Abstract
Objectives:
To evaluate the clinical characteristics, hospital courses, outcomes after hospitalization, and factors associated with outcomes in patients with nontraumatic spinal cord injuries (NTSCI).
Design:
Retrospective analysis.
Setting:
A large for-profit United States health care system.
Participants:
2807 inpatients with NTSCI between 2014 and 2020 were identified using International Classification of Disease codes.
Main Outcome Measure:
Demographic, clinical characteristics, hospital course, and disposition data collected from electronic health record.
Results:
The mean age was 57.91 ± 16.41 years with 69.83% being male. Incomplete cervical level injury was the most common injury type, spinal stenosis was the most common diagnostic etiology and central cord syndrome was the most common clinical syndrome. The average length of stay was 9.52 ± 15.8 days, with the subgroup of 1308 (46.6%) patients who were discharged home demonstrating a shorter length of stay (6.42 ± 10.24 days). Falls were the most common hospital-acquired complication (n = 424, 15.11%) and 83 patients deceased. There were increased odds of non-home discharge among patients with the following characteristics: older age, Medicare insurance, non-black racial minority, increased Charlson Comorbidity Index (CCI), intensive care unit (ICU) stay, use of steroid or anticoagulant medications, and hospital-acquired pulmonary complications. Increased in-hospital mortality was observed in those with Medicaid insurance, ICU stay, increased CCI, diagnosis of degenerative spine disease, other unspecified level of injury, and hospital-acquired pulmonary complications.
Conclusions:
NTSCI in this sample were predominantly incomplete cervical central SCIs. Increased CCI, ICU stay, and hospital-acquired pulmonary complications were associated with poorer outcomes after acute care hospitalization among patients with NTSCI.
Keywords: Nontraumatic spinal cord injury, Hospital complications, Outcomes after hospitalization, Post-acute care, Home discharge
Introduction
The incidence of nontraumatic spinal cord injury (NTSCI) is widely accepted as being much greater than that of traumatic spinal cord injuries (TSCI), yet NTSCI remains less well classified and studied.1,2 The National Spinal Cord Injury Model Systems (SCIMS) Database is the oldest and most well-known mechanism for capturing SCI statistical data in the United States, but focuses almost exclusively on tracking the epidemiology of TSCI. Previous studies that have compared NTSCI and TSCI functional recovery and Functional Independence Measure (FIM) scores have looked exclusively at NTSCI attributed to metastatic cancer and have consistently shown the NTSCI group to have poorer functional outcomes.3,4 Metastatic cancer, however, is only one of many etiologies of NTSCI, which may include degenerative spine conditions, tumors, vascular disease, inflammatory disease, toxic and metabolic disorders, as well as idiopathic causes, all of which tend to be insidious in nature (with the notable exception of spinal cord infarction) and are less easily studied. This heterogeneity of underlying cause and onset stands in contrast to the better defined, acute onset, and comparatively homogenous presentation of TSCI. As a result, research into basic epidemiology, tracking outcomes, and prognostication of NTSCI is challenging.5–7
Previous NTSCI studies have been limited by small sample size and have focused on non-heterogeneous populations in under-developed countries.8,9 SCI Medicine has long relied upon these older reports and empiric observations of providers that suggest an increasing incidence of NTSCIs associated with advancing age and accumulation of medical comorbidities.1,10 Many of these studies have been underpowered and unable to effectively analyze the complex relationships between the varied etiologies of NTSCSI, the multitude of medical comorbidities, characteristics of the acute hospital stay, and resulting discharge outcomes.8,10–12 A contemporary, large multicenter study analyzing the current epidemiology of NTSCI, associated acute hospital course, medical comorbidities, and discharge outcomes would provide a more advanced understanding of this population, improve their management, and help patients, families, and treatment teams manage and plan their post-acute care.
The primary objective of this study is to review the acute care hospitalization of a large number of patients with NTSCI across a large private for-profit health care system in the United States and describe their characteristics, course of treatment while hospitalized, and outcomes after acute care hospital discharge. The secondary objective is to identify NTSCI patient variables that may help predict clinical outcomes after acute care hospitalization. The third objective is to evaluate common medication use in patients with NTSCI including opioid analgesics and analyze their potential impact on discharge disposition.
Methods
The data for this study was drawn from a large for-profit health care system, comprised of 182 acute care hospitals across 20 states in the United States. All of these hospitals have acute medical units but the system contains hospitals with inpatient rehabilitation units. Inclusion criteria for this study were adults 18 years of age and older admitted to acute inpatient units across the healthcare system from January 1st, 2014 to December 31st, 2020 who received International Classification of Disease, Ninth and Tenth Revision, Clinical Modification (ICD-9-CM and ICD-10-CM) NTSCI diagnoses (Appendix). Excluded from the study were patients with concomitant central nervous system (brain and spinal cord) traumatic lesions (G46.0–46.8, G20, G21.11, G21.19, G21.4, G 21.8, G 21.9), patients readmitted to the health care system and patients admitted to inpatient rehabilitation units (IRU) due to distinct heterogeneity of the group in hospital course (i.e. longer length of stay, disposition at discharge, and large overlap with other exclusion criteria [such as concomitant lesions, re-admission to the system]). In addition, patients missing any demographic variables were removed from the final analysis.
Variables
The demographic variables assessed included age, sex, race, ethnicity, medical insurance coverage, length of hospital stay, and discharge destination. Nontraumatic spinal cord injury diagnoses were determined using ICD-9-CM and ICD-10-CM diagnostic codes. The primary diagnosis assigned on admission was reviewed and used to help determine the etiology of NTSCI including degeneration, vascular, infectious, cancer, and others. These diagnostic codes were further classified anatomically by regional level (cervical, thoracic, lumbosacral), injury completeness (complete versus incomplete), and clinical syndrome (anterior, central cord, Brown-Sequard, if applicable) to enhance the categorization of patient variables.13
Medical comorbidities found in this population (exclusive of admitting diagnoses) included: spondylosis, disc disease, spinal stenosis, infection, cancer, falls, cardiovascular disorders, peripheral neuropathy, alcohol abuse, musculoskeletal disorders, and pain (chronic or other). The Charlson Comorbidity Index (CCI), a validated, weighted scoring system was calculated based on 17 comorbid conditions such as congestive heart failure, myocardial infarction, cerebrovascular disease, chronic pulmonary disease, dementia, diabetes without complication, mild liver disease, peptic ulcer disease, peripheral vascular disease, rheumatologic disease, paraplegia, diabetes with complications, renal disease, cancer, moderate or severe liver disease, metastatic cancer, and acquired immune deficiency syndrome.14 It has been validated to quantify overall severity of illness, quantifies an individual’s burden of disease, and helps determine 1-year mortality risk.15,16 The CCI has also been used to stratify risk for adverse outcomes following medical conditions including physically disabling conditions including stroke,17 trauma,18 and falls.19,20
Hospital-acquired conditions studies include: pulmonary complications, venous thromboembolic disease (deep venous thrombosis and pulmonary embolism), falls, cardiovascular disease, gastrointestinal complications, genitourinary complications (urinary tract infection and acute kidney injury), peripheral neuropathy, musculoskeletal disorders, pain (chronic or other), pressure injury, and intraoperative and postoperative complications.21,22
Medications received during the acute hospitalization were organized using the therapeutic classification system from the United States Pharmacopeia Drug Classification System.23 The 10 most frequently used medications were analyzed among the groups with different outcomes after hospitalization (community discharge vs post-acute facility-based care).
Statistics
Descriptive statistics were used to analyze demographic and clinical characteristics via either a T-test or chi-square test, depending on the normal distribution pattern. A stepwise logistic regression model was used to evaluate significant relationships between predictive variables and the outcome variables of non-home discharge (home discharge as 0 and discharge other than home [institutional post-acute care] as 1) and in-hospital mortality (no in-hospital mortality as 0 and in-hospital mortality as 1). While performing stepwise selection, an attempt was made to remove any insignificant variables from the model before adding a significant variable to the model.24 A P value of < .05 was considered significant. Analyses were performed using SAS, version 9.4 (SAS Institute, Inc., Cary, NC) and Stata version 12 (StataCorp, College Station, TX).
Results
There were 7566 unique patient encounters that met the inclusion and exclusion criteria between January 1st, 2014 to December 31st, 2020 that were pooled from 182 hospitals across 20 states. Of these patients, 4759 patients were removed because 249 were missing basic demographic information and 4419 had co-existing traumatic nervous system disorders. There were 91 patients who were hospitalized in inpatient rehabilitation units in the national hospital network and therefore excluded, leaving 2807 patients that were included in the study for further analysis.
69.83% of patients were male with the mean age ± standard deviation being 57.91 ± 16.41 years. The age of the patient population increased over the 6 years studies increasing from 55.85 ± 16.31 in 2014 to 61.19 ± 15.75 in 2020 (P = .0001). There were 1308 (46.60%) patients discharged home (with or without home services) and 1499 patients were discharged other than home. Among the 1499 patients with non-home discharges, 677 (45.16%) were discharged to inpatient rehabilitation units, 400 (26.68%) were discharged to skilled nursing facilities, 127 (8.47%) were discharged to long term care hospitals, and 160 (10.67%) constituted discharge other (including transfer to other hospitals, law enforcement, left against medical advice, etc). Fifty-two patients (3.47%) were discharged to hospice care and 83 (5.54%) patients deceased during acute care hospitalization.
The group discharged home was younger (52.80 ± 15.78 vs 62.36 ± 15.64, P < .0001) without a significant difference in sex ratio (P = 0.131). The group with in-hospital mortality was found to be older 66.45 ± 15.98 compared to patients that survived hospitalization 57.65 ± 16.36 (P < .0001). The overall duration of hospitalization (6.42 ± 10.24 vs 12.22 ± 18.98 days, P < .0001) and percentage of intensive care unit care utilization (31.80% vs 58.11%, P < .0001) were significantly lower in the group discharged home (Table 1).
Table 1.
Demographic and clinical information based on the discharge destination.
| Total, n = 2807 | Home discharge, n = 1308 (46.60%) | Discharge-other than home, n = 1499 (53.40%) | P value | |
|---|---|---|---|---|
| Age, mean ± standard deviation (SD) | 57.91 ± 16.41 | 52.81 ± 15.78 | 62.36 ± 15.64 | <.0001 |
| Male sex Female sex |
1960 (69.83%) 847 (30.17%) |
895 (68.43%) 413 (31.57%) |
1065 (71.05%) 434 (28.95%) |
0.1311 |
| Race Black Other White |
532 (18.95%) 317 (11.29%) 1958 (69.75%) |
237 (18.12%) 139 (10.63%) 932 (71.25%) |
295 (19.68%) 178 (11.87%) 1026 (68.45%) |
0.2658 |
| Insurance Medicaid Medicare No Insurance Other Private |
373 (13.29%) 1319 (46.99%) 242 (8.62%) 245 (8.73%) 628 (22.37%) |
187 (14.30%) 445 (34.02%) 175 (13.38%) 123 (9.40%) 378 (28.90%) |
186 (12.41%) 874 (58.31%) 67 (4.47%) 122 (8.14%) 250 (16.68%) |
<.0001 |
| Length of Stay, mean ± SD | 9.519 ± 15.80 | 6.42 ± 10.24 | 12.22 ± 18.98 | <.0001 |
| Admitting Location ICU Neuro Ortho Other Surgery |
N = 482 162 (33.61%) 33 (6.85%) 1 (0.21%) 250 (51.87%) 36 (7.47%) |
N = 169 47 (27.81%) 6 (3.55%) 1 (0.59%) 99 (58.58%) 16 (9.47%) |
N = 313 115 (36.74%) 27 (8.63%) 0 (0.00%) 151 (48.24%) 20 (6.39%) |
0.0157 |
| ICU stay during stay | 1302 (44.93%) | 429 (31.36%) | 873 (57.06%) | <.0001 |
| Presence of pain diagnosis at admission Neck pain Dorsalgia Back pain/back ache Other (unspecified) |
386 (13.75%) 187 (6.66%) 144 (5.13%) 80 (2.85%) |
213 (16.28%) 109 (8.33%) 81 (6.19%) 39 (2.98%) |
173 (11.54%) 78 (5.20%) 63 (4.20%) 41 (2.74%) |
0.0003 0.0009 0.0171 0.6954 |
| Nontraumatic Nervous System Disorder | 31 (1.10%) | 6 (0.46%) | 25 (1.67%) | 0.0019 |
| Admission based on year 2014 2015 2016 2017 2018 2019 2020 |
380 (13.54%) 423 (15.07%) 387 (13.79%) 455 (16.21%) 393 (14.00%) 386 (13.75%) 383 (13.64%) |
188 (14.37%) 200 (15.29%) 194 (14.83%) 204 (15.60%) 184 (14.07%) 174 (13.30%) 164 (12.54%) |
192 (12.81%) 223 (14.88%) 193 (12.88%) 251 (16.74%) 209 (13.94%) 212 (14.14%) 219 (14.61%) |
0.3786 |
The vast majority of NTSCIs in this study were designated as incomplete with the most common level of injury identified as cervical, followed by thoracic, lumbar and sacral, with unspecified level of injury being the least common. Central cord syndrome represented the most common clinical syndrome and spinal stenosis was the most diagnostic etiology for NTSCI followed by disc disorder and spondylosis. Other leading diagnostic etiologies included tumor/cancer in 115 patients (4.10%), vascular in 56 patients (2.00%), infection in 38 patients (1.35%), and spondyloarthropathy in 21 patients (.75%) (Table 2). Regarding medical comorbidities, diabetes without complication was the most common overall comorbidity, followed by COPD and renal disease.
Table 2.
Classification of spinal cord injuries based on the diagnostic codes.
| Total | Home discharge | Discharge-other than home | P value | |
|---|---|---|---|---|
| Anatomical classification | ||||
| Cervical level (C1–C7) | 1701 (60.60%) | 757 (57.87%) | 944 (62.98%) | 0.0058 |
| Complete | 19 (0.68%) | 6 (0.46%) | 13 (0.87%) | 0.1879 |
| Incomplete | ||||
| Central cord syndrome | 763 (27.18%) | 280 (21.41%) | 483 (32.22%) | <.0001 |
| Anterior Cord | 21 (0.75%) | 6 (0.46%) | 15 (1.00%) | 0.0965 |
| Brown Sequard | 12 (0.43%) | 2 (0.15%) | 10 (0.67%) | 0.0373 |
| Other Incomplete syndrome | 206 (7.34%) | 92 (7.03%) | 114 (7.61%) | 0.5625 |
| Unspecified Completeness | 797 (28.39%) | 389 (25.95%) | 408 (31.19%) | 0.002 |
| Thoracic (T1–12) | 291 (10.37%) | 123 (9.40%) | 168 (11.21%) | 0.1178 |
| Complete | 5 (0.17%) | 3 (0.22%) | 2 (0.13%) | |
| Incomplete | ||||
| Anterior Cord | 5 (0.18%) | 2 (0.15%) | 3 (0.20%) | |
| Brown Sequard | 2 (0.07%) | 1 (0.08%) | 1 (0.07%) | |
| Others | 67 (2.39%) | 22 (1.68%) | 45 (3.00%) | 0.0223 |
| Unspecified Completeness | 221 (7.87%) | 98 (7.49%) | 123 (8.21%) | 0.4841 |
| Lumbar and sacral | 191 (6.80%) | 114 (8.71%) | 77 (5.14%) | 0.0002 |
| Complete | 2 (0.07%) | 1 (0.08%) | 1 (0.07%) | |
| Incomplete | 5 (0.18%) | 3 (0.23%) | 2 (0.13%) | |
| Unspecified | 183 (6.52%) | 110 (8.41%) | 73 (4.87%) | |
| Conus Medullaris | 2 (0.07%) | 0 (0.00%) | 2 (0.13%) | |
| Other (Unspecified level and syndrome) | 199 (7.09%) | 87 (6.65%) | 112 (7.47%) | 0.3983 |
| Etiological classification | ||||
| Degeneration | ||||
| • Spondylosis | 358 (12.75%) | 153 (11.70%) | 205 (13.68%) | 0.1170 |
| • Spinal stenosis | 1018 (36.27%) | 392 (29.97%) | 626 (41.76%) | <.0001 |
| • Disc disorders | 471 (16.78%) | 223 (17.05%) | 228 (16.54%) | 0.7212 |
| Tumor and cancer | 115 (4.10%) | 36 (2.75%) | 79 (5.27%) | 0.0008 |
| Vascular | 56 (2.00%) | 17 (1.30%) | 39 (2.60%) | 0.0139 |
| Spinal instability | 45 (1.60%) | 21 (1.61%) | 24 (1.60%) | 0.9926 |
| Osteomyelitis, discitis | 38 (1.35%) | 14 (1.07%) | 24 (1.60%) | 0.2246 |
| Multiple sclerosis | 2 (0.07%) | 0 (0.00%) | 2 (0.13%) | |
| Transverse myelitis | 14 (0.50%) | 7 (0.54%) | 7 (0.47%) | 0.7981 |
| Spina bifida | 14 (0.50%) | 12 (0.92%) | 2 (0.13%) | 0.0033 |
| Syringomyelia | 11 (0.39%) | 6 (0.46%) | 5 (0.33%) | 0.5965 |
| Spondyloarthropathy | 21 (.75%) | 5 (.38%) | 16 (1.07%) | 0.035 |
The mean number of hospital-acquired complications was .333 ± 0.61 per patient with the subgroup home discharge group showing the lowest rate of complications (.21 ± 0.46 vs .44 ± 0.70, P < .0001). The most common hospital-acquired complications were falls (424 patients, 15.11%), followed by peripheral nerve injuries (133 patients, 4.74%), venous thromboembolic disease (94 patients, 3.35%), intra/perioperative complications (82 patients, 2.92%), and cardiovascular disorders (53 patients, 1.89%). Other than musculoskeletal disorders, all hospital-acquired complications were higher in the group discharged to a non-home destination (Table 3).
Table 3.
Medical comorbidities of the patients with nontraumatic spinal cord injury and hospital acquired complications.
| Total, n = 2807 | Home discharge, n = 1308 (46.60%) | Discharge-other than home, n = 1499 (53.40%) | P value | |
|---|---|---|---|---|
| Charlson Comorbidity Index (mean ± SD) | 2.962 (2.543) | 2.067 ± 2.013 | 3.743 ± 2.696 | <.0001 |
| Diabetes without complications | 474 (16.89%) | 172 (13.15%) | 302 (20.15%) | <.0001 |
| COPD | 300 (10.69%) | 117 (8.94%) | 183 (12.21%) | 0.0052 |
| Renal disease | 206 (7.34%) | 50 (3.82%) | 156 (10.41%) | <.0001 |
| CHF | 171 (6.09%) | 43 (3.29%) | 128 (8.54%) | <.0001 |
| Diabetes with complications | 163 (5.81%) | 48 (3.67%) | 115 (7.67%) | <.0001 |
| Cerebrovascular disease | 100 (3.56%) | 21 (1.61%) | 79 (5.27%) | <.0001 |
| Myocardial infarction | 91 (3.24%) | 30 (2.29%) | 61 (4.07%) | 0.0081 |
| Peripheral vascular disease | 83 (2.96%) | 27 (2.06%) | 56 (3.74%) | 0.0091 |
| Malignancy | 71 (2.53%) | 22 (1.68%) | 49 (3.27%) | 0.0076 |
| Dementia | 63 (2.24%) | 12 (0.92%) | 51 (3.40%) | <.0001 |
| Hospital acquired complications | ||||
| Count of acquired conditions | 0.33 ± 0.61 | 0.21 ± 0.46 | 0.44 ± 0.70 | <.0001 |
| Falls | 424 (15.11%) | 151 (11.54%) | 273 (18.21%) | <.0001 |
| Peripheral nerve injuries | 133 (4.74%) | 47 (3.59%) | 86 (5.74%) | 0.0077 |
| Deep vein thrombosis and pulmonary embolism | 94 (3.35%) | 20 (1.53%) | 74 (4.94%) | <.0001 |
| Intra/Postop complications | 82 (2.92%) | 15 (1.15%) | 67 (4.47%) | <.0001 |
| Cardiovascular disorders | 53 (1.89%) | 13 (0.99%) | 40 (2.67%) | 0.0011 |
| Pulmonary complications | 49 (1.75%) | 3 (0.23%) | 46 (3.07%) | <.0001 |
| Gastrointestinal complications | 34 (1.21%) | 6 (0.46%) | 28 (1.87%) | 0.0007 |
| Musculoskeletal disorders | 36 (1.28%) | 17 (1.30%) | 19 (1.27%) | 0.9397 |
| Genitourinary system disorders | 33 (1.18%) | 6 (0.46%) | 27 (1.80%) | 0.0010 |
| Pressure injuries | 19 (0.68%) | 8 (0.61%) | 11 (0.73%) | 0.694 |
The prevalence of spinal pain on admission was relatively low (neck pain in 386 patients [13.75%], dorsalgia in 187 patients [6.66%], and back pain in 80 patients [2.85%]). However, opioid analgesics were the second most frequently used medications (n = 2538, 90.42%) after electrolyte replacement (n = 2570, 91.56%). Laxatives (n = 2042, 72.75%), anticoagulants (n = 1546, 55.08%), and muscle relaxants (n = 1417, 50.48%) were also commonly used medications. Other than muscle relaxants and antiepileptics, medications were used less frequently by the home discharge group, compared to the non-home discharge group (Table 4).
Table 4.
Ten most frequently used medications used during hospitalization.
| Total, n = 2807 | Home discharge, n = 1308 (46.60%) | Discharge-other than home, n = 1499 (53.40%) | P value | |
|---|---|---|---|---|
| Opioids | 2538 (90.42%) | 1165 (89.07%) | 1373 (91.59%) | 0.0233 |
| Electrolyte replacement | 2570 (91.56%) | 1170 (89.45%) | 1400 (93.40%) | 0.0002 |
| Laxative | 2042 (72.75%) | 877 (67.05%) | 1165 (77.72%) | <.0001 |
| Unfractionated and low molecular weight heparin | 1546 (55.08%) | 605 (46.25%) | 941 (62.78%) | <.0001 |
| Muscle relaxant | 1417 (50.48%) | 676 (51.68%) | 741 (49.43%) | 0.2345 |
| Antiemetic | 1340 (47.74%) | 604 (46.18%) | 736 (49.10%) | 0.1221 |
| Acetaminophen | 1315 (46.85%) | 525 (40.14%) | 790 (52.70%) | <.0001 |
| Anticonvulsant | 1212 (43.18%) | 542 (41.44%) | 670 (44.70%) | 0.0820 |
| GI protectants | 1186 (42.25%) | 490 (37.46%) | 696 (46.43%) | <.0001 |
| Steroid | 965 (34.38%) | 383 (29.28%) | 582 (38.83%) | <.0001 |
Regarding discharge destination, a stepwise regression analysis of the relationship between non-home discharge (dependent variable) and contributing factors (independent variables in Tables 1–4) revealed the following characteristics were strongly associated with non-home discharge (institutional post-acute care) after acute hospitalization: older age (odds ratio [OR] 1.012, 95% confidence interval [CI] 1.003–1.021), Medicare insurance (OR 1.599, CI 1.285–1.989), and non-black (other) racial minority (OR 1.358, CI 1.023–1.803). Additionally, increased CCI (OR 1.231 per each point, 95% CI 1.158–1.309), need for ICU care during anytime in hospital stay, not limited to initial ICU admission (OR 2.452, 95% CI 2.033–2.957), and increased length duration of hospitalization (OR 1.028 per each day CI 1.017–1.040) were all associated with increased odds of non-home discharge after acute hospitalization. Specific medications including anticoagulants and steroid administration were associated with increased odds of discharge other than home (OR 1.443, CI 1.199–1.737 and OR 1.252, CI 1.030–1.521, respectively). Finally, hospital-acquired pulmonary complications were shown to increase the odds of a non-home discharge after acute hospitalization (OR 7.689, CI 1.652–35.796) (Table 5).
Table 5.
Stepwise logistic regression analysis of relationship between discharge destination and contributing factors (odds ratio for other than home discharge).
| Independent variables | Odds Ratio | P value | [95% Conf. Interval] | |
|---|---|---|---|---|
| Demographic | ||||
| Age per each year | 1.012 | 0.008 | 1.003 | 1.021 |
| Black race | 1.240 | 0.071 | 0.982 | 1.566 |
| Other (non-Black race) | 1.358 | 0.034 | 1.023 | 1.803 |
| Medicare | 1.599 | 0.000 | 1.285 | 1.989 |
| No insurance | 0.526 | 0.000 | 0.370 | 0.748 |
| Spinal cord injury classification | ||||
| Thoracic level | 1.726 | 0.001 | 1.260 | 2.366 |
| Other level | 2.023 | 0.000 | 1.539 | 2.658 |
| Degenerative spine disorder | 1.664 | 0.000 | 1.334 | 2.077 |
| Comorbidities | ||||
| Charlson comorbidity index | 1.231 | 0.000 | 1.158 | 1.309 |
| Neck pain | 0.672 | 0.031 | 0.468 | 0.965 |
| Hospital course and hospital acquired complications | ||||
| Intensive care unit (ICU) during stay | 2.452 | 0.000 | 2.033 | 2.957 |
| Length of stay | 1.028 | 0.000 | 1.017 | 1.040 |
| Pulmonary complications | 7.689 | 0.009 | 1.652 | 35.796 |
| Intra and postoperative complications | 1.884 | 0.088 | 0.911 | 3.900 |
| Medications | ||||
| Heparin and other | 1.443 | 0.000 | 1.199 | 1.737 |
| Corticosteroid | 1.252 | 0.024 | 1.030 | 1.521 |
| Acetaminophen | 1.171 | 0.090 | 0.976 | 1.405 |
With respect to in-hospital mortality, a stepwise regression analysis of contributing factors revealed increased odds of mortality in patients with Medicaid insurance (OR 2.431, CI 1.209–4.891), other (unspecified) spinal level of injury (OR 3.754, CI 1.877–7.507), diagnosis of degenerative spine disease (OR 2.267, CI 1.004–5.121), need for ICU level of care during stay (OR 22.869, CI 8.163–64.072), CCI (OR 1.373 per one point increment, CI 1.248–1.512), and pulmonary complications (OR 5.711, CI 2.388–13.949) (Table 6).
Table 6.
Stepwise logistic regression analysis of variables for in-hospital mortality.
| Independent variables | Odds ratio | P value | [95% Conf. Interval] | |
|---|---|---|---|---|
| Demographic | ||||
| Black race | 0.407 | 0.021 | 0.190 | 0.871 |
| Medicaid insurance | 2.431 | 0.013 | 1.209 | 4.891 |
| Classification and comorbidity | ||||
| Other spine level | 3.754 | 0.000 | 1.877 | 7.507 |
| Degenerative spine disease | 2.267 | 0.049 | 1.004 | 5.121 |
| CCI | 1.373 | 0.000 | 1.248 | 1.512 |
| Hospital course and complication | ||||
| ICU during stay | 22.869 | 0.000 | 8.163 | 64.072 |
| Pulmonary complication | 5.771 | 0.000 | 2.388 | 13.949 |
| Cardiovascular complication | 3.082 | 0.056 | 0.973 | 9.760 |
| Medication | ||||
| Benzodiazepine | 0.329 | 0.003 | 0.157 | 0.689 |
| Laxative | 0.254 | 0.000 | 0.147 | 0.437 |
Discussion
Similar to prior studies and the SCIMS findings, the results of this study show NTSCI is associated with older age with the mean age of NTSCI being 58 years compared to TSCI reported at 43 years.8,25,26 Additionally, within the 7 year study period, the mean age of NTSCI patients was observed to increase, which is also consistent with previously reported results and likely reflects demographic changes occurring in the U.S.8,25 The length of stay during acute care among NTSCI population was slightly shorter than that reported for the TSCI population (9.5 days vs 11 days).27 Other noteworthy demographic differences included less male predominance (69.83%) among the NTSCI in this study versus previously reported TSCI data (80.4%).27 In-hospital mortality was slightly higher in the current study NTSCI population than reported in TSCI data (2.96% vs 2.1%). The difference in underlying mechanisms can explain the demographic difference as previously suggested.8,28 The slightly higher in-hospital mortality despite shorter length of stay in the NTSCI population is an interesting finding that may be explained by the older age of patients (58 vs. 43), supported by evidence showing older age was shown to be independently associated with in-hospital mortality, and possibly due to the higher rate of medical comorbidities in the NTSCI population (for example diabetes present in 22.7% in this population vs 10.6% in TSCI data). The high rate of associated injuries and surgeries among TSCI group (37.3% and 80%, respectively) versus the NTSCI group may have contributed to the increased length of stay among TSCI group and may be due to exclusion of patients with associated nervous system injuries in the current study.27
Within this large sample, NTSCI patients discharged to facility-based post-acute care were both older and experienced longer lengths of hospital stay. Given this information, early rehabilitation intervention and planning for post-acute care rehabilitation should be a goal for these patients.29
Medicare insurance and non-black racial minority were associated with increased odds of non-home discharge. Medicare enrollment typically begins at age 65, so these patients were expectedly older and demonstrated a higher CCI which likely contributed to their need for non-home discharge. Our study is unable to determine whether or not race is a determinant of discharge destination after hospitalization for NTSCI as it may be a surrogate for other patient characteristics that were not assessed, including socioeconomic status.30,31
Cervical incomplete SCI was the most common level of NTSCI with central cord syndrome being the most common clinical syndrome. Similar to findings from prior studies, cervical spinal stenosis, followed by disc disorders, were the most common cause of NTSCI.2,8 Conversely, tumors, cancer, and vascular etiologies were found to be lower than previously reported. This difference may be attributable to the data sources, where previous studies collected data from specialized SCI units or trauma centers, and this study used data from acute community-based hospitals and excluded specialized rehabilitation facilities. Furthermore, these differences in clinical settings may have selected for less severe cases that were able to be managed in a non-specialized community hospital setting.
The high rate of opioid analgesics administration during hospitalization (90.42%), despite the relatively low prevalence of spinal pain diagnosis on admission (13.75% with neck pain), may be indicative of anticipatory and aggressive pain control practices in community hospital settings, overestimation of pain needs within this population, or inappropriate medication selection. The importance of elucidating pain etiology (nociceptive or neuropathic) in order to optimize treatment cannot be overstated and the current study is limited in further classifying the type of pain and indication for opioid analgesics versus other analgesic medications. Early management of acute pain may be important in mitigating the development of chronic pain following NTSCI because it has been previously demonstrated to develop following TSCI.32–36 Further complicating this issue is the association between spinal conditions and opioid use disorder that has been reported in prior studies.37,38
The association between oral steroid and subcutaneous heparin administration and increased odds of non-home discharge may be a proxy for NTSCI severity, as steroids are often administered in the presence of neurologic deficits or pain following NTSCI, and venous thromboembolic disease prophylaxis is often started in the presence of motor impairment following NTSCI.
The most common hospital-acquired complications for people with NTSCI were falls during acute care hospitalization, but these did not have a significant impact on discharge destination. Although pulmonary complications were not common (n = 49, 1.75%), they were strongly associated with non-home discharge after acute hospitalization. While prior studies of NTSCI in rehabilitation settings have suggested that pulmonary infections, urinary tract infections, and skin pressure injuries are the most common hospital-acquired medical complications,8,9 the larger, more diverse population of acutely hospitalized patient used in this study showed a lower incidence of these medical complications.9,39 Surprisingly, peripheral nerve injuries were 2nd most common in this population with post-procedural complications (hemorrhage, hematoma, other complications) accounting for most peripheral nerve injuries followed by unspecified polyneuropathy diagnosis discovered in post-hoc analysis. Many of the complications associated with NTSCI seen in this study are preventable and our results suggest that efforts should be focused on reducing these complications particularly pulmonary complications as they can have a major effect on discharge destination and long-term outcome.
The group with non-home discharge after acute hospitalization had a higher CCI compared to the group with home discharge in this study (3.743 ± 2.696 vs. 2.067 ± 2.013, P < .0001), and each additional CCI point increased the odds for non-home discharge by 1.231. The CCI was also associated with increased in-hospital mortality in this study by odds of 1.373. The CCI has been previously used to study for prognostic variables for outcomes following stroke,17 trauma,18 and falls19,20; this study is the first to evaluate its applicability to NTSCI.
Limitations
This study has a number of limitations. The study population was limited to hospitals within a private for-profit healthcare system and not be fully applicable to non-profit healthcare systems such as those within academic centers or the Veterans Health Administration. In addition, the large number of records excluded was a limitation although exclusion of coexisting traumatic nervous system disorders decreased the heterogeneity of the group. This study relied on a retrospective review of hospital discharge-based data including ICD-9-CM and ICD-10-CM codes and was fundamentally limited by potential errors in coding which is often based on clinical documentation. Further, the use of these codes made it impossible to determine whether or not identified NTSCIs were new or established diagnoses and also led to a large number of patients identified with an unspecified level of injury. The database used for this study lacked baseline functional data and descriptions of the residences for patients which often are an important determinant of functional outcome and discharge location, respectively. This study also is unable to determine causality and indications for prescribed medications.
Conclusions
Current knowledge regarding the demographics, hospital course, and outcomes after acute hospitalization for NTSCI has been lacking and this information would be useful for patients, families, and providers to improve care and anticipate functional outcomes. To our knowledge, this is the largest and most comprehensive study of NTSCI, using a large database of electronic health records from a large for-profit health care system in the United States. This study upholds previous knowledge that NTSCI is more common that TSCI and is associated with older age. NTSCI in this sample were predominantly incomplete cervical central SCIs. Increased CCI, ICU stay, hospital-acquired pulmonary complications were associated with poor outcome after acute care hospitalization among patients with NTSCI.
Disclaimer statements
Disclaimer This research was supported (in whole or in part) by HCA and/or an HCA affiliated entity. The views expressed in this publication represent those of the author(s) and do not necessarily represent the official views of HCA healthcare or any of its affiliated entities.
Conflicts of interest No conflicts of interest have been reported by the authors or by any individuals in control of the content of this article.
Appendix. Inclusion and exclusion criteria.
| Inclusion Criteria | Exclusion criteria |
|---|---|
| S14.0XXA S14.101A-S14.109A S14.111A-S14.119A S14.121A-S14.129A S14.131A-S14.138A, S24.131A-S24.134A, S14.139A S14.141A-S14.149A S14.151A-S14.159A S24.0XXA S24.101A-S24.104A, S24.109A S24.111A-S24.114A, S24.119A S24.131 A-S24.134A, S24.139A S24.141A-S24.144A, S24.149A S24.151A-S24.154A, S24.159A S34.01XA, S34.02XA S34.101A-S34.105A, S34.109A S34.111A-S34.115A, S34.119A S34.121A-S34.125A, S34.129A S34.131A, S34.132A, S34.139A |
S12, S22, S32 |
Funding Statement
None.
References
- 1.New PW, Sundararajan V.. Incidence of non-traumatic spinal cord injury in Victoria, Australia: a population-based study and literature review. Spinal Cord 2008;46(6):406–11. [DOI] [PubMed] [Google Scholar]
- 2.Ullah S, Qamar I, Qureshi AZ, Abu-Shaheen A, Niaz A.. Functional outcomes in geriatric patients with spinal cord injuries at a tertiary care rehabilitation hospital in Saudi Arabia. Spinal Cord Ser Cases 2018;4:78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Fortin CD, Voth J, Jaglal SB, Craven BC.. Inpatient rehabilitation outcomes in patients with malignant spinal cord compression compared to other non-traumatic spinal cord injury: a population based study. J Spinal Cord Med 2015;38(6):754–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.De la Garza-Ramos R, Abt NB, Kerezoudis P, Krauss W, Bydon M.. Provider volume and short-term outcomes following surgery for spinal metastases. J Clin Neurosci 2016;24:43–6. [DOI] [PubMed] [Google Scholar]
- 5.Ge L, Arul K, Mesfin A.. Spinal cord injury from spinal tumors: prevalence, management, and outcomes. World Neurosurg 2019;122:e1551–e1556. [DOI] [PubMed] [Google Scholar]
- 6.Grassner L, Marschallinger J, MW Dunser, HF Novak, Zerbs A, Aigner L, et al. Nontraumatic spinal cord injury at the neurological intensive care unit: spectrum, causes of admission and predictors of mortality. Ther Adv Neurol Disord 2016;9(2):85–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.PW New, RK Reeves, É Smith, Townson A, Eriks-Hoogland I, Gupta A, et al. International retrospective comparison of inpatient rehabilitation for patients with spinal cord dysfunction epidemiology and clinical outcomes. Arch Phys Med Rehabil 2015;96(6):1080–7. [DOI] [PubMed] [Google Scholar]
- 8.McKinley WO, Seel RT, Hardman JT.. Nontraumatic spinal cord injury: incidence, epidemiology, and functional outcome. Arch Phys Med Rehabil 1999;80(6):619–23. [DOI] [PubMed] [Google Scholar]
- 9.New PW, Rawicki HB, Bailey MJ.. Nontraumatic spinal cord injury: demographic characteristics and complications. Arch Phys Med Rehabil 2002;83(7):996–1001. [DOI] [PubMed] [Google Scholar]
- 10.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;93(2):325–31. [DOI] [PubMed] [Google Scholar]
- 11.SN Cosar, OU Yemisci, Oztop P, Cetin N, Sarifakioglu B, SA Yalbuzdag, et al. Demographic characteristics after traumatic and non-traumatic spinal cord injury: a retrospective comparison study. Spinal Cord 2010;48(12):862–6. [DOI] [PubMed] [Google Scholar]
- 12.Quintana-Gonzales A, Sotomayor-Espichan R, Martínez-Romero M, Kuroki-García C.. Nontraumatic spinal cord injury: etiology, demography and clinics. Rev Peru Med Exp Salud Publica 2011;28(4):633–8. [PubMed] [Google Scholar]
- 13.New PW, Marshall R.. International spinal cord injury data sets for non-traumatic spinal cord injury. Spinal Cord 2014;52(2):123–32. [DOI] [PubMed] [Google Scholar]
- 14.Ofori-Asenso R, Zomer E, KL Chin, Si S, Markey P, Tacey M, et al. Effect of comorbidity assessed by the Charlson Comorbidity Index on the length of stay, costs and mortality among older adults hospitalised for acute stroke. Int J Environ Res Public Health 2018;15(11):2532. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Austin SR, Wong Y-N, Uzzo RG, Beck JR, Egleston BL.. Why summary comorbidity measures such as the Charlson Comorbidity Index and Elixhauser score work. Med Care 2015;53(9):e65–e72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bannay A, Chaignot C, P-O Blotière, Basson M, Weill A, Ricordeau P, et al. The best use of the Charlson Comorbidity Index with electronic health care database to predict mortality. Med Care 2016;54(2):188–94. [DOI] [PubMed] [Google Scholar]
- 17.Tessier A, Finch L, Daskalopoulou SS, Mayo NE.. Validation of the Charlson Comorbidity Index for predicting functional outcome of stroke. Arch Phys Med Rehabil 2008;89(7):1276–83. [DOI] [PubMed] [Google Scholar]
- 18.Gabbe BJ, Magtengaard K, Hannaford AP, Cameron PA.. Is the Charlson Comorbidity Index useful for predicting trauma outcomes? Acad Emerg Med 2005;12(4):318–21. [DOI] [PubMed] [Google Scholar]
- 19.Ishida Y, Maeda K, Nonogaki T, Shimizu A, Yamanaka Y, Matsuyama R, et al. Malnutrition at admission predicts in-hospital falls in hospitalized older adults. Nutrients 2020;12(2):541. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Bates DW, Pruess K, Souney P, Platt R.. Serious falls in hospitalized patients: correlates and resource utilization. Am J Med 1995;99(2):137–43. [DOI] [PubMed] [Google Scholar]
- 21.Rahman M, Neal D, Fargen KM, Hoh BL.. Establishing standard performance measures for adult brain tumor patients: a nationwide inpatient sample database study. Neuro Oncol 2013;15(11):1580–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Michel J, Nghiem HD, Jackson TJ.. Using ICD-10-AM codes to characterise hospital-acquired complications. Health Inf Manag 2009;38(3):18–25. [DOI] [PubMed] [Google Scholar]
- 23.USP Model Guidelines Expert Committee and US Pharmacopeia Staff . Narrative review: the U.S. pharmacopeia and model guidelines for medicare part D formularies. Ann Intern Med 2006;145(6):448–53. [DOI] [PubMed] [Google Scholar]
- 24.Citterio A, Franceschini M, Spizzichino L, Reggio A, Rossi B, Stampacchia G.. Nontraumatic spinal cord injury: an Italian survey. Arch Phys Med Rehabil 2004;85(9):1483–7. [DOI] [PubMed] [Google Scholar]
- 25.Ge L, Arul K, Ikpeze T, Baldwin A, Nickels JL, Mesfin A.. Traumatic and nontraumatic spinal cord injuries. World Neurosurg 2018;111:e142–e148. [DOI] [PubMed] [Google Scholar]
- 26.Facts and Figures at a Glance . 2021.
- 27.Spinal cord injury model systems 2019 annual report-complete public version. NSCISC National Spinal Cord Injury Statistical Center; 2019. https://www.nscisc.uab.edu/Public/2019%20Annual%20Report%20-%20Complete%20Public%20Version.pdf.
- 28.Scivoletto G, Farchi S, Laurenza L, Molinari M.. Traumatic and non-traumatic spinal cord lesions: an Italian comparison of neurological and functional outcomes. Spinal Cord 2011;49(3):391–6. [DOI] [PubMed] [Google Scholar]
- 29.Stucki G, Stier-Jarmer M, Grill E, Melvin J.. Rationale and principles of early rehabilitation care after an acute injury or illness. Disabil Rehabil 2005;27(7–8):353–9. [DOI] [PubMed] [Google Scholar]
- 30.Van der Cruyssen K, Vereeck L, Saeys W, Remmen R.. Prognostic factors for discharge destination after acute stroke: a comprehensive literature review. Disabil Rehabil 2015;37(14):1214–27. [DOI] [PubMed] [Google Scholar]
- 31.Mendoza J, DJ Pangal, Cardinal T, PA Bonney, Lechtholz-Zey E, BA Strickland, et al. Systematic review of racial, socioeconomic, and insurance status disparities in neurosurgical care for intracranial tumors. World Neurosurg 2021;158:38–64. [DOI] [PubMed] [Google Scholar]
- 32.IJ Kim, DH Kim, YW Song, Guermazi A, MD Crema, DJ Hunter, et al. Prevalence of chronic pain after spinal cord injury: a systematic review and meta-analysis. Reg Anesth Pain Med 2021;46(4):328–36. [DOI] [PubMed] [Google Scholar]
- 33.van Gorp S, Kessels AG, Joosten EA, van Kleef M, Patijn J.. Pain prevalence and its determinants after spinal cord injury: a systematic review. Eur J Pain 2015;19(1):5–14. [DOI] [PubMed] [Google Scholar]
- 34.Dijkers M, Bryce T, Zanca J.. Prevalence of chronic pain after traumatic spinal cord injury: a systematic review. J Rehabil Res Dev 2009;46(1):13–29. [PubMed] [Google Scholar]
- 35.Burke D, Fullen BM, Stokes D, Lennon O.. Neuropathic pain prevalence following spinal cord injury: a systematic review and meta-analysis. Eur J Pain 2017;21(1):29–44. [DOI] [PubMed] [Google Scholar]
- 36.Siddall PJ, McClelland JM, Rutkowski SB, Cousins MJ.. A longitudinal study of the prevalence and characteristics of pain in the first 5 years following spinal cord injury. Pain 2003;103(3):249–57. [DOI] [PubMed] [Google Scholar]
- 37.Guan Q, ME Hogan, Calzavara A, McCormack D, AK Lofters, Patel T, et al. Prevalence of prescribed opioid claims among persons with nontraumatic spinal cord dysfunction in Ontario, Canada: a population-based retrospective cohort study. Spinal Cord 2021;59(5):512–9. [DOI] [PubMed] [Google Scholar]
- 38.SJT Guilcher, ME Hogan, McCormack D, AJ Calzavara, SL Hitzig, Patel T, et al. Prescription medications dispensed following a nontraumatic spinal cord dysfunction: a retrospective population-based study in Ontario, Canada. Spinal Cord 2021;59(2):132–40. [DOI] [PubMed] [Google Scholar]
- 39.McKinley WO, Tewksbury MA, Godbout CJ.. Comparison of medical complications following nontraumatic and traumatic spinal cord injury. J Spinal Cord Med 2002;25(2):88–93. [DOI] [PubMed] [Google Scholar]