Rehospitalization

Cardenas et al. 2004
USA
Case series
N (Initial)=8668; N (End)=1252

Population: SCI: Level of injury: C1-4, C5-8, T1-S5; Severity of injury: AIS: A-D.
Treatment: Retrospective analysis of cases of traumatic SCI for persons with anniversary dates of 1, 5, 10, 15 or 20 years post-discharge occurring between 1995-2002  within the US Model Systems database .
Outcome Measures: Discharge destination, causes for rehospitalization, predictors of rehospitalization.

• 90% of patients were discharged home from acute rehabilitation.
• The most common reasons for rehospitalizations included:
  • Diseases of the genitourinary system.
  • Diseases of skin and subcutaneous tissue.
  • Diseases of the respiratory system.
  • Other unclassified diseases.
  • Diseases of the musculoskeletal system.
• At first year follow up the average number of rehospitalizations were significantly higher than other follow-up years (p<0.001). Rate was 55% in first year and 36-38% thereafter.
• Rehospitalization rates were not significantly different among the different age groups.
• At 1 year rehospitalization was significantly related to:
  • Lower motor FIM scores (p=0.000).
  • Patients funded by state or federal programs (p=0.010).
• At 5 years follow-up, rehospitalization was significantly related to:
  • Lower motor FIM scores (p=0.000).
  • Race, with Hispanics (p=0.009) and other races (p=0.027) were less likely than African Americans.
• At 10 year follow-up, only payer remained significantly related to rehospitalization rates (p=0.004).

Charlifue et al. 2004
USA
Case series
N = 7981

Population: Traumatic SCI: Age 3254≤40, 2908≥41; Level of injury: All levels; Severity of injury: AIS: A-D.
Treatment: Retrospective analysis of cases of traumatic SCI with onset between 1973-1998 from the US Model Systems database .
Outcome Measures: Number of & causes for rehospitalization, days rehospitalized, number of pressure ulcers, self-assessed health status and Satisfaction with Life Scale collected at 1, 5, 10, 15, 20 and 25 years post-injury.

• Rate of rehospitalization was 41% in year 5 and significantly less (35-36%) thereafter (p=0.000)
• Average number of days rehospitalized was highest at year 5 (6.0 days) and significantly less thereafter in a progressive fashion (from 5.4 days at year 10 to 3.7 days by year 25). (p=0.002)
• Perceived health status and SWLS was generally high and pain scores generally low
• Both # of rehospitalizations and a greater # of days rehospitalized 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.

Jaglal et al. 2009
Canada
Case Series
N=559

Population: Traumatic SCI; Age: 47.3±18.4 years; Gender: males = 423, females = 136; Level of injury: Cervical = 350, Thoracic = 126, Lumbar = 62, Other = 21.
Treatment: Retrospective analysis (population-based) of cases of traumatic SCI between 2003-2006 from 6 administrative healthcare databases (Province of Ontario).
Outcome Measures: Rehospitalization rates, causes, predictors collected over a 1 year period following rehabilitation discharge.

• 27.5% were rehospitalized to acute care in the 1st year following initial rehab discharge
• Main causes were musculoskeletal (23.1%), respiratory (11.5%), gastrointestinal (11.0%), urological (10.5%), cardiovascular (10.3%), psychological (9%) and skin (7.3%) disorders
• Factors significantly associated with 1-year rehospitalization in multivariate logistic regression were longer length of rehabilitation stay, rural residence, >50 outpatient physician visits and >50 specialists visits following the initial admission
• Individual factors with highest likelihood (i.e., highest odds ratios) 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). 

Middleton et al. 2004
Australia
Case Series
N=432 (253 persons requiring one or more rehospitalizations)

Population: Rehospitalized SCI patients between 1990-1991, 1999-2000; Traumatic SCI; Gender: males = 338, females = 94; Level of injury: paraplegia = 199, tetraplegia = 229, unclassified = 4; Severity of injury: AISA: A = 206, B = 27, C = 67, D = 132
Treatment: Data from spinal cord injured patients was retrospectively analyzed.
Outcome Measures: causes for rehospitalization, predictors of rehospitalization.

• 253 persons (58.6%) (³12 months post injury) required rehospitalization for a spinal-related cause on at least one occasion during the 10-year study period (total readmissions = 977; 15,127 bed-days; avg length of stay = 15.5 days; median 5 days)
• ~ 10% were readmitted five times or more
• Overall rehospitalization rate in the first 12 months post discharge = 0.64 readmissions per person at risk and decreases to ~0.4 readmissions per person at risk 10 years post acute admission)
• Average length of stay (ALOS) was significantly longer for those with AIS A, B and C (22.2 – 17.0 days) compared to AIS D (11.3 days)
• The most common causes for rehospitalization included:
  • Complications of the genitourinary system (n = 235 (24.1%)), (125 persons (28.9%))
  • Gastrointestinal (GIT)-related (n = 107 (11.0%)), (69 persons (16.0%))
  • Skin pressure areas (n = 87 (8.9%)), (40 persons (9.3%))
  • Musculoskeletal (n = 84 (8.6%)), (60 persons (13.9%))
  • Other causes included Neurological (n = 30 (3.1%)); Respiratory (n = 44 (4.5%)); Cardiovascular (n = 47 (4.8%)); Endocrine (n = 7 (0.7%)); Psychiatric (n = 66 (6.8%)); Other (n = 270 (27.6%))
• The most costly cause of readmission in terms of bed-occupancy, were the skin-related complications (pressure sores: 6.6% of all readmissions, accounted for 27.9% of bed-days and ALOS = 65.9 days)
• Depending on the complication, age and level and completeness of neurological impairment influenced differential rates of readmission; AIS D = 43.2%; AIS A, B and C = 55.2-67.0% (p<0.0001)
• Mean duration to first readmission = 46 months (AIS A-C = 26-36 months, AIS D = 60 months)
• Overall rehospitalization (and bed occupancy) rates trended downwards over time, yet rates were high in the first 4 years after discharge (0.64 readmissions per person, 12.6 bed-days)) before decreasing to 0.35 (2.0 bed-days) as the 10th year approached

Savic et al. 2000
UK
Case series
N=198

Population: Mean age = 57.5 yrs; Gender: males = 84.8%, females = 15.2%; Level and severity of injury (AIS): paraplegic ABC = 97, tetraplegic ABC = 61, D = 40; Time since injury = 33 yrs.
Treatment: SCI patients were interviewed three times from 1990-1996 and their medical records were reviewed.
Outcome Measures: Readmission rates, reasons for readmission, LOS, FIM score, CHART score.

• 64% of patients had 1 or more readmissions between 1990 and 1996.
• Mean length of stay per readmission was 12.03 days.
• Reasons for readmission included:
  • Urinary system complications (40.5%).
  • Skin problems (17%).
  • Digestive system (10%).
  • Musculosketal system (8.7%).
  • Nervous system complications (6.9%).
• Highest reason for bed occupancy was skin problems.
• No significant difference in readmission rates was seen in:
  • Level of injury of the patients.
  • Current age of patients.
• Patients with Frankel/AIS grade D had significantly shorter LOS than patients with A,B or C grade (p=0.005).
• There was significant difference between hospitalized patients and non-hospitalized patients in:
  • Patients hospitalized were paralyzed for 2 years longer than the non hospitalized group (p=0.012).
  • Hospitalized patients had a lower FIM score than non-hospitalized (p=0.031).
  • Hospitalized patients had a lower CHART physical independence score (p=0.003) and CHART occupation score (p=0.001).

Franceschini et al. 2003
Italy
Case series
Initial N=380
Final N=146

Population: All SCI patients with hospitalization between 1989-1994. Mean age = 37.8 yrs; Gender: males = 104, females = 42; Level of injury: Cervical=36.4%, Thoracolumbar=63.7%; Severity of injury (Frankel): A=44.6%, B=2.7%, C = 13%, D=39.7%; Time since injury = 6.1 yrs; Traumatic = 74.7%, Nontraumatic 25.3%..
Treatment: Cross-sectional telephone questionnaire of various rehabilitation outcomes.
Outcome Measures: Custom questionnaire including rehospitalization among other things (i.e., state of health, occupation, mobility, autonomy, social and partner relationships, satisfaction with QoL) collected at mean of 6.1 years post-discharge.

• 25.3% respondents had been hospitalized once in the past year, most frequently for urological problems (22.9%), spasticity (11.4%) and rehab treatment (11.4%).

Paker et al. 2006
Turkey
Case series
N=56

Population: Rehospitalized SCI patients: Mean age = 35yrs; Gender: males = 39, females = 17; Level of injury: cervical = 13, thoracic = 27, lumbar = 16, paraplegia = 44, tetraplegia = 12; Severity of injury: AIS: A = 29, B = 9, C = 12, D = 6, complete = 29, incomplete = 27; Time since injury = 18.4 mnths.
Treatment: Patient data was retrospectively reviewed.
Outcome Measures: Reasons for rehospitalization.

• 7.6% of patients were rehospitalized within the same hospital, of these 71% had been hospitalized at other hospitals making the determination of a true rate uncertain.
• Mean rehospitalization LOS was 72.21 days during the 5 year period.
• Cause of rehospitalization was:
  • Spasticity in 25%.
  • Pressure sores, 17.9%.
  • Urinary tract infections, 16.1%.
  • Spinal surgery, 8.9%.
  • Urinary tract surgery, 5.4%.
  • Pain, 5.4%.
• Rehospitalization due to spinal surgery was significantly related to lower age (p=0.04).
• Reason for rehospitalization was related to length of stay (p=0.07), ASIA score (p=0.06), mobility (p=0.09).

Dorsett & Geraghty 2008
Australia
Case series
N=46 (N=32 @ 10 year follow-up)

Population: Mean age = 32yrs; Gender: males = 42, females = 4; Level of injury: paraplegia=19, tetraplegia=27; Severity of injury: complete =16, incomplete=30
Treatment: 10 year data from acute traumatic SCI patients discharged from the Spinal Injuries Unit of the Queensland Spinal Cord injuries Service from November 1992 to March 1994 was assessed.
Outcome Measures: Mortality, Life situation questionnaire, medical service utilization, hospital admission (including reason for admission) and occurrence of pressure sores collected at discharge, 12 months, 24 months, 36 months and 10 years.

• 9% mortality rate was seen within 3 yrs of study.
• Life situation questionnaire mean scores remained consistent over the 10 years.
• The highest percentage of medical service utilization (10 or more) was at 2 years, while the lowest was at the 10th year (only 3) 9%.
• No significant change was seen in the number of hospitalizations or length of stay over time.
• Overall 32% of patients were rehospitalized in the first 2 years and 52% by the 10th year.
• Only 11% of individuals required rehospitalization for longer than 28 days.
• Common reasons for rehospitalization included: pressure sores, urinary tract infections, bowel obstructions, pneumonia, surgical removal spinal instrumentation, fractures and renal tract calculi.
• At 2 years, reasons for rehospitalization were directly related to SCI, while at 10th year SCI complications were not related to rehospitalization.
• Pressure sore occurrence was highest at the 2nd year, however no significant change in the number of pressure sores occurred over time.
• Half the patients reported no pressure sores over the study period, while 30% tended to have pressure sores at multiple points of time.

Healthcare Utilization

Guilcher et al. 2010
Canada
Case Control
N=1562

Population: Nontraumatic (n=1002) and Traumatic (n=560) SCI; Age at admission: 46.9±17.3 and 61.6±15.8 years; Gender: males = 75.4% and 52.2%, females = 24.6% and 47.8%; Level of injury: Paraplegia = 38.6% and 39.5%, Tetraplegia = 47.1% and 18.6%, Other = 14.3% and 41.9%.
Treatment: Retrospective analysis (population-based) of cases of traumatic SCI between 2003-2006 from 3 administrative healthcare databases (Province of Ontario).
Outcome Measures: Health care utilization collected over a 1 year period following rehabilitation discharge. Predictors of health care utilization included length of stay in rehab, FIM score, rurality index, comorbities (Charlson Index), Socioeconomic Status.

• Mean # of overall physician visits was 31.2 and 29.7 for  nontrauma and trauma respectively. 16.5 and 17.0 for specialist visits. In both cases there was no significant difference in # of visits between nontrauma and trauma although there were differences in the types of physicians being visited.
• Individual factors with highest likelihood (i.e., highest odds ratios) of ≥ 30 physician visits included: lowest quartile FIM @ discharge (OR=1.83), urban (OR=1.59), comorbidities (OR=1.56), ≥ 60 years old (OR=1.54). 
• Individual factors with highest likelihood (i.e., highest odds ratios) of ≥ 20 specialist visits included: comorbidities (OR=2.05), urban (OR=1.92), paraplegia (OR=1.53), lowest quartile FIM @ discharge (OR=1.51).

Munce et al. 2009
Canada
Case Series
N=559

Population: Traumatic SCI; Age: 47.3±18.4 years; Gender: males = 423, females = 136; Level of injury: Cervical = 350, Thoracic = 126, Lumbar = 62, Other = 21.
Treatment: Retrospective analysis (population-based) of cases of traumatic SCI between 2003-2006 from 5 administrative healthcare databases (Province of Ontario).
Outcome Measures: Physician utilization (including family physician, specialist, emergency physician, etc.), rurality index, comorbities (Charlson Index) collected over a 1 year period following rehabilitation discharge.

• Women had significantly more physician visits than men (37.0 vs 30.0, p=0.006)
• Women had significantly more visits to their family physician than men (15.4 vs 10.3, p<0.001)
• Men had significantly more visits to their physiatrists than women (6.6 vs 4.5, p<0.028)
• Individual factors with highest likelihood (i.e., highest odds ratios) of ≥ 50 physician visits included: >70 years old (OR=3.64), direct discharge to chronic care (OR=3.62), in-hospital complication (OR=2.34), thoracic injury level (OR=1.81), direct discharge to rehabilitation (OR=1.69). 
• Individual factors with highest likelihood (i.e., highest odds ratios) of ≥ 50 specialist visits included: direct discharge to chronic care (OR=11.52), direct discharge to rehabilitation (OR=2.45), in-hospital complication (OR=1.99). 

Dryden et al. 2004
Canada
Case Control
N=233 (1165 matched controls)

Population: Traumatic SCI; Median age: 34.0 years; Gender: males = 176, females = 57; Level of injury: Cervical = 117, Thoracic, Lumbar, Sacral or Cauda Equina = 98; Severity: Complete = 43, Incomplete = 69, Unknown = 121.
Treatment: Retrospective analysis (population-based) of cases of traumatic SCI between 1992-1994 from 5 administrative healthcare databases (Province of Alberta).Control subjects registered with the Alberta health system were matched by age, gender and region at a ratio of 5:1).
Outcome Measures: Rehospitalization, Health care utilization, mortality and secondary complications followed over a 6 year period post-injury.

• 57.3% of persons were rehospitalized over the 6 year follow-up period with a median LOS of 4.0 days/hospital stay.
• After initial discharge, persons with SCI had 2.6 more hospital visits than matched controls.
• Persons with SCI had a median # of physician contacts of 22.0 in year 1, declining to 8.0 by year 2 and to 4.0 by year 6. Controls had fewer physician contacts for each year (median = 3.0)
• 20 (8.6%) died during initial hospitalization and 16 (7.5%) died during 6 month follow-up and this was a greater mortality rate with SCI as compared to controls (p<0.001)
• Over the 6 year follow-up 47.6% were treated for a UTI, 33.8% for pneumonia, 19.7% for decubitus ulcer and 15.5% for septicemia