259) | 0.129 |
| I3 | 0.661 (0.389, 1.119) | 0.123 |
| I4 | 0.600 (0.168, 2.021) | 0.415 |
| I8 | 1.770 (1.061, 2.972) | 0.029 |
| I9 | 1.713 (0.822, 3.586) | 0.149 |
| I10 | 0.666 (0.355, 1.238) | 0.201 |
| LOS | 0.992 (0.986, 0.998) | 0.018 |
| HADS Depression | 0.856 (0.792, 0.919) | <0.001 |
| TFIM at discharge | 1.035 (1.023, 1.047) | <0.001 |
| AUC (95% CI): 0.823 (0.7516–0.8961) | ||
| Sensitivity: 0.758 | ||
| Specificity: 0.733 | ||
| Precision: 0.763 | ||
| Recall: 0.781 | ||
| F: 0.763 |
Notes: American Spinal Injury Association Impairment Scale; TFIM: Total Functional Independence Measure; HADS: Hospital Anxiety and Depression Scale; TSI: time since injury to inpatient rehabilitation admission; LOS: length of stay; VIF variance inflation factor. Social work interventions: financial (I1), emotional and mental health support services (I2), coordination of care continuity (I3), education (I4), community service (I5), applications and reporting (I6), housing (I7), mobility and transport (I8), provision, repair, maintenance, and/or replacement of medical/orthopedic equipment (I9) and legal services, link to attorneys and law enforcement (I10).
Causal mediators of community integration
We constructed individual single mediator models for each of the hypothesized mediators (motor FIM at discharge, motor FIM efficiency, HADS at discharge and social work interventions I8 and I10) using the one at a time mediator approach as presented in Table 5. All causal models were adjusted for age, sex, AIS, etiology, FIM at admission and time since injury to admission (Fig. 2). We used 1000 bootstrapped simulations to generate 95% confidence intervals. Both motor FIM models in Table 5 show a significant mediation effect (P < 0.05 for ACME, ADE and Total effect), meanwhile no mediation effect was observed for HADS and social work interventions. Details on the construction of the mediator models are presented in Supplementary material.
Table 5.
Effect decomposition for each single-mediator model on total CIQ outcome using the one moderator at a time approach.
| Variable | Measures | Estimates | 95% CI | P |
|---|---|---|---|---|
| mFIM at discharge | ACME | −0.005 | (−0.010, 0.000) | 0.040 |
| ADE | −0.015 | (−0.025, 0.000) | 0.014 | |
| Total Effect | −0.020 | (−0.032, −0.010) | 0.002 | |
| Prop. Mediated | 0.258 | (0.023, 0.650) | 0.042 | |
| mFIM efficiency | ACME | −0.006 | (−0.010, 0.000) | 0.004 |
| ADE | −0.013 | (−0.027, 0.000) | 0.022 | |
| Total Effect | −0.020 | (−0.033, −0.010) | <0.001 | |
| Prop. Mediated | 0.315 | (0.102, 0.820) | 0.004 | |
| HADS depression at discharge | ACME | −0.00037 | (−0.0038, 0.00) | 0.746 |
| ADE | −0.0199 | (−0.0324, −0.010) | <0.001 | |
| Total Effect | −0.0203 | (−0.0328, −0.010) | 0.002 | |
| Prop. Mediated | 0.01817 | (−0.2286, 0.200) | 0.744 | |
| I8 | ACME | 0.0061 | (−0.047, 0.050) | 0.893 |
| ADE | −0.020 | (−0.033, −0.010) | <0.001 | |
| Total Effect | −0.0144 | (−0.0713, 0.030) | 0.33 | |
| Prop. Mediated | −0.428 | (−7.592, 6.360) | 0.67 | |
| I10 | ACME | 0.0065 | (−0.0678, 0.070) | 0.988 |
| ADE | −0.0187 | (−0.031, −0.001) | 0.002 | |
| Total Effect | −0.0122 | (−0.093, 0.060) | 0.514 | |
| Prop. Mediated | −0.5367 | (−7.710, 10.341) | 0.492 |
Note: mFIM: motor Functional Independence Measure; ACME: average causal mediation effect; ADE: average direct effect; I8: mobility and transport; I10: legal services, link to attorneys and law enforcement.
Figure 2.
Directed acyclic graph to describe causal assumptions required to avoid confounding and selection bias.
Discussion
In this study we investigated predictive factors of community integration addressed during inpatient rehabilitation (e.g. functional independence, anxiety, depression, social work interventions) and identified causal mediators of community integration, for individuals with traumatic or non-traumatic etiologies, adding to the scarce literature.
In relation to our first objective, analyzing baseline demographic and clinical characteristics of the included patients, our results are consistent with recent traumatic SCI epidemiological results reporting on Spanish population.
In our case injuries were dated between 2007 and 2021, Barbara-Bataller and colleagues analyzed the 2001–2015 period (36), our study includes 76.1% of patients whose date of injury was between 2007 and 2015. Mean age at injury was 42.8 (17.91) years, 80.1% male, being traffic accidents the main cause of injury (36.5%), complete injuries (AIS A) were the most frequent (52.1%) and 23.2% AIS D. In our case mean age at injury was 43.5 (17.1) years, 80.8% male, traffic accidents were the main cause of injury (39.3%), complete injuries (AIS A) were the most frequent (45.1%) and 23.8% AIS D.
Non-traumatic epidemiological studies are scarce, nevertheless our results were also similar to previous reports from Spanish population (37). Van den Berg and cols reported mean age at injury was 51.7(19.3) years, in our case it was 56.1 (14.7), 53% male and 55% in our case, also the most common causes of injury were similar (tumors, vascular).
Community Integration has been reported as an essential component of the entire rehabilitation process especially for traumatic SCI since these mostly occur among healthy working age adults who have an active family and social life (38). Nevertheless, in our case participants with traumatic SCI consistently reported higher values of total CIQ when compared to non-traumatic SCI (Table 2, Fig. 3), suggesting that non-traumatic SCI may require additional support.
In relation to objective 2, our results add to the scarce literature on predictors of community integration. Ahmed and colleagues reported paraplegia and sex (with females 2.5 times higher community integration score) as significant predictors (39).
According to our results, when analyzing predictors of total CIQ (Table 3), total FIM at discharge was found significant, for every 22.8 points on FIM score an extra 1.8 points on total CIQ score are obtained. In relation to age at injury, according to our results every 17.2 years of age at injury, represent a decrease of 1.41 points in total CIQ. Ahmed et al. reported age as a statistically significant positive predictor; “an increase of age by a year increases total community integration by 0.09” (39), in our case age was a negative predictor. A possible explanation of this can be that their mean age was remarkably younger, 34 years meanwhile it was 49 in our case.
Regarding psychological factors, every 3.5 points on HADS depression at discharge represent a decrease of one point in total CIQ (Table 3). In relation to educational level a shift from our reference group (high education, i.e. university) to low education (< six years) represents a 30% decrease in total CIQ score. The decrease is 16% when the shift is from high to mid education (<12 years). Regarding AIS grade, a shift from our reference group (AIS D) to B represents a 10% increase in total CIQ score.
Logistic regression (Table 4) shows that each additional year of age at injury is associated with an 3.1% decrease in the odds of achieving “good” community integration (total CIQ score higher than 15.1 points), meanwhile each additional point in total FIM at discharge is associated with an 3.5% increase and each additional point in HADS depression at discharge is associated with a 14.4% decrease.
In relation to educational level, an individual with low education experiences a reduction of 80% in the odds of achieving “good” community integration, compared to an individual with high education, in the case of individuals with mid education such reduction is 67%.
We identified a social work intervention (I8, mobility, transportation, e.g. provision of transportation fares and passes; linkage with transportation services) as a significant predictor of good community integration. An increase of 77% is experienced by those patients receiving I8 (Table 4). This is confirmed by our linear regression model, receiving such intervention represents a 9.3% increase in total CIQ score (Table 3). A larger proportion of patients with traumatic etiology received I8 when compared to non-traumatic (39.7% vs 28.0%, P = 0.01, as presented in Table 2). Social work interventions in transportation for individuals with SCI typically include coordinating accessible transportation options such as specialized transit services and subsidized fares. Social workers facilitate non-emergency medical transportation for regular healthcare access and advocate for infrastructural improvements to enhance public transportation accessibility. Additionally, for younger patients with more complete injuries, who often require more intensive transportation support, interventions may include assistance with obtaining adaptive driving equipment and training for vehicle modification. These patients may also benefit from programs that coordinate travel training to teach independent public transportation use, enhancing their mobility and overall independence. Furthermore, social workers may arrange for recreational transport services that enable participation in community and social activities, which are particularly important for the social reintegration of younger individuals (40,41). Although the application of mediation analyses (our objective 3) in health-related research is increasing (35, 42) we found no previous research on the causal effects of rehabilitation interventions in community integration after SCI. In this regard our results can be considered as an initial step, we identified motor FIM at discharge and motor FIM efficiency as significant causal mediators.
One of the main limitations of this study are potential biases inherent in its retrospective design, besides, it was conducted in a single-centre, and although this ensured the uniformity of data collection, this limited the generalizability of the results. Further, given that this was a 17-year study, management of these injuries has evolved over time, with changes in the patient pathways continually moving with the demands and resources available.
Additionally, previous reports indicate that social relationships and informal networks significantly enhance participation and help manage secondary health conditions and life with SCI (43, 44). Such informal social support and networking was not included in our analysis as well as other aspects of environment (e.g. access to technology, policy, community characteristics). People with SCI also differ based on the frequency and magnitude of environmental barriers that they face in the community (43,45).
Nevertheless, a recent scoping review concludes that “there is lack of clarity on the impact of injury-related factors on community reintegration” (44). We focused on such clinical factors addressed during inpatient rehabilitation as well as those factors that can be targeted by social workers when they are in direct contact with inpatients and their informal caregivers.
Conclusions
This study identified associations between clinical and demographic variables, rehabilitation outcomes, and social work interventions for patients admitted to rehabilitation within three months of sustaining a traumatic or non-traumatic spinal cord injury, with community integration assessed two years post-discharge. We found that aspects such as functional independence, levels of depression, and targeted social work interventions during rehabilitation are closely linked to community reintegration. Future research should explore how these predictors evolve over time and investigate differences between traumatic and non-traumatic injuries using interaction terms or stratified models.
Supplementary Material
Acknowledgements
Special thanks to Olga Araujo from Institut Guttmann – Documentation department for her continuous support in accessing related literature.
Disclaimer statements
Funding None
Disclosure statement No potential conflict of interest was reported by the author(s).
Equal contributions Authors Alejandro García-Rudolph and Hector Cusso equally contributed as first authors to this manuscript
Supplemental data
Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/10790268.2024.2386738.
References
- 1.Ko HY. Current epidemiology of spinal cord injuries. In: Management and rehabilitation of spinal cord injuries. 2nd ed. Singapore: Springer; 2022. p. 1–17. doi: 10.1007/978-981-19-0228-4_12. [DOI] [Google Scholar]
- 2.Harvey LA. Physiotherapy rehabilitation for people with spinal cord injuries. J Physiother. 2016 Jan;62(1):4–11. doi: 10.1016/j.jphys.2015.11.004. [DOI] [PubMed] [Google Scholar]
- 3.Cieza A, Fayed N, Bickenbach J, Prodinger B.. Refinements of the ICF linking rules to strengthen their potential for establishing comparability of health information. Disabil Rehabil. 2019 Mar;41(5):574–583. doi: 10.3109/09638288.2016.1145258. [DOI] [PubMed] [Google Scholar]
- 4.Barclay L, Robins L, Migliorini C, Lalor A.. Community integration programs and interventions for people with spinal cord injury: a scoping review. Disabil Rehabil. 2021 Dec;43(26):3845–3855. doi: 10.1080/09638288.2020.1749889. [DOI] [PubMed] [Google Scholar]
- 5.Willer BS, Rosenthal M, Kreutzer JS, Gordon WA, Rempel RG.. Assessment of community integration following rehabilitation for traumatic brain injury. J Head Trauma Rehabil. 1993;8:75–87. [Google Scholar]
- 6.Unger J, Singh H, Mansfield A, Hitzig SL, Lenton E, Musselman KE.. The experiences of physical rehabilitation in individuals with spinal cord injuries: a qualitative thematic synthesis. Disabil Rehabil. 2019 Jun;41(12):1367–1383. doi: 10.1080/09638288.2018.1425745. [DOI] [PubMed] [Google Scholar]
- 7.Craig A, Nicholson Perry K, Guest R, Tran Y, Middleton J.. Adjustment following chronic spinal cord injury: determining factors that contribute to social participation. Br J Health Psychol. 2015 Nov;20(4):807–823. doi: 10.1111/bjhp.12143. [DOI] [PubMed] [Google Scholar]
- 8.Burns AS, Santos A, Cheng CL, Chan E, Fallah N, Atkins D, Dvorak MF, Ho C, Ahn H, Paquet J, Kwon BK, et al. Understanding length of stay after spinal cord injury: insights and limitations from the access to care and timing project. J Neurotrauma. 2017;34(20):2910–2916. doi: 10.1089/neu.2016.4935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Barclay L, Lalor A, Migliorini C, Robins L.. A comparative examination of models of service delivery intended to support community integration in the immediate period following inpatient rehabilitation for spinal cord injury. Spinal Cord. 2020 May;58(5):528–536. doi: 10.1038/s41393-019-0394-x. [DOI] [PubMed] [Google Scholar]
- 10.Walsh B, Smith S, Wren MA, Eighan J, Lyons S.. The impact of inpatient bed capacity on length of stay. Eur J Health Econ. 2022 Apr;23(3):499–510. doi: 10.1007/s10198-021-01373-2. Epub 2021 Sep 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Backus D, Gassaway J, Smout RJ, Hsieh CH, Heinemann AW, DeJong G, Horn SD.. Relation between inpatient and postdischarge services and outcomes 1 year postinjury in people with traumatic spinal cord injury. Arch Phys Med Rehabil. 2013 Apr;94(4 Suppl):S165–S174. doi: 10.1016/j.apmr.2013.01.012. [DOI] [PubMed] [Google Scholar]
- 12.Müller-Jensen L, Ploner CJ, Kroneberg D, Schmidt WU.. Clinical presentation and causes of non-traumatic spinal cord injury: an observational study in emergency patients. Front Neurol. 2021 Aug 9;12:701927. doi: 10.3389/fneur.2021.701927. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Alito A, Filardi V, Famà F, Bruschetta D, Ruggeri C, Basile G, Stancanelli L, D'Amico C, Bianconi S, Tisano A.. Traumatic and non-traumatic spinal cord injury: demographic characteristics, neurological and functional outcomes. A 7-year single centre experience. J Orthop. 2021 Nov 20;28:62–66. doi: 10.1016/j.jor.2021.11.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Vandenbroucke JP, von Elm E, Altman DG, Gøtzsche PC, Mulrow CD, Pocock SJ, Poole C, Schlesselman JJ, Egger M, et al. Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. PLoS Med. 2007 Oct 16;4(10):e297. doi: 10.1371/journal.pmed.0040297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Biering-Sørensen F, DeVivo MJ, Charlifue S, Chen Y, New PW, Noonan V, Post MW, Vogel L.. International spinal cord injury core data set (version 2.0)-including standardization of reporting. Spinal Cord. 2017;55:759–764. [DOI] [PubMed] [Google Scholar]
- 16.Linacre JM, Heinemann AW, Wright BD, Granger CV, Hamilton BB.. The structure and stability of the functional independence measure. Arch Phys Med Rehabil. 1994 Feb;75(2):127–132. [PubMed] [Google Scholar]
- 17.Loni E, Moein S, Bidhendi-Yarandi R, Akbarfahimi N, Layeghi F.. Changes in functional independence after inpatient rehabilitation in patients with spinal cord injury: a simultaneous evaluation of prognostic factors. J Spinal Cord Med. 2024;47(3):369–378. doi: 10.1080/10790268.2022.2064264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Tokunaga M, Higashi T, Inoue R, Ohkubo T, Watanabe S.. Comparison of functional independence measure (FIM) gain between groups with extreme differences in FIM score at admission. Int J Phys Med Rehabil. 2017;5:402. doi: 10.4172/2329-9096.1000402. [DOI] [Google Scholar]
- 19.Kee KM, Mohamad NZ, Koh PPW, Yeo JPT, Ng YS, Kam JC, Asano M.. Return to work after spinal cord injury: a Singaporean pilot community-based rehabilitation program. Spinal Cord. 2020 Oct;58(10):1096–1103. doi: 10.1038/s41393-020-0459-x. [DOI] [PubMed] [Google Scholar]
- 20.Gontovsky S, Russum P, Stokie D.. Comparison of the CIQ and CHART short form in assessing community integration in individuals with chronic spinal cord injury: a pilot study. NeuroRehabilitation. 2009;24:185–192. doi: 10.3233/NRE-2009-0467. [DOI] [PubMed] [Google Scholar]
- 21.Kratz A, Chadd E, Jensen M, Kehn M, Kroll T.. An examination of the psychometric properties of the community integration questionnaire (CIQ) in spinal cord injury. J Spinal Cord Inj. 2015;38:446–455. doi: 10.1179/2045772313Y.0000000182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chhabra HS, Bhalla AM.. Influence of socio-economic status on access to different components of SCI management across Indian population. Spinal Cord. 2015 Nov;53(11):816–820. doi: 10.1038/sc.2015.80. [DOI] [PubMed] [Google Scholar]
- 23.Callaway L, Enticott J, Farnworth L, McDonald R, Migliorini C, Willer B.. Community integration outcomes of people with spinal cord injury and multiple matched controls: a pilot study. Aust Occup Ther J. 2017 Jun;64(3):226–234. doi: 10.1111/1440-1630.12350. [DOI] [PubMed] [Google Scholar]
- 24.Zigmond AS, Snaith RP.. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983 Jun;67(6):361–370. doi: 10.1111/j.1600-0447.1983.tb09716.x. [DOI] [PubMed] [Google Scholar]
- 25.Munce SE, Straus SE, Fehlings MG, Voth J, Nugaeva N, Jang E, Webster F, Jaglal SB.. Impact of psychological characteristics in self-management in individuals with traumatic spinal cord injury. Spinal Cord. 2016;54(1):29–33. doi: 10.1038/sc.2015.91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Russell HF, Richardson EJ, Bombardier CH, Dixon TM, Huston TA, Rose J, Sheaffer D, Smith SA, Ullrich PM.. Professional standards of practice for psychologists, social workers, and counselors in SCI rehabilitation. J Spinal Cord Med. 2016;39(2):127–145. doi: 10.1080/10790268.2015.1119966. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ruth BJ, Wachman MK, Marshall JW, Backman AR, Harrington CB, Schultz NS, Ouimet KJ.. Health in all social work programs: findings from a US national analysis. Am J Public Health. 2017 Dec;107(S3):S267–S273. doi: 10.2105/AJPH.2017.304034. Erratum in: Am J Public Health. 2018 Apr;108(4):e13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bako AT, Taylor HL, Wiley K Jr., Zheng J, Walter-McCabe H, Kasthurirathne SN, Vest JR.. Using natural language processing to classify social work interventions. Am J Manag Care. 2021 Jan 1;27(1):e24–e31. doi: 10.37765/ajmc.2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.R Core Team . R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2023; https://www.R-project.org/. [Google Scholar]
- 30.O’Brien RM. A caution regarding rules of thumb for variance inflation factors. Qual Quant. 2007;41(5):673–690. doi: 10.1007/s11135-006-9018-6. [DOI] [Google Scholar]
- 31.Kuhn M. Building predictive models in R using the caret package. J Stat Softw. 2008;28(5):1–26. doi: 10.18637/jss.v028.i05.27774042 [DOI] [Google Scholar]
- 32.Imai K, Keele L, Tingley D.. A general approach to causal mediation analysis. Psychol Methods. 2010;15(4):309–334. [DOI] [PubMed] [Google Scholar]
- 33.Tingley D, Yamamoto T, Hirose K, Keele L, Imai K.. Mediation: R package for causal mediation analysis. J Stat Softw. 2014;59(5):1–38. doi: 10.18637/jss.v059.i05.26917999 [DOI] [Google Scholar]
- 34.Yimer BB, Lunt M, Beasley M, Macfarlane GJ, McBeth J.. Bayesgmed: An R-package for Bayesian causal mediation analysis. PLoS One. 2023;18(6):e0287037. doi: 10.1371/journal.pone.0287037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Cashin AG, Lee H, Traeger AC, Moseley GL, Hübscher M, Kamper SJ, Skinner IW, McAuley JH.. Producing clinically meaningful reductions in disability: a causal mediation analysis of a patient education intervention. J Pain. 2022 Feb;23(2):236–247. doi: 10.1016/j.jpain.2021.07.007. [DOI] [PubMed] [Google Scholar]
- 36.Bárbara-Bataller E, Méndez-Suárez JL, Alemán-Sánchez C, Sánchez-Enríquez J, Sosa-Henríquez M.. Change in the profile of traumatic spinal cord injury over 15 years in Spain. Scand J Trauma Resusc Emerg Med. 2018 Apr 5;26(1):27. doi: 10.1186/s13049-018-0491-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.van den Berg ME, Castellote JM, Mahillo-Fernandez I, de Pedro-Cuesta J.. Incidence of nontraumatic spinal cord injury: a Spanish cohort study (1972–2008). Arch Phys Med Rehabil. 2012 Feb;93(2):325–331. doi: 10.1016/j.apmr.2011.08.027. Erratum in: Arch Phys Med Rehabil. 2012 Mar;93(3):550. van den Berg, Maayken EE [corrected to van den Berg, Maayken EL]. [DOI] [PubMed] [Google Scholar]
- 38.Sekaran P, Vijayakumari F, Hariharan R, Zachariah K, Joseph SE, Kumar RK.. Community reintegration of spinal cord-injured patients in rural south India. Spinal Cord. 2010 Aug;48(8):628–632. doi: 10.1038/sc.2010.6. [DOI] [PubMed] [Google Scholar]
- 39.Ahmed N, Quadir MM, Rahman MA, Alamgir H.. Community integration and life satisfaction among individuals with spinal cord injury living in the community after receiving institutional care in Bangladesh. Disabil Rehabil. 2018 May;40(9):1033–1040. doi: 10.1080/09638288.2017.1283713. [DOI] [PubMed] [Google Scholar]
- 40.Ferdiana A, Post MWM, Bültmann U, van der Klink JJL.. Barriers and facilitators for work and social participation among individuals with spinal cord injury in Indonesia. Spinal Cord. 2021 Oct;59(10):1079–1087. doi: 10.1038/s41393-021-00624-6. [DOI] [PubMed] [Google Scholar]
- 41.Gross-Hemmi MH, Fekete C, Post MWM, Scheel-Sailer A, Schwegler U, Brinkhof MWG.. Detecting subgroups in social participation among individuals living with spinal cord injury: a longitudinal analysis of community survey data. Spinal Cord. 2021 Apr;59(4):419–428. doi: 10.1038/s41393-020-00576-3. [DOI] [PubMed] [Google Scholar]
- 42.Nguyen TQ, Schmid I, Stuart EA.. Clarifying causal mediation analysis for the applied researcher. Defining effects based on what we want to learn. Psychol Methods. 2021;26:255–271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.LaVela SL, Landers K, Etingen B, Karalius VP, Miskevics S.. Factors related to caregiving for individuals with spinal cord injury compared to caregiving for individuals with other neurologic conditions. J Spinal Cord Med. 2015 Jul;38(4):505–514. doi: 10.1179/2045772314Y.0000000240. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Gupta S, Jaiswal A, Norman K, DePaul V.. Heterogeneity and its impact on rehabilitation outcomes and interventions for community reintegration in people with spinal cord injuries: an integrative review. Top Spinal Cord Inj Rehabil. 2019 Spring;25(2):164–185. doi: 10.1310/sci2502-164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Barclay L, Mcdonald R, Lentin P, Bourke-Taylor H.. Facilitators and barriers to social and community participation following spinal cord injury. Aust Occup Ther J. 2016;63(1):19–28. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.