See All Evidence Sections
Rehabilitation Practices

Specialized vs General SCI Units (Acute Care)

Donovan et al. (1984) contend that best practice for SCI care consists of a situation in which every individual sustaining a SCI is admitted to an integrated, comprehensive system where expertise, facilities and equipment are focused on optimal patient care and cost effectiveness. At the other extreme is the situation condemned by Bedbrook and Sedgley (1980), of piecemeal care for those with SCI characterized by “the occasional patient being treated by the occasional doctor”. In practice, care provided by most SCI centers likely falls somewhere in between these extremes of specialized vs. general care. The present section outlines the studies that are focused on examining the hypothesis that care provided through specialized SCI centers is more efficient and effective than that delivered at general centers. The reader should note that while the majority of these studies were conducted from rehabilitation centers, the experimental manipulation of interest concerns the degree to which specialist care is delivered during the acute care period.

Author Year; Country
Research Design
PEDro Score
Total Sample Size

Methods Outcome

McKechnie et al. (2019); Australia

Population: 2007 (n=856): specialist=334, non-specialist=522; 2010 (n=808): specialist=312, non-specialist=496; 2013 (n=972): specialist=548, non-specialist=424; 2016 (n=1078): specialist=533, non- specialist=545
Treatment: Comparison of patients in specialized versus non-specialized rehabilitation units over 10-year period.
Outcome Measures: Patient demographics, onset from injury, LOS, FIM
Chronicity: Inpatient rehabilitation

  1. Across time-groups, there were a greater proportion of those with SCI in non-specialist units.
  2. Specialist units admit more males and the average age is lower.
  3. Onset from injury and rehab LOS were longer for specialist units.
  4. Total LOS for SCI in specialized units was 90 days, nearly double non- specialized units.
  5. Mean FIM admission was lower in specialized unit versus non- specialized.
  6. Patients admitted to specialized units had higher burden of care at admission and greater functional outcomes (absolute functional gain)

Cheng et al. (2017); Canada Cohort

RHSCIR (n=1138): median age=47y (IQR: 33); Gender: male=79%, female=21%; Level of injury: Cervical=59.4%, other=40.6%
No RHSCIR (n=403): median age=56 (IQR: 30); Gender: male=73.2%, female=26.8%; Level of injury: Cervical=64.9%, other=35.1%
Non-RHSCIR (n=58): median age=56 (IQR: 30); Gender: male=75.9%, female=24.1%; Level of injury: Cervical=78.8%, other=21.2%
Treatment: Patient trajectory was analyzed after being discharged from a specialized acute SCI facility. 3 groups were formed: RHSCIR group received rehab at a specialized facility, No RHSCIR did not receive rehab at a specialized facility, and non-RHSCIR did not attend a specialized facility. Authors then matched 159 RHSCIR and No RHSCIR participants and compared their discharge destination afterward.
Outcome Measures: Predictors of returning home after attending a specialized (RHSCIR) rehab centre, difference in returning home between receiving rehabilitation at RHSCIR or no rehab.
Chronicity: Post-acute

  1. Receiving rehabilitation, age, and AIS D at admission, and acute LOS were significant predictors of being discharged home after attending RHSCIR (p<0.05)
  2. In the matched sample of n=159, there was a significant difference in discharge destination (home or other) between RHSCIR rehab or no RHSCIR rehab (p=0.0004) with RHSCIR rehab having an increased likelihood to discharge home.

Smith (2002); UK Observational

Population: Patients that received rehabilitation within the UK National Health Service.
Treatment: Spinal cord injured patients who received rehabilitation from either a specialized spinal injury units (SIU) or non- specialized spinal injury units completed a postal self-report questionnaire.
Outcome Measures: Functional outcome, satisfaction, social activity.

  1. 13.6% of patients did not use the SIU system.
  2. SIU group had significantly lower:
    1. Superficial pressure sores (p=0.048).
    2. Need for assistance in grooming (p=0.004), eating (p=0.001), and drinking (p<0.001) in patients with complete tetraplegia.
  3. Patients in SIU group were significantly more satisfied with the amount of assistance received (p=0.017).
  4. SIU group was more likely to have:
    1. A partner (p=0.012).
    2. Paid employment (p=0.017).
    3. Voluntary employment (p=0.025).
    4. Satisfaction with sex in those with either tetraplegia (p=0.006) or paraplegia (p=0.05).
  5. No significant difference was seen in general life satisfaction between the two groups.

Tator et al. (1995); Canada
Case Control NInitial=552; NFinal=552

Population: Traumatic SCI; 201/220 consecutive admissions to a newly established specialized interdisciplinary acute SCI unit versus 351 admissions to one of two general hospital trauma units; tetraplegia, paraplegia; incomplete, complete; Male/female ~ 4/1; Median age – 27yr (SCI Specialist unit), 32.0 years (general hospital).
Treatment: Comparison of those treated in a SCI specialist spinal unit (1973-1981) versus a general hospital trauma unit (1947- 1973).
Outcome Measures: LOS, Mortality rate, Cord Injury Neurological Recovery Index. All collected at 6mo (complete) or 12mo (incomplete).

  1. Subjects who were admitted to the specialized SCI unit had significantly shorter acute care LOS than those admitted to the general units (p<0.001). Within the specialized unit subsample, an increased delay from accident to admission resulted in longer LOS (p=0.032).
  2. Subjects who were admitted to the specialized SCI unit had significantly reduced mortality than those admitted to the general units (p=0.022). This was especially evident in those with complete SCI.
  3. Subjects who were admitted to the specialized SCI unit had significantly greater neurologic recovery (p<0.001).

Heinemann et al. (1989); USA
Case Control NInitial=338; NFinal=338

Population: 338 SCI admitted to Rehabilitation, paraplegia, tetraplegia, complete, incomplete.
Treatment: N=185 initially treated in a specialized short-term acute care unit; Control: N=153 initially treated in general hospitals.
Outcome Measures: MBI, MRSCICS Patient Functional Level Scheme, Rehabilitation LOS, Efficiency of Rehabilitation Gains (MBI / natural logarithm of LOS)

  1. Those receiving specialized care made functional gains with significantly greater efficiency and were transferred to rehabilitation significantly faster (p<0.001).
  2. A significantly greater number of people were transferred from general centres with spine instability than from specialized SCI centres (p=0.02).
  3. There was no difference between specialized and general acute care with respect to functional status at rehabilitation admission or discharge nor on rehabilitation LOS.

Yarkony et al. (1985); USA
Case Control NInitial=181 NFinal=181

Population: Traumatic SCI admitted to a specialized rehabilitation unit; Males (n=149) and females (n=32); Avg age 28 years; Tetraplegia (54%), paraplegia (46%); incomplete (58%), complete (42%).
Treatment: Comparison of those treated acutely in a specialized interdisciplinary spinal unit (n=90) versus a general hospital unit (n=91).
Outcome Measures: Joint motion, time to rehabilitation admission, all collected at admission to rehabilitation.

  1. Those admitted from the specialized SCI unit had significantly improved joint motions (i.e., reduced contractures). More had normal range of motion (p<0.05) and fewer abnormalities.
  2. Those admitted from the specialized SCI unit were admitted significantly earlier for rehabilitation as compared to those admitted from the general hospital unit (p<0.01). Those admitted earlier to rehabilitation had reduced numbers of contractures (p<0.01).
  3. Those with tetraplegia had an increased incidence of contractures (p<0.01).

Donovan et al. (1984); USA
Case Control NInitial=1,672 NFinal=1,672

Population: Traumatic SCI, admitted to a specialized, integrated rehabilitation unit in Australia (n=66) versus those admitted to the United States Model Systems (n=1606); tetraplegia, paraplegia; incomplete, complete.
Treatment: Those treated in an integrated, specialized interdisciplinary spinal unit (Australia) admitted <48 hours post-injury versus those admitted to the United States Model Systems at 1-15, 16-30, 31-45 or 46- 60 days post-injury (reflecting progressively less specialized care).
Outcome Measures: Incidence of 7 complications collected at 1-15, 16-30, 31- 45 or 46-60 days post-injury.

  1. Subjects who were cared for in the integrated, specialized unit (Australia) encountered the fewest complications (no statistical analysis was performed).
  2. People sustained progressively more complications with longer periods of delayed admission (US Model Systems). Individuals admitted at these longer delays were cared for initially in general hospital units.


The majority of the studies examining the effect of specialist vs. general SCI care settings focused on this issue during the acute period of care only, with the primary outcome measures being taken at admission to rehabilitation and no follow-up after this point. Of the five studies reviewed, two investigated the results associated with a specialized, integrated unit comprised of both acute and rehabilitation services (Donovan et al. 1984; Smith 2002). Donovan et al. (1984) noted rates of six of seven different medical secondary complications typically encountered by individuals with SCI were lowest for the cohort admitted initially (i.e., typically within 48 hours post-injury) to the specialist SCI centre. This cohort was analyzed retrospectively with complication rates determined at various times throughout rehabilitation (i.e., 1-15, 16-30, 31-45, 46-60 days) and compared with those being admitted to specialist SCI centers from more general care settings at similar time periods. Most striking was the absence of decubitis ulcers during any time period for those under more specialized care vs. a progressively greater incidence for those with greater time spent in general care. No statistical analysis was conducted for this study. Smith (2002) conducted a postal survey (i.e., observational study) of 800 persons who had received care through either a specialist spinal injury unit (n=701) or in a general setting (n=99) within the UK. This cross-sectional sample reported significantly improved outcomes for 10 of 18 health outcomes, 16 of 18 functional outcomes and 5 of 10 social outcomes for those who had received care from the specialist vs non-specialist setting. Notable findings included reduced pressure sores (p=0.048), and a lower level of required assistance for the group who had received specialist care, and there was a trend but no statistically significant difference noted between the groups for life satisfaction (p=0.07).

In the remaining 3 studies all comparisons were limited to specialized vs. general acute care and were retrospective in nature. Two of these studies compared subjects as they were being admitted for comprehensive rehabilitation (Yarkony et al. 1985; Heinemann et al. 1989). In both studies, patients were transferred significantly faster to comprehensive inpatient rehabilitation from more specialized acute care settings than from general hospital settings. In the remaining study by Tator et al. (1995), the same issue was investigated by examining outcomes associated with a seven year experience of a newly developed specialist SCI unit as compared to historical data culled from pre-existing trauma units reflecting more general settings (Tator et al. 1995). In this study, subjects were also transferred to rehabilitation faster from the specialist SCI unit resulting in a reduced length of stay (LOS) in acute care. In general, all of these studies demonstrated improved medical outcomes associated with more specialized care. In addition to the reduced complication rates noted above by Donovan et al. (1984) and Smith (2002), others have noted that more specialized acute care resulted in less spine instability (Heinemann et al. 1989) and significantly improved joint motion with reduced incidence of contractures (Yarkony et al. 1985) upon admission to a comprehensive rehabilitation program. In addition, reduced mortality and improved neurological recovery (as demonstrated by higher scores on the Cord Injury Neurological Recovery Index) were seen in the newly developed specialist SCI unit as compared to the data from pre-existing general trauma units (Tator et al. 1995). It should be noted that a gradual reduction of mortality was seen over the entire study period and that reductions attributed to the specialist unit might also be due to many general gradual improvements in medical care, especially as a historical control was used as the primary basis for comparison.

Only one study has examined the functional benefits realized during rehabilitation associated with SCI-specific acute care vs. that delivered in more general settings. Heinemann et al. (1989) used the Modified Barthel Index to show that those individuals receiving more specialist care made functional gains during subsequent rehabilitation with significantly greater efficiency (i.e., functional change/LOS) than those referred from general settings. No statistically significant differences were seen between the specialist vs. general groups for either admission or discharge functional levels nor were significant differences seen with LOS. There was, however, a significant reduction in the time from injury to rehabilitation admission for those receiving care in the specialist SCI unit. This implies an overall reduced length of total hospitalization for this group, although this data was not reported. Functional benefits associated with early admission and reduced LOS will be reviewed in the next section.

A primary limitation of all studies reported here was the use of retrospective data collection methods and in the case of Tator et al. (1995), the use of historical controls. Another important limitation of some of these studies is the failure to control for (or at least adequately describe) the time to admission to initial care following injury, especially with respect to control subjects (e.g., Donovan et al. 1984; Yarkony et al. 1985; Heinemann et al. 1989). This is an important confounding variable as early admission to a specialized system of care is likely associated with better outcomes as demonstrated in the following section. Therefore, the present conclusions are limited to a Grade 3 level of evidence and some findings have been reduced to Grade 4 if not corroborated and involving inadequate controls. While more carefully controlled prospective studies would be difficult to implement, they would be required to strengthen the evidence in this area.


Based on several retrospective, case-control studies there is level 3 evidence that individuals cared for in interdisciplinary, specialist SCI acute care units soon after injury (most being admitted within 48 hours) begin their rehabilitation program earlier.

There is level 3 evidence that individuals cared for in interdisciplinary, specialist acute care SCI units have fewer complications upon entering and during their rehabilitation programs.

There is level 4 evidence that individuals initially cared for in interdisciplinary, specialist acute care SCI units make more efficient functional gains during rehabilitation (i.e., more or faster improvement).

There is level 4 evidence that individuals cared for in interdisciplinary, specialist SCI units have reduced mortality.

Related Downloads
Outcome Measures
Related Toolkits