Spinal cord injury rehabilitation in Riyadh, Saudi Arabia: Time to rehabilitation admission, length of stay and functional independence

Abstract

Objectives

1) To describe functional status, length of stay (LOS), and time to rehabilitation admission trends. 2) To identify independent predictors of motor function following rehabilitation for traumatic and nontraumatic spinal cord injury (SCI).

Participants

Adult patients with SCI discharged from inpatient rehabilitation between 2009–2014.

Design

Retrospective cohort study.

Setting

Spinal injury rehabilitation unit at the King Fahad Medical City, Riyadh, Saudi Arabia.

Methods

From retrospective chart reviews of 312 traumatic and 106 nontraumatic SCI, we extracted information on time from injury to rehabilitation admission, rehabilitation LOS, and Functional Independence Measure (FIM) score (admission and discharge). Additionally we collected information on SCI patients’ demographics and injury characteristics. Hierarchical regression was employed to investigate variables associated with FIM motor score at discharge for traumatic SCI and nontraumatic SCI.

Outcome measures

FIM motor score measured at discharge.

Results

Mean and median days from injury to rehabilitation admission were not significantly different for traumatic SCI (377±855, 150) and nontraumatic SCI (288±403, 176). Mean and median days for rehabilitation LOS were significantly longer for traumatic SCI (85±60, 70) compared to nontraumatic SCI (64±59, 49) (p=…..). FIM scores improved significantly from admission to discharge in both groups. For individuals with traumatic SCI, after accounting for admission FIM motor score, tetraplegia and time from injury to rehabilitation, admission had a significant but small negative association with discharge FIM motor score. For individuals with nontraumatic SCI, increasing age and AIS score of A or B had a significant negative association with discharge FIM motor score. LOS in rehabilitation was not a significant contributor for either model.

Conclusions

Decreasing the time from injury to rehabilitation admission may improve outcomes for those with traumatic SCI. Since time spent in rehabilitation was shorter than most other countries, a change in practice in this area may be warranted. Developing new strategies to improve outcomes for older patients with nontraumatic SCI would also be beneficial.

INTRODUCTION

Compared with Western countries, limited research on spinal cord injury (SCI) has been conducted in Saudi Arabia, a country with one of the highest rates of traumatic SCI in the world. ^1,2 Spinal cord injury requires intensive and specialized clinical rehabilitation³ and starting rehabilitation as soon as possible has been found to improve functional outcome at discharge.^4–6 Therefore, it is important to identify trends in time to rehabilitation admission and rehabilitation length of stay (LOS), and explore their relationship with functional recovery.

Internationally, there is wide variation in the time from injury to rehabilitation admission among people with traumatic SCI. In the United States, time from injury to rehabilitation (usually spent in acute care) was recently reported to be 11 days on average.⁷ In China, a mean acute care LOS of 32 days (median 21 days) has been estimated.⁸ An average of 45 days from injury to admission to a specialized spinal unit has been reported in India.⁹ Two multi-center studies from Italy found that mean time from injury to admission was 36.8 days¹⁰ and 55 days.¹¹ 37 days mean time was reported in a recent study in Saudi Arabia.¹²

The onset of a nontraumatic SCI across conditions is variable; however, researchers often use admission to acute care as the date of injury when assessing time to rehabilitation admission. With this caveat in mind, in Italy, Celani et al.¹¹ reported a time from onset to rehabilitation admission for nontraumatic SCI of 167 days (median 57 days). In a Dutch study, nontraumatic SCI patients stayed a median of 23 days in an acute care hospital prior to rehabilitation admission.¹³ We were unable to identify any such published information in Saudi Arabia.

Once patients are admitted to rehabilitation the average LOS for traumatic SCI varies considerably between countries. In the United States, median LOS ranges from 29 to 59 days, depending on injury.¹⁴ Examples of mean rehabilitation LOS in European countries span from a low of 47 days in Russia,¹⁵ up to 240 days in the Netherlands.¹⁶

Nontraumatic SCI LOS in rehabilitation also varies, but is generally shorter in comparison with LOS for traumatic SCI. A retrospective analysis of nontraumatic SCI in 9 countries (Australia, Canada, Italy, India, Ireland, The Netherlands, Switzerland, the United Kingdom, and the United States), showed median LOS in spinal rehabilitation units ranging from 13 days in the United States to 143 days in the United Kingdom and an overall median LOS for all countries of 46.5 days.¹⁷ In Saudi Arabia, an average rehabilitation LOS for traumatic SCI of 59 days and nontraumatic SCI of 46 days respectively has been reported.¹⁸

Etiology of SCI appears to have an effect on functional outcomes. Functional independence has generally been shown to be lower in nontraumatic SCI than traumatic SCI^19,20 though Ones et al.²¹ showed no such difference. Variation in motor functional outcome for traumatic SCI has been related to severity of neurological injury, degree of disability at admission, age, complications and other demographic variables.^22–27 In a study of 100 patients with nontraumatic SCI, Van der Putten²⁸ found the primary factors associated with improved motor function (accounting for 54% of the variance) were having a lower admission score and reduced time from symptom onset to admission. In a Dutch study, completeness and level of injury, and LOS have been associated with changes in FIM motor scores.¹³ A study in Saudi Arabia among those with traumatic SCI found functional independence at admission, level and severity of injury, anxiety/depression score, time between injury and admission to rehabilitation, and LOS were associated with motor functional outcome.¹²

Given high rates of spinal cord injury in Saudi Arabia and limited research on predictors of functional outcomes, especially among those with non-traumatic SCI, we conducted a study with the following objectives: 1) to describe the amount of time from injury to rehabilitation admission and LOS in rehabilitation for individuals with traumatic and nontraumatic SCI, and 2) to identify factors impacting motor functional independence at discharge from rehabilitation for individuals with traumatic and nontraumatic SCI. For both groups we expected that time to admission and LOS in rehabilitation would be longer than that for other countries. We hypothesized that after controlling for admission FIM motor score, time to admission and LOS in rehabilitation would significantly contribute to predicting discharge FIM motor score.

METHODS

Participants

In this retrospective cohort study, data were obtained from chart review of individuals with traumatic and nontraumatic SCI discharged from the inpatient rehabilitation program at King Fahd Medical City- Rehabilitation Hospital located in Riyadh, Saudi Arabia between 2009 and 2014.

Nontraumatic SCI was defined as an injury resulting from spinal stenosis, tumor, ischemia, transverse myelitis, and infection.²⁰ Charts were included for review if individuals were new admissions to rehabilitation and were ≥ 16 years of age. No patients with AIS D injury were included in this study because the Rehabilitation Hospital manages such patients through outpatient care. Patients are admitted to this facility from a variety of sources, including home, acute care and facility placement.

The study protocol was approved by the King Fahd Medical City medical ethics committee.

Data collection

Data extracted from chart review included: age, sex, injury level (paraplegia/tetraplegia), ASIA AIS score, time from injury to rehabilitation admission, length of stay in rehabilitation, discharge setting (e.g., community dwelling, assisted living, chronic hospital), and the Functional Independence Measure (FIM) motor score, which was assessed at admission and discharge from inpatient rehabilitation.

The FIM measures the level of independence in activities of daily living. The FIM is an 18-item assessment composed of 13 motor tasks (eating, grooming, bathing, upper body dressing, lower body dressing, toileting, bladder management, bowel management, bed to chair transfer, toilet transfer, shower transfer, locomotion (ambulatory or wheelchair level), and stairs) and 5 cognitive tasks that were not included in our analyses as the FIM cognitive score is affected by a ceiling effect when used with persons with SCI.^23,29 Each item is scored on a 7-point ordinal scale ranging from 1 (total dependence) to a score of 7 (total independence). The inter-rater reliability, validity, and responsiveness of the FIM are well established.^30–32 The FIM is part of the U.S. Uniform Data System for Medical Rehabilitation and is the most often used registry of standardized information on medical rehabilitation inpatients worldwide.³³

Data analysis

Descriptive statistics (means, standard deviations, medians, frequencies) for patient demographics were performed. For the FIM motor score, a mixed model ANOVA was used to investigate changes between admission and discharge for individuals with traumatic versus nontraumatic SCI. Effect size (partial eta squared) was calculated (small effect size=0.01, a medium effect size=0.06 and a large effect size=0.14).³⁴

Separate hierarchical regressions were used for patients with traumatic and nontraumatic SCI to identify independent predictors of discharge FIM motor score while controlling for admission FIM motor score, demographic variables, and injury characteristics. The variable entered into the first block was admission FIM motor score, variables entered into the second block included demographics (age and gender) and type of injury (tetraplegia/paraplegia and ASIA AIS score), and variables entered into the third block included time from injury to rehabilitation admission and length of stay in rehabilitation. Diagnostics were performed to look for the presence of multivariate outliers and to ensure that statistical assumptions were not violated.

Alpha was set at p≤0.05. Statistical software, SPSS 21 (SPSS Inc., Chicago, IL USA) was used for all analysis.

RESULTS

Patient Demographics

A total of 532 patients with SCI were discharged from inpatient rehabilitation between 2009 to December 2014. Of these, 40 were readmissions, 18 were under 16 years of age, and 57 were categorized as having SCI other than paraplegia or tetraplegia, and were excluded. Of the 423 patients eligible for entry into the study, five (1%) patients were not included in analyses because FIM data were not available. Thus, 418 patients were included in the final data analysis. Demographic information is provided in Table 1. Male patients with traumatic SCI were approximately 75% of the sample, and ≥90% were discharged home. The mean age was 35±17 years though this depended on etiology, with individuals of non-traumatic etiology being significantly older (Table 1) due to a typical increase in traumatic SCI for those between 20–30 years of age (Figure 1).

Table 1.

Demographics for individuals with SCI discharged from inpatient rehabilitation, 2009–2014

Variable	Traumatic SCI	Nontraumatic SCI	p
n	312	106	~
Sex (M/F), n(%)	244/68 (78/22)	78/28 (74/26)	~
Paraplegia/Tetraplegia, n(%)	239/73 (77/23)	95/11 (90/10)	~
AIS score (A,B,C,D), n(%)	261/25/26/0 (84/8/8/0)	58/12/36/0 (55/11/34/0)	~
Discharge destination (1/2/3/4/5/6/7/8/9), n(%)	297/0/1/1/1/4/1/1/1 (98/0/0/0/0/1/0/0/0)	95/1/2/0/3/2/1/0/0 (90/1/2/0/3/2/1/0/0)	~
Age, mean (SD), median, years	31.4 (14.4), 26	43.8 (19.4), 43	<0.0005*
DOI to admission, mean (SD), median, days	377.2 (854.8), 150	288.0 (402.8), 176	0.30
LOS in rehabilitation, mean (SD), median, days	84.8 (59.7), 70	63.9 (58.7), 49	0.002*

^*p<0.05;

SCI= spinal cord injury; n= number of patients; AIS= ASIA Impairment Scale; Discharge destination values= 1(home), 2(board and care), 3(transitional living), 4(intermediate care (nursing home)), 5(acute unit of own facility), 6(acute unit of other facility), 7(rehabilitation facility), 8(other), 9(alternative level of care unit); DOI= date of injury; LOS= length of stay.

Figure 1.

Individuals with SCI Discharged from Inpatient Rehabilitation-Riyadh, Saudi Arabia (2009–2014)

Time from injury to rehabilitation admission

For individuals with traumatic SCI, time from injury to rehabilitation admission averaged 377±855 days, the median value was 150 days and the range was 4 to 9820 days (Table 1). The average time from injury to rehabilitation admission was approximately 100 days lower in individuals with nontraumatic SCI (288±403) (median =176 and range 5 to 2211 days).

Rehabilitation length of stay

For traumatic SCI, the average LOS in rehabilitation was 84±60 days, median days were 70, and range was 4 to 419 days. In nontraumatic SCI, average LOS was significantly lower at 64±59 days, a median value of 49 days, and a range of 9 to 449 days (Table 1).

Clinical outcome

There was a statistically significant main effect for time [F(1,416)= 724.2, p<0.0005, partial eta squared=0.6] where both individuals with traumatic and nontraumatic SCI exhibited improved FIM motor scores from admission to discharge (Table 2). There was a statistically significant main effect for group [F(1,416)=8.0, p=0.005, partial eta squared=0.02] where individuals with nontraumatic SCI had higher FIM motor scores at admission and discharge than individuals with traumatic SCI (Table 2). As evidenced by the partial eta squared, despite reaching statistical significance, the actual difference in mean scores between groups was quite small.

Table 2.

Motor FIM scores at admission and discharge from inpatient SCI rehabilitation

Variable	Admission	Discharge	Change
Traumatic SCI	33.6 (17.9)	58.9 (21.2)	25.3 (16.4)
Nontraumatic SCI	40.5 (16.9)	63.4 (21.7)	22.9 (14.5)

FIM= Functional Independence Measure (Motor score= 13–91).

Predictors of clinical outcome

For traumatic SCI, the largest contributor in the regression model was admission FIM motor score (which accounted for 44% of the model). Discharge FIM motor score had a significant negative association with tetraplegia compared to paraplegia and with days from injury to rehabilitation admission but these variables explained only a small part of the variance of discharge FIM motor score (6%). Days from injury to rehabilitation admission was a weak predictor of discharge FIM motor score (Table 3). LOS in rehabilitation was not a significant predictor.

Table 3.

Hierarchical regression for FIM motor score at discharge from rehabilitation in individuals with traumatic SCI

	Model 1				Model 2				Model 3
	b	SE	β	95% CI	b	SE	β	95% CI	b	SE	β	95% CI
(Constant)	32.58	1.93		28.80 – 36.37	34.66	4.73		25.36 – 43.96	35.71	5.026		25.82 – 45.60
FIM Motor Score at Admission	0.78	0.051	0.66	0.68 – 0.88	0.71	0.052	0.59	0.60 – 0.81	0.73	0.054	0.62	0.63 – 0.84
Age (years)					−0.093	0.062	−0.063	−0.22 – 0.030	−0.090	0.062	−0.061	−0.21 – 0.032
Male/Female					0.85	2.13	0.016	−3.35 – 5.04	1.09	2.10	0.021	−3.05 – 5.23
Paraplegia/Tetraplegia					9.26	2.17	0.19	4.99 – 13.53	8.00	2.17	0.16	3.73 – 12.28
AIS A					−4.54	3.21	−0.079	−10.84 – 1.77	−4.27	3.19	−0.075	−10.54 – 2.01
AIS B					−5.75	4.35	−0.074	−14.31 – 2.82	−5.69	4.32	−0.073	−14.19 – 2.82
DOI to Admission (days)									−0.003	0.001	−0.14	−0.005 – 0.001
LOS (days)									−0.003	0.016	−0.008	−0.033 – 0.028
R2			0.44				0.48				0.50

Bold=p≤0.05; b=unstandardized coefficients; SE=standard error; β=standardized coefficients; CI=confidence interval; FIM= Functional Independence Measure; AIS= ASIA Impairment Scale; DOI= date of injury; LOS= length of stay.

For nontraumatic SCI, as for traumatic SCI, the largest contributor in the regression model was admission FIM motor score (accounting for 56% of the model). Discharge FIM motor score had a significant negative association with age and with AIS score of A or B compared to AIS C. These variables explained 12 percent of the variance of discharge FIM motor score. After admission FIM motor score, AIS score was the strongest predictor of discharge FIM motor score (Table 4). Time from injury to admission and LOS in rehabilitation were not significant predictors.

Table 4.

Hierarchical regression for FIM motor score at discharge from rehabilitation in individuals with nontraumatic SCI

	Model 1				Model 2				Model 3
	b	SE	β	95% CI	b	SE	β	95% CI	b	SE	β	95% CI
(Constant)	24.48	3.71		17.13 – 31.83	45.88	7.93		30.15 – 61.61	43.95	8.61		26.85 – 61.04
FIM Motor Score at Admission	0.96	0.084	0.75	0.79 – 1.13	0.72	0.09	0.56	0.54 – 0.9	0.76	0.093	0.59	0.57 – 0.94
Age (years)					−0.19	0.07	−0.17	−0.33 – −0.06	−0.19	0.07	−0.17	−0.33 – −0.049
Male/Female					−1.2	2.97	−0.024	−7.09 – 4.7	−1.20	2.96	−0.025	−7.09 – 4.68
Paraplegia/Tetraplegia					7.44	4.26	0.11	−1.01 – 15.89	6.65	4.35	0.094	−1.99 – 15.29
AIS A					−15.508	3.22	−0.36	−21.89 – −9.12	−15.84	3.23	−0.37	−22.25 – −9.43
AIS B					−4.71	4.59	−0.069	−13.82 – 4.41	−4.27	4.59	−0.063	−13.38 – 4.82
DOI to Admission (days)									−0.004	0.003	−0.065	−0.01 – 0.003
LOS (days)									0.032	0.023	0.085	−0.014 – 0.077
R2			0.56				0.67				0.68

Bold=p≤0.05; b=unstandardized coefficients; SE=standard error; β=standardized coefficients; CI=confidence interval; FIM= Functional Independence Measure; AIS= ASIA Impairment Scale; DOI= date of injury; LOS= length of stay.

DISCUSSION

This is one of the few studies to explore time from injury to rehabilitation admission among people with SCI in Saudi Arabia and the first, to our knowledge, to look at predictors of functional outcome among people with non-traumatic SCI in this country.

Time from injury to rehabilitation admission

With means above 280 days and median values at or greater than 150 days from injury to admission into SCI rehabilitation for both traumatic SCI and nontraumatic SCI, our values are notably larger than that of other countries where mean and median values for time in acute care often do not exceed 2 months.^7,8,11,13 Our findings also differ from a recent investigation in a rehabilitation hospital in Jeddah, Saudi Arabia, by Abdul-Sattar¹² who reported a much shorter time between injury and rehabilitation admission of 37.7±17.4 days. This difference may result from that study’s inclusion criteria, as patients were only included if they were referred from acute hospitals. In many countries, individuals are admitted to acute care upon injury and are then, if appropriate, directly admitted to inpatient rehabilitation once they are stable enough, so time in acute care and time from injury to inpatient admission would be effectively synonymous. However, it is important to note that time from injury to rehabilitation admission is not necessarily the same as time spent in acute care.

In our study, while some individuals were admitted to inpatient rehabilitation directly after acute care, others may not have been admitted directly. Some may have received homecare services or possibly unreported rehabilitation services in another facility. Patients could potentially not report previous rehabilitation if they consider the services insufficient, or if they believe they will receive better services that way. Although this likely affected a small number of patients, this may have artificially inflated the time from injury to rehabilitation reported here.

Rehabilitation length of stay

Individuals with traumatic SCI in our study had a significantly longer LOS than individuals with nontraumatic SCI, which reflects findings from other countries.^19,35 The average rehabilitation LOS for traumatic SCI of 85 (median, 70) days in our study is higher than the average 29–59 days¹⁴ and median value of 38 days⁷ observed in the United States, similar to that reported in Greece (mean, 88 days),³⁶ and much lower than that seen in a number of countries including Japan (267days),²⁴ the Netherlands (205–240 days),^16,37. For nontraumatic SCI, the average LOS in rehabilitation in our study (64 days) was shorter than that seen in Italy (92 days),¹¹ and the Netherlands (85 days).³⁷ In Saudi Arabia, Al-Jadid and Robert¹⁸ reported values of rehabilitation LOS for patients with traumatic SCI (59 days) and nontraumatic SCI (46 days) that were about 3 weeks shorter than values in our study. However, Abdul-Sattar¹² reported a LOS of 123 days for individuals with traumatic SCI which is some 40 days longer than in our study. Given that Abdul-Sattar’s study excluded more complex patients (those who had additional central nervous system lesions or peripheral nervous system pathology) and our study excluded patients with ASIA D, these differences may well reflect substantial regional or rehabilitation center differences.

A prolonged stay in a specialized center may be positively associated with improved rehabilitation outcomes. Notwithstanding more recent estimations of even shorter rehabilitation LOS,^7,14 data from 1990 to 1998 in the United States showed that the reduction in LOS from 74 to 60 days for individuals with traumatic SCI was associated with negative outcomes including increased in hospital readmissions and discharge to nursing homes instead of discharge to the community.²² Given the LOS we observed was similar to the reduced number of days reported by Eastwood et al.²² it is possible patients in our study did not receive sufficient time in rehabilitation.

Clinical outcome

The improvement in FIM motor score from rehabilitation admission to discharge for individuals with traumatic SCI is similar to that seen in the study by Abdul-Sattar¹² who reported a change score of 30±21. While values from that study were higher, our sample did not include patients with AIS D injury, who are expected to have higher function at discharge.

This is the first report of FIM motor score for individuals with nontraumatic SCI in Saudi Arabia. Our results are similar to those of McKinley et al.¹⁹ in the United States, who reported an admission and discharge FIM motor score of 37 and 56, respectively, and a change score of 19.

We found that those with nontraumatic SCI had higher FIM scores at admission and discharge than individuals with traumatic SCI, which is contrary to the findings of McKinley et al.¹⁹ who showed no difference at admission and a higher score for traumatic SCI at discharge. Yet, in a study from Turkey, Ones et al.²¹ found that individuals with nontraumatic SCI had higher FIM motor score at admission and no difference between groups at discharge. The conflicting findings may reflect differences in the inclusion/exclusion criteria of study samples. For example, in the study by Ones et al.,²¹ the majority of participants had incomplete paraplegia and functional status was better at the time of the hospitalization in the non-traumatic SCI group. Furthermore, the patients were only referred to rehabilitation therapy after being neurologically stabilized. In our study, a greater proportion of individuals with paraplegia in the nontraumatic group may have contributed to higher FIM scores. Study design may also play a role. Our retrospective review may have allowed for more variability in results compared with the longitudinal matched block control design utilized by McKinley et al.¹⁹

Predictors of clinical outcome

Previous research on functional outcome in traumatic SCI in Saudi Arabia ¹² has shown FIM motor change scores are negatively associated with total FIM score at admission, time to rehabilitation admission, and severity of injury and anxiety/depression score, whereas LOS in rehabilitation is the only factor positively associated. Our findings for traumatic SCI were similar as we found predictors of functional outcome at discharge were motor FIM score at admission, level of injury, and time between injury and admission.

The length of time between injury and admission for rehabilitation in this setting is of potential concern, given that early admission to a comprehensive SCI trauma center (i.e., including both acute care and subacute inpatient rehabilitation) has been found to result in shorter acute care LOS and fewer complications,³⁸ whereas our study highlights a notably longer time to rehabilitation than previously reported¹² in Saudi Arabia. This finding suggests a potential area to improve efficiency in the course of SCI rehabilitation.

Our investigation of predictors of functional outcome for nontraumatic SCI differ from the findings of van der Putten et al.²⁸ in that we found higher FIM motor score at admission related to higher score at discharge, age was negatively associated, and time from onset to rehabilitation admission was not significantly related.

Limitations

As this was a retrospective cohort study design from a single rehabilitation hospital, any generalization about the results with regard to other facilities or countries should be made with caution.

CONCLUSION

Our study provides a unique contribution to understanding the state of SCI rehabilitation in Saudi Arabia. Our findings for LOS in rehabilitation in this facility demonstrate, for both traumatic and nontraumatic SCI, fewer days than most countries save the United States. This raises a question about whether inpatient SCI rehabilitation LOS may be too short to optimize patient outcomes. Further, we highlight a notably longer time from injury to inpatient rehabilitation admission than previously reported. Given that longer time since injury predicted poorer functional motor outcomes among those with traumatic SCI, this represents another potential chance in practice. Furthermore, among patients with non-traumatic SCI, the treatment of older patients may need to be re-evaluated to determine if poorer outcomes reflect potentially negative expectations on the part of staff or age-related effects that should be addressed through policy and practice.