Rehospitalization and Healthcare Utilization after Initial Rehabilitation in SCI

Persons with SCI are at greater risk for numerous secondary health complications than the general population and therefore are at far greater likelihood of being admitted to hospital or seeking medical care for one reason or another. At least some of the causes for these admissions or other forms of healthcare utilization have been deemed as preventable (e.g., pressure sores, UTIs) and therefore there has been much interest in understanding the patterns and antecedents for rehospitalization/healthcare utilization so as to inform effective preventative strategies.

Table 15: Individual Studies – Rehospitalization and Healthcare Utilization



Of the nine papers reviewed across six distinct jurisdictions (i.e. Australia, Canada, Italy, Turkey, UK, USA), differences in methods of calculating readmission rates and specific inclusion criteria made comparisons tenuous at best. Regardless, it is apparent that hospital re-admission is a very significant issue across all regions with universally high re-admission rates (Savic et al. 2000; Cardenas et al. 2004; Charlifue et al. 2004; Dryden et al. 2004; Franceschini et al. 2003; Jaglal et al. 2009; Middleton et al. 2004; Paker et al. 2006; Dorsett and Geraghty 2008). Cardenas et al. noted an average rehospitalization rate of 55% (defined as the number of patients rehospitalized within a particular anniversary of injury year) for the first year post injury and then rates of 36-38% for subsequent anniversary years from 5-20 years post injury. This analysis was conducted on the large multi-centre US model systems dataset (n=8668) between 1995-2002. This was very similar to the rates reported by Charlifue et al. (2004) which was not surprising as she had examined the same database, albeit, over different years (1973-1998).

The only other high-quality, population-based data on which to base a comparison exists for the jurisdiction of Ontario, Canada. Jaglal et al. (2009) defined rehospitalizations over the first year after initial rehabilitation discharge, thereby circumventing the primary limitation of most other studies associated with a variable follow-up period. Multiple administrative healthcare databases were linked to overcome the other common limitation inherent in several other studies, that being the variances which may occur with participant self-report. These authors reported a rehospitalization rate of 27.5% – approximately half that reported in the US. This appears to be similar to the rates reported over a somewhat similar time period in Queensland, Australia (n=46) by Dorsett and Geraghty (2008) as participants reported rehospitalization rates of ~18% from 0-6 months post-discharge (estimated from graph), ~25% from 6-12 months, ~31% for year 2, ~18% for year 3 and ~38% for year 10. Overall cumulative rehospitalization rates were reported at 32.6% over the first 2 years and 52% by the 10th year. Middleton et al. (2004) reported slightly higher 10 year (i.e., cumulative) rehospitalization rates for the jurisdiction of New South Wales, Australia (n=432) with 58.6% of persons with SCI being rehospitalized due to a SCI-related issue and an additional 10.8% being admitted to hospital for a non-SCI-related issue. Another report indicated an overall re-admission rate of 64% over 6 years involving 3 longitudinal interviews of community dwelling persons (n=198) with a mean of 33 years injury duration associated with two large SCI specialist centres in the UK (Savic et al. 2000).

One trend that can be gleaned from these reports is that the rehospitalization rates generally decline following the initial year or two post-discharge (Cardenas et al. 2004; Charlifue et al. 2004; Middleton et al. 2004). Regardless, in the context of informing initial rehabilitation practice, rehospitalization rates in the first year post-discharge are of particular importance and the data associated with the largest and highest-quality studies demonstrate a higher rate in the US versus other jurisdictions (i.e., Australia, Canada). It is difficult to speculate on why this may be the case given the variation between these countries in terms of health care and social systems although one suggestion has been that the high rehospitalization rate may be linked to a shortened rehabilitation stay, especially present in the US (Cardenas et al. 2004). It is certainly the case that the US has the shortest rehabilitation LOS than any other jurisdiction reporting data (See Section 4.2).

There is reasonable agreement for the primary reasons for hospital readmission following initial SCI inpatient rehabilitation across most studies (Cardenas et al. 2004; Dorsett and Geraghty 2008; Dryden et al. 2004; Franceschini et al. 2003; Jaglal et al. 2009; Middleton et al. 2004; Paker et al. 2006; Savic et al. 2000). All reports included issues associated with skin (e.g., pressure ulcers) and the genitourinary system (e.g., UTIs and to a lesser extent complications of the upper urinary tract) as among the highest reasons for readmission. Other issues that were associated with significant rates of readmission included diseases of the respiratory system (e.g., infections, especially in persons with tetraplegia), musculoskeletal complaints (e.g., spasticity, pain) and digestive system problems (e.g., bowel). Of note, musculoskeletal issues were found to be most prominent as a cause of readmission in the report by Jaglal et al. (2009) than any other issue. It should be noted that although readmission rates were significant due to pressure ulcers, when considering the subsequent length of stay often associated with this specific complication, the impact of pressure ulcers are even more consequential (Savic et al. 2000; Middleton et al. 2004).

Cardenas et al. (2004) conducted multivariate logistic regression on the large US Model Systems dataset and determined that motor FIM™ scores at discharge and the payer were the two most significant predictors of rehospitalization within the first year (i.e., those with lower motor score state or federal funded persons vs those with private insurance were more likely to be hospitalized). Payer, motor FIM™ and race were also noted as predictors of readmission at later points in time. A similar analysis was conducted by Jaglal et al. (2009) and the factors most significantly associated with rehospitalization in the first year were longer length of rehabilitation stay, rural residence, ≥50 outpatient physician visits and ≥50 specialist visits following the initial admission. Odds ratios for individual factors associated with the highest likelihood of being rehospitalized included: Total physician visits ≥ 50 (OR=3.69), Total specialist visits ≥ 50 (OR=2.95), rural residence (OR=1.94), presence of comorbidities with Charlson score ≥ 3 (OR=2.08), >70 years old (OR=1.72). Charlifue et al. (2004) noted that both the number and length of rehospitalizations were predicted by being older at injury, being unmarried, having an indwelling catheter, having a more severe SCI and having been hospitalized 5 years earlier.

Healthcare Utilization

Similar to that evident with hospital readmissions, it is apparent that persons with SCI utilize other aspects of the healthcare system more frequently than most other persons, especially in the first year following rehabilitation discharge. Three Canadian studies from two separate jurisdictions (i.e., provinces of Alberta and Ontario) determined the rates of physician contacts for persons returning to the community following initial rehabilitation. Guilcher et al. (2010) and Munce et al. (2009) examined the linked results from several province-wide (Ontario) administrative healthcare databases to investigate differences in the number of physician contacts in the first year following rehabilitation associated with etiology (i.e., nontrauma / trauma) or gender respectively. There were no significant differences due to etiology with similar numbers of overall physician visits for those with nontraumatic vs traumatic SCI (31.2 vs 29.7 respectively), however there were differences in the types of physicians seen between the 2 groups (Guilcher et al. 2010). Women with SCI had significantly more physician visits than men in the first year following discharge (37.0 vs 30.0) although they were more likely to visit their family physician, whereas men had significantly more visits to their physiatrist (Munce et al. 2009). Some of the individual factors associated with a greater likelihood of having more physician visits included age, lower function (i..e., lower FIM scores), direct discharge to a chronic care / other rehabilitation facility, urban vs rural residence or the presence of comorbidities / prior (in-hospital) complications (Munce et al. 2009; Guilcher et al. 2010). Dryden et al. (2004) used similar methodologies in another Canadian province (i.e., Alberta) and found a median number of physician contacts of 22.0 in the first year and this declined dramatically to 8.0 visits by year 2 and to 4.0 visits by year 6. In all cases, control subjects identified in the overall health registry and matched by age, gender and geographic region had significantly fewer physician contacts for each year (median = 3.0 visits).


There is level 4 evidence that at least 25% of persons with SCI (moreso in some jurisdictions including the US) may expect a hospital readmission in the first year following discharge from SCI rehabilitation.

There is level 4 evidence from three studies that hospital re-admission rates are highest in the first year post injury and then stabilize at a still significantly high rate.

There is level 4 evidence from eight studies that urinary problems (UTIs), pressure ulcers, respiratory infections and musculoskeletal problems are consistently among the most frequent causes of hospital readmission among persons with SCI.

There is level 4 evidence from three studies that factors such as increased age, lower function / greater severity of injury, prior contact with the health system, funding, rural habitation and being unmarried are associated with a greater chance of a hospital readmisssion.

There is level 3 evidence from 1 study and supported by two level 4 studies that persons with SCI have an increased number of physician contacts as compared to matched controls from the general population, especially moreso in the first year post-injury.

  • Hospital readmission occurs frequently for persons with SCI (especially within the first year post-injury), with UTIs, pressure ulcers, respiratory infections and musculoskeletal problems among the most frequent causes.
  • Persons with SCI have more physician contacts than the general population, especially more so in the first year post-injury.