Venous Thromboembolism After Spinal Cord Injury: Incidence, Time Course, and Associated Risk Factors in 16,240 Adults and Children

Transcription

1 2240 ORIGINAL ARTICLE Venous Thromboembolism After Spinal Cord Injury: Incidence, Time Course, and Associated Risk Factors in 16,240 Adults and Children Tracey Jones, MD, Viviane Ugalde, MD, Peter Franks, MD, Hong Zhou, PhD, Richard H. White, MD ABSTRACT. Jones T, Ugalde V, Franks P, Zhou H, White RH. Venous thromboembolism after spinal cord injury: incidence, time course, and associated risk factors in 16,240 adults and children. Arch Phys Med Rehabil 2005;86: Objective: To analyze the incidence of venous thromboembolism (VTE) after spinal cord injury (SCI). Design: Retrospective cohort analysis of all SCI cases (16,240) in California from 1991 through Setting: All public hospitals in California. Participants: Subjects (cases) coded as having complete or incomplete SCI. Interventions: Not applicable. Main Outcome Measure: Diagnosis of VTE or death within 91 days of the day of hospital admission. Results: For all cases, the 91-day cumulative incidence of VTE was 5.4%. In a multivariate model, significant predictors of VTE included male sex (odds ratio [OR] 1.4; 95% confidence interval [CI], ), African-American race (OR 1.6; 95% CI, ), complete paraplegia versus tetraplegia (OR 1.8; 95% CI, ), and presence of 3 or more comorbid conditions versus none (OR 1.6; 95% CI, ). Age less than 14 years was predictive of not developing VTE (OR 0.2; 95% CI, ). The incidence of VTE did not change significantly over the 11-year time period (P.07), and VTE was not a significant predictor of death in the first 91 days after hospitalization. Conclusions: The incidence of VTE in SCI patients in California did not change between 1991 and We identified specific risk factors for VTE. Further studies are needed to determine if prompt initiation of medical prophylaxis in high risk subjects reduces the incidence of symptomatic VTE. Key Words: Pulmonary embolism; Rehabilitation; Spinal cord injuries; Thromboembolism; Venous thrombosis by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation VENOUS THROMBOEMBOLISM (VTE), a term that encompasses both deep vein thrombosis (DVT) and pulmonary embolism (PE), has been reported to develop in 8% to From the Department of Physical Medicine and Rehabilitation (Jones, Ugalde), Center for Health Services Research in Primary Care, Department of Family & Community Medicine (Franks), and Division of General Medicine (Zhou, White), University of California, Davis, CA. Supported by the Hibbard E. Williams Endowment for Medical Research, UC Davis. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. Reprint requests to Richard H. White, MD, Div of General Internal Medicine, Ste 2400, PSSB, 4150 V Street, Sacramento, CA 95817, rhwhite@ucdavis.edu /05/ $30.00/0 doi: /j.apmr % of patients with spinal cord injury (SCI). 1-3 Clinical studies using venography have found that SCI patients have an extremely high incidence of asymptomatic VTE, as high as 48% to 100% among patients not given any medical thromboprophylaxis. 2,4,5 Variations in the reported incidence of symptomatic VTE after SCI are probably due to differences in the severity and location of the injury, presence of associated trauma or comorbid conditions, and differences in the use, dose, or duration of medical thromboprophylaxis. 6-9 Prevention of VTE is important because PE is a leading cause of death of SCI patients. 10,11 The incidence of VTE in SCI patients decreased significantly from the late 1970s until the early 1990s, 12 a period characterized by increased use of subcutaneous heparin thromboprophylaxis. It is not known whether the incidence of VTE has continued to decrease in the last 15 years, a period characterized by widespread use of low molecular weight heparin (LMWH) prophylaxis. Moreover, although some reports have suggested that in SCI patients, age over 50 years, lowerextremity fracture, and delayed use of prophylaxis, are predictors of VTE, 4 it is not clear what independent risk factors are most strongly associated with clinically symptomatic VTE in these patients. One difficulty in determining the important risk factors associated with development of VTE is the small size and low power of most SCI studies. 4,9 Even the recent National Spinal Cord Injury Statistical Center registry report 1 provided information on only 700 selected SCI patients concerning the incidence of VTE. We assembled a large cohort of SCI cases using the State of California Patient Discharge Data Set and determined the incidence of VTE over an 11-year period, from January 1991 to December Analysis of a data set of that size can identify certain demographic and clinical risk factors associated with the development of symptomatic VTE. Based on reports in the literature, we hypothesized that: (1) the incidence of VTE would decrease slowly over the 11-year period, 1 (2) prepubertal children would have a significantly lower incidence of VTE than adults and postpubertal teenagers, 13 (3) patients with complete paraplegia or tetraplegia would have a higher incidence of VTE than patients with partial cord injuries, 3 and (4) patients with VTE would not survive as long as those who did not develop VTE. 10,11 METHODS Design and Database The California Patient Discharge Data Set has been described elsewhere. 14 All nonfederal hospitals supply specific information about each inpatient, including basic demographic data, the principal diagnosis, up to 24 secondary diagnoses, a principal procedure, and as many as 20 secondary procedures, using International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes. Since July 1990, encrypted record linkage numbers have made it possible to

2 VENOUS THROMBOEMBOLISM AFTER SPINAL CORD INJURY, Jones 2241 link serial hospitalizations. 15 This study was approved by the California Health and Welfare Agency Committee for the Protection of Human Subjects and the Human Subjects Committee, University of California, Davis. Study Cohort We analyzed all cases admitted from January 1, 1991, to December 31, 2001, that were coded as acute SCI with fracture of the vertebral column (ICD-9-CM, code 806.xx), or SCI without fracture (code 952.xx). Cases were subcategorized as complete paraplegia, complete tetraplegia, incomplete paraplegia, incomplete tetraplegia, unspecified paraplegia, unspecified tetraplegia, or unspecified injury (appendix 1). Fewer than 1% of all cases were hospitalized for longer than 91 days and they were excluded from this analysis as outliers. All cases with a prior diagnosis of VTE were also excluded. Index hospitalization was defined as any period of continuous hospitalization in 1 or more hospitals after the injury (eg, acute care facility to rehabilitation hospital). Placement of an inferior vena cava (IVC) filter (ICD-9-CM code 38.7) within 91 days was noted. Definition of Outcomes Principal outcomes were death and the incidence of VTE (PE or DVT) during the index hospital stay, within 91 days, and within 1 year of the date of hospitalization. Deaths were determined from a state death registry linked to the discharge database. PE was defined by the ICD-9-CM code 415.1x, and DVT in the lower extremity was defined by codes 451.1x, 451.2, , 451.9, 453.1, 453.2, 453.8, and Codes for superficial venous thrombosis and upper-extremity venous thrombosis were not included. Cases coded as having both DVT and PE were categorized as having PE. Each VTE case was categorized as very likely or likely based on the following criteria. Very likely included all cases with a secondary diagnosis of VTE coupled with a code for a diagnostic test for thromboembolism, specifically compression venous ultrasound, ventilation perfusion lung scan, helical scan of chest, etc (codes or 88.44, 88.66, 92.15, 88.77, 87.41, 89.59, 38.7, 99.19), or cases readmitted with a principal diagnosis of VTE coupled with a length of stay (LOS) of 3 or more days (unless death occurred). Likely included cases with a secondary code for VTE and LOS of 3 or more days (unless death occurred). Validation studies have shown that the positive predictive value among cases meeting criteria for very likely is more than 90%, whereas cases meeting criteria for likely have a positive predictive value for objectively confirmed VTE between 72% to 95%. 17,18,20 However, 2 factors associated with lower positive predictive value were excluded from the definition of VTE namely, prior VTE and an LOS for VTE of less than 2 days. Comorbidity Comorbidity was defined using a modified Elixhauser Comorbidity Index. 21 Major elements in this index of 30 serious comorbidities include congestive heart failure, chronic obstructive pulmonary disease, diabetes, etc, but VTE is not included. For our analysis, we removed paralysis from the index. Presence of other factors associated with development of VTE was determined based on specific ICD-9-CM codes: pelvic or lowerextremity fractures (codes 808.x, 820.x, 821.x, 823.x), tracheostomy (code 31.x), and any rehospitalization within 91 days of the injury. We also analyzed the effect of index hospital size on outcomes. Hospitals (N 470) were categorized as large university or university-affiliated (n 9), hospitals with more than 350 beds (n 27), hospitals with between 251 and 350 beds (n 67), hospitals with 125 to 250 beds (n 106), and hospitals with fewer than 125 beds (n 261). We determined the number of cases that were transferred to a second hospital from the index hospital. The source of payment was also analyzed after cases were classified according to the principal payer: insurance or health maintenance organization, Medi-Cal, Medicare, indigent, and other. Statistical Analysis Data were analyzed using Stata. a Continuous variables were compared using the Student t test, and categoric data were analyzed using the chi-square test or Fisher exact test. We used linear regression to test for trends in the incidence of VTE and of SCI in California. Logistic rather than proportional hazard models were generated because the exact day of VTE diagnosis could not be determined for a significant proportion of the cases. Analyses were adjusted for the clustering of observations within hospitals by using a generalized estimating equations approach, with robust sandwich estimators of variance and an independent within-group correlation structure assumed. Age was modeled as a discrete variable, using age 30 to 49 years as the referent. Hospital LOS was not included because diagnosis of VTE likely leads to an increase in the length of hospitalization. To adjust for any secular trends in the incidence of VTE over time, we included calendar year in the models. To determine which specific chronic comorbid conditions were associated with the development of VTE, variables not significantly associated (P.05) were removed by backward elimination. RESULTS There were 16,240 cases coded as having SCI during the 11-year study period. The demographic and clinical characteristics of the entire cohort and of the cases that did or did not develop VTE within 1 year of hospitalization are summarized in table 1. The subjects mean age was 44 years, and the average length of index hospitalization was 18 days. Figure 1 shows the age and sex distribution of the SCI cohort. As shown in figure 2, the standardized incidence ( popest/archives) of SCI patients admitted to California hospitals decreased significantly over the 11-year period (cases/ 100,000 population/year, r 0.9, P.002), whereas there was a trend toward a higher percentage of SCI cases developing VTE over the same period, which was not statistically significant (P.07). The number of ultrasound tests that were coded as being performed increased from approximately 115 (7.3% of cases) from 1991 to 1993 to 183 cases in 1994 (11.8% of cases), but thereafter the number decreased to 115 (7.5% of cases) in The cumulative incidence of VTE was 774 (4.8%) during the index hospital stay, 883 (5.4%) within 91 days and 977 (6.0%) within 1 year after index hospitalization. The relation between age and the incidence of VTE during these 3 time periods is shown in figure 3. Among cases diagnosed within 91 days, 45% were classified as very likely and 55% were classified as likely; 76% had DVT and 24% had PE. Cases 60 to 69 years old had the highest cumulative incidence of VTE (8%). In bivariate analyses, the incidence of VTE within 91 days of hospitalization was significantly higher among men (5.9%) than women (4.1%) (P.001). Only 2 of 181 children (1.1%) 13 years or younger were diagnosed with VTE within 91 days, compared with 4.4% among cases 14 to 19 years old (P.035) and 5.6% among cases age 20 to 39 (P.13). Cases with complete paraplegia had the highest 91-day incidence of VTE,

3 2242 VENOUS THROMBOEMBOLISM AFTER SPINAL CORD INJURY, Jones Table 1: Clinical and Demographic Characteristics of Patients With SCI Variable Total VTE (1y) No VTE (1y) No. of cases 16, (6) 15,263 Mean age SD (y) Sex, n (%)* Male 11,777 (72.5) 768 (6.5) 11,009 Female 4463 (27.5) 209 (4.7) 4254 Race/ethnicity, n (%) White 9747 (60) 541 (5.6) 9206 African American 1919 (11.8) 177 (9.2) 1742 Hispanic 3188 (19.6) 198 (6.2) 2990 Asian/Pacific Islander 931 (5.7) 42 (4.5) 889 Other 455 (7.3) 19 (4.2) 436 Age, n (%) (y) (1.1) (4.8) (5.6) (6.7) (6.7) (6.4) (7.4) (6.1) (4.2) 1324 Extent of paralysis, n (%) Complete paraplegia 1017 (6.3) 121 (11.9) 896 Complete tetraplegia 1218 (7.5) 104 (8.5) 1114 Incomplete paraplegia (or not specified) 5608 (34.5) 368 (7.4) 5240 Incomplete tetraplegia (or not specified) 7395 (45.5) 354 (5.0) 7041 Unspecified level or extent 1002 (6.2) 30 (3) 972 Mean hospital LOS SD (d) Fracture of spine, n (%) 9489 (58.4) 683 (7.2) 8806 Tracheostomy performed, n (%) 1244 (7.7) 139 (11.7) 1105 Deaths, n (%) Died in acute care hospital 1052 (6.4) 60 (5.7) 992 Death 91d 1436 (8.8) 86 (6.0) 1350 Death 181d (not in hospital) 1623 (10) 101 (6.2) 1522 Death 365d (not in hospital) 1868 (11.5) 123 (6.6) 1745 Abbreviation: SD, standard deviation. *One patient did not have an assigned sex. 11%, compared with only 7.8% among cases with complete tetraplegia (P.009). There was but a slight difference in the incidence of VTE between university hospitals (6.7%) and the Percent of All Thromboembolism Cases (1y) 20% 16% 12% 8% 4% 0% > 80 Age (y) Men Fig 1. Age and sex distribution of the SCI cohort. Women large hospitals (7.4%) that had more than 350 beds (P.31), but both groups had a higher incidence than moderate-size hospitals with 250 to 350 beds (5.5%, P.001) and small hospitals (4.2%, P.001). Of the total 883 patients who developed a VTE within 91 days, 453 (51%) were diagnosed in the index hospital, and the remaining were diagnosed after a transfer or at the time of a rehospitalization in the index hospital. The results of multivariate logistic modeling of potential risk factors for developing VTE within 91 days of SCI are shown in table 2. Compared with the referent group of cases that were 30 to 49 years old, cases 8 to 13 years old had significantly lower odds of being diagnosed with VTE (odds ratio [OR] 0.2; 95% confidence interval [CI], ; P.01). Although not statistically significant, VTE risk was also lower among cases ages 14 to 19 (OR 0.7; 95% CI, ; P.03). There was no significant increase in the risk of VTE among cases over the age of 50. Other strong risk factors for VTE included being a male (OR 1.4; 95% CI, ), having complete paraplegia (relative to complete tetraplegia, OR 1.8; 95% CI, ), and being an African American (OR 1.6; 95% CI, ). Incomplete injuries (with the exception of incomplete paraplegia) and unspecified level injuries were associated with a lower risk of VTE compared with cases with complete tetraplegia. An

4 VENOUS THROMBOEMBOLISM AFTER SPINAL CORD INJURY, Jones 2243 Incidence VTE Incidence Per 100 SCI Cases SCI Incidence Per 100,000 Adults Linear (SCI Incidence Per 100,000 Adults) Fig 2. Standardized yearly incidence of SCI and yearly incidence of VTE. increasing number of chronic comorbid conditions was associated with a modest increase in the odds of developing VTE, rising from an OR of 1.3 (95% CI, ) for 1 condition up to an OR of 1.6 (95% CI, ) when 3 or more conditions were present. When the specific comorbidities were analyzed, metastatic cancer had the strongest association (OR 2.5; 95% CI, ). Other significant comorbidities included: underlying chronic neurologic disease (OR 1.4; 95% CI, ), obesity (OR 1.9; 95% CI, ), history of psychiatric disease (OR 1.5; 95% CI, ), and depression/anxiety (OR 1.4; 95% CI, ). Only 322 (2%) of all the cases had an IVC filter; 119 (37%) of those cases were coded as having VTE, 33 (28%) with PE. In 60 of these cases it was determined that the filter was placed after the VTE event. Thirty-eight (15.7%) of the remaining 241 cases that presumably had a prophylactic filter were diagnosed with VTE after the filter was placed. An additional 21 cases had both a filter placed and VTE diagnosed during the same hospitalization, but the temporal sequence could not be determined. Thus, the incidence of VTE among cases with filters could have been as high as 59 (22.5%) of 262. In the first 91 days after injury 1436 (8.8%) subjects died. Mortality rose exponentially with age, from 3.3% of children under age 14, to 3.6% among cases age 30 to 39, to 7.4% among cases age 50 to 59, to 20.3% among cases 70 to 79 years old. As shown in table 2, significant predictors of death were male sex, advancing age, increasing comorbidity, hospitalization in a small hospital, and being tetraplegic. Indigent care cases were also associated with higher risk of dying within 91 days. Cases diagnosed with VTE within the first 91 days were not significantly more likely to die in the first 3 months after injury. DISCUSSION Comprehensive analysis of the incidence of VTE among SCI patients has been hampered because of the small numbers of patients reported in most series, 9,12,22 and by the highly select nature of the cases enrolled in registry studies. 1 To overcome these problems, we analyzed a large population-based SCI cohort. There were 2 unexpected findings: no change in the risk-adjusted incidence of VTE between 1991 and 2001, and no evidence that VTE was a significant predictor of death within 3 months of the injury. Strong predictors of VTE identified were male sex, African American race, an increasing number of comorbid medical conditions, tracheostomy placement, and paraplegia. Age less than 30 years was associated with a lower risk of developing VTE; age less than 14 had the lowest risk. Year Possible explanations for the lack of any reduction in VTE incidence over the 11-year period include: (1) no difference in the effectiveness of regular heparin and LMWH 4 coupled with no meaningful change in the frequency, intensity, or duration of medical thromboprophylaxis, (2) relative ineffectiveness of thromboprophylaxis in preventing symptomatic VTE in high risk patients, (3) an increase in the number of cases with asymptomatic VTE detected by screening ultrasound, offsetting a true decrease in the incidence of symptomatic VTE, (4) a secular trend of SCI patients being sicker and more prone to develop VTE, and (5) increased diagnostic suspicion of VTE by health workers and/or more complete diagnostic coding for VTE. Thus, the absence of any observed change might be explained by opposing secular trends, with an increase in the level of diagnostic suspicion leading to more diagnoses, combined with the use of more effective thromboprophylaxis, leading to a lower incidence, resulting in a net overall effect of no change in VTE incidence. However, it is also possible that thromboprophylaxis for SCI patients was fully optimized by the early 1990s and that use of LMWH in the past 10 years has not caused a reduction in the incidence of symptomatic VTE events. We did not find any evidence to suggest that VTE screening using ultrasound increased significantly during the study period. Because the administrative database we used contained no information about medication use or treatment, we were unable to assess the effect of the use or nonuse of thromboprophylaxis. However, it is highly unlikely that sicker patients, such as those with comorbid medical conditions, lower-extremity fracture, or who required tracheostomy, were not given any thromboprophylaxis. It is more likely that it was either started too late for some patients, or the dose was simply insufficient. Even after adjusting for multiple potential confounders, including hospital size, source of medical payment, etc, African Americans had a significantly higher risk of developing VTE, a finding that has not been previously reported. Although it is known that African Americans have a higher overall incidence of VTE events 23 and idiopathic VTE events, 16 no studies have specifically analyzed the effect of race and ethnicity on the incidence of VTE in SCI patients. Potential unmeasured confounders that might explain this finding include higher levels of factor VIII among blacks with SCI, 24 and potential underuse of Fig 3. Time of thromboembolic events diagnosed within 1 year of SCI.

5 2244 VENOUS THROMBOEMBOLISM AFTER SPINAL CORD INJURY, Jones Table 2: Multivariate Analysis of Risk Factors for VTE or Death Within 91 Days Variable OR for VTE 91 Days 95% CI OR for Death 91 Days 95% CI Year (continuous by year) Sex Women (referent) Men Race/ethnicity White (referent) African American Asian/Pacific Islander Hispanic Other Age (y) * * * (referent) * Insurance status HMO (referent) Medi-Cal 1.3* Medicare Self-pay or other government Indigent * Elixhauser Comorbidity Index score 0 (referent) Hospital size (beds) University hospitals (referent) beds beds 0.7* beds 0.7* VTE NA NA Spinal fracture Trauma to lower extremity 1.3* Intervening hospitalization Tracheostomy Neurologic impairment Tetraplegia (referent) Paraplegia Unspecified tetraplegia Incomplete tetraplegia 0.8* Unspecified paraplegia 1.0* Incomplete paraplegia 1.2* Unspecified defect Abbreviations: HMO, health maintenance organization; NA, not applicable. *P.05. P.01. The number with VTE was too small to determine reliably the odds of VTE. Not entered into the model to predict death. medical prophylaxis. 25,26 It is also possible that race may be a marker for socioeconomic status, which in turn, has been associated with higher levels of morbidity than are captured in comorbidity measures 27 ; we did find an association between having Medicaid insurance and risk of VTE. Our findings confirm several prior reports that found a higher incidence of VTE in the weeks immediately after the SCI. 28,29 Although we could not determine the exact day when VTE was diagnosed, 774 (88%) of all the cases diagnosed with VTE during the first 3 months occurred in the index hospitalization or during a contiguous stay in a second hospital. Only 94 additional cases were diagnosed with VTE between 3 months and 12 months after injury. Finally, the time immediately after injury may be when thromboprophylaxis is least likely to be used. 30 Only a few studies have analyzed outcomes such as VTE among young children with SCI, probably because of the very

6 VENOUS THROMBOEMBOLISM AFTER SPINAL CORD INJURY, Jones 2245 low incidence of SCI in the preadolescent population. 31 Although it is possible that a severe illness such as SCI might overwhelm the protective effects of younger age on thrombosis, our findings confirm that the risk of VTE was indeed significantly lower in young subjects age 13 through 19, and very low among children younger than 14 years old compared with the risk in middle-aged adults. In the 11-year period we studied, there were just 181 VTE cases age 13 or under, and only 8 of those were coded as having complete paraplegia or quadriplegia. This fact underscores the rarity of severe neurologic impairment in this population. In the risk-adjusted model, men were significantly more likely of developing VTE than women (OR 1.4). 1 The reason is unclear but it may be related to unmeasured differences in the overall type and severity of injuries sustained. Most studies of the epidemiology of VTE have reported little difference in the incidence between men and women after trauma or surgery. 23,32-34 The finding that complete paraplegia was associated with significantly greater risk of VTE than was complete tetraplegia has been reported by several investigators. 1,6 We have extended these findings by noting a higher risk of VTE among patients coded as having incomplete paraplegia compared with those with incomplete tetraplegia (OR 1.4; 95% CI, ). Further research is needed to determine if degree of spasticity, intensity of thromboprophylaxis, or other unidentified comorbidities explain this difference. The strongest predictors of death were age over 80 years, placement of a tracheostomy, presence of 3 or more chronic comorbid medical conditions, and tetraplegia. Diagnosis of VTE was not a significant predictor, perhaps because of the relatively low number of cases with PE among the VTE cases. Overall, only 24% of all the SCI cases with VTE had PE, which is much lower than the prevalence of PE (33%) typically reported among cohorts with VTE. 35 The finding that cases that were treated with a prophylactic IVC filter had a very high incidence of VTE, approximately 15% to 23% is difficult to interpret. Most of these patients were very sick with multiple injuries, which may have precluded use of thromboprophylaxis. Because of this bias, no firm conclusions should be drawn about the effectiveness of IVC filters. The finding that cases from small hospitals had a higher incidence of VTE in the risk-adjusted model may reflect a difference in diagnostic bias or severity of initial injury. Transfers of sicker patients to larger hospitals does not account for this finding since we used only the index hospital in the analysis. CONCLUSIONS The incidence of VTE within 91 days of acute SCI averaged 5.4%, with a range of 3% to 12%, depending on whether specific risk factors were present. Unexpectedly, the incidence did not change significantly during the 11-year period. If further studies verify that VTE is indeed associated with the specific risk factors we identified, the incidence of VTE might be further reduced by identifying the highest risk patients as soon as possible after hospital admission and initiating a thromboprophylaxis regimen. 30 Conversely, the very low risk of VTE associated with children younger than 14 years old may lead to standards of care that limit exposure of this age group to the risks of prophylaxis. APPENDIX 1: CODING CATEGORIZATION BY SPINAL CORD LEVEL AND COMPLETENESS OF INJURY Complete Paraplegia Incomplete Paraplegia Unspecified Paraplegia Complete Tetraplegia Incomplete Tetraplegia Unspecified Tetraplegia Unspecified Complete lesion of thoracic or lumbar cord Anterior cord Central cord Posterior cord Sacrum and coccyx, complete cauda equina Cauda equina Sacrum and coccyx, other cauda equina Unspecified SCI of thoracic or lumbar cord Lumbar Sacral Sacrum and coccyx, unspecified or other SCI Complete lesion of cervical cord Anterior cord Central cord Posterior cord Unspecified SCI of cervical cord

7 2246 VENOUS THROMBOEMBOLISM AFTER SPINAL CORD INJURY, Jones Explanations for categorization 806: SCI with fracture : sacrum and coccyx, closed with unspecified SCI categorized as unspecified. Although most likely this represents incomplete injury, the coding makes it unspecified : sacrum and coccyx, closed with complete cauda equina lesion categorized as incomplete injury. Cauda equina injuries typically have motor incomplete pattern of injury : sacrum and coccyx, closed with other cauda equina injury categorized as incomplete. As stated above, unless otherwise specified, cauda equina will have an incomplete motor pattern of injury : sacrum and coccyx, closed with other SCI categorized as unspecified. Most likely incomplete pattern, but the coding is nonspecific : sacrum and coccyx, open with unspecified SCI categorized as unspecified. See above code : sacrum and coccyx, open with complete cauda equina lesion categorized as complete. See above code : sacrum and coccyx, open with other cauda equina injury categorized as unspecified. See above code : sacrum and coccyx, open with other SCI categorized as unspecified. See above code : lumbar, closed categorized as unspecified : lumbar, open categorized as incomplete. 952: SCI without fracture : lumbar categorized as unspecified : sacral categorized as incomplete... para/unable to state complete or incomplete : cauda equine categorized as incomplete , 806.9, 952.8, and are unspecified levels of SCI categorized as unspecified. References 1. Chen D, Apple D, Hudson L, Bode R. Medical complications during acute rehabilitation following spinal cord injury current experience of the Model Systems. Arch Phys Med Rehabil 1999; 80: Deep K, Jigajinni MV, McLean AN, Fraser MH. Prophylaxis of thromboembolism in spinal injuries results of enoxaparin used in 276 patients. Spinal Cord 2001;39: Weingarden SI. Deep venous thrombosis in spinal cord injury; overview of the problem. Chest 1992;102:636S-9S. 4. Spinal Cord Injury Thromboprophylaxis Investigators. Prevention of venous thromboembolism in the acute treatment phase after spinal cord injury: a randomized, multicenter trial comparing low-dose heparin plus intermittent pneumatic compression with enoxaparin. J Trauma 2003;54: ; discussion Green D. Diagnosis, prevalence, and management of thromboembolism in patients with spinal cord injury. J Spinal Cord Med 2003;26: Waring W, Karunas R. Acute spinal cord injuries and the incidence of clinically occurring thromboembolic disease. Paraplegia 1991;29: Geerts WH, Code KI, Jay RM, Chen E, Szalai JP. A prospective study of venous thromboembolism after major trauma. N Engl J Med 1994;331: Geerts WH, Pineo GF, Heit JA, et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126(3 Suppl):338S- 400S. 9. Green D, Hartwig D, Chen D, Soltysik R, Yarnold P. Spinal Cord Injury Risk Assessment for Thromboembolism (SPIRATE Study). Am J Phys Med Rehabil 2003;82: Green DC. Prevention of thromboembolism in spinal cord injury. Consortium for spinal cord medicine clinical practice guidelines. Washington (DC): Paralyzed Veterans of America; DeVivo MJ, Krause JS, Lammertse DP. Recent trends in mortality and causes of death among persons with spinal cord injury. Arch Phys Med Rehabil 1999;80: Winemiller M, Stolp-Smith K, Silverstein M, Therneau T. Prevention of venous thromboembolism in patients with spinal cord injury: effects of sequential pneumatic compression and heparin. J Spinal Cord Med 1999;22: Levy M, Granvill R, Hart D, Meltzer H. Deep venous thrombosis in children and adolescents. J Neurosurg Pediatr 2004; 101: White RH, Zhou H, Romano PS. Length of hospital stay for treatment of deep venous thrombosis and the incidence of recurrent thromboembolism. Arch Intern Med 1998;158: Meux E. Encrypting personal identifiers. Health Serv Res 1994; 29: White R, Zhou H, Romano P. Incidence of idiopathic deep venous thrombosis and secondary thromboembolism among ethnic groups in California. Ann Intern Med 1998;128: White RH, Brickner LA, Scannell KA. ICD-9-CM codes poorly identified venous thromboembolism during pregnancy. J Clin Epidemiol 2004;57: Zakai N, Wright J, Cushman M. Risk factors for venous thrombosis in medical inpatients: validation of a thrombosis risk score. J Thromb Haemostasis 2004;2: Kniffin W Jr, Baron J, Barrett J, Birkmeyer J, Anderson F Jr. The epidemiology of diagnosed pulmonary embolism and deep venous thrombosis in the elderly. Arch Intern Med 1994;154: Cushman M, Tsai A, White R, et al. Deep vein thrombosis and pulmonary embolism in two cohorts: the longitudinal investigation of thromboembolism etiology. Am J Med 2004;117: Elixhauser A, Steiner C, Harris D, Coffey R. Comorbidity measures for use with administrative data. Med Care 1998;36: Merli GJ, Crabbe S, Doyle L, Ditunno JF, Herbison GJ. Mechanical plus pharmacological prophylaxis for deep vein thrombosis in acute spinal cord injury. Paraplegia 1992;30: Stein PD, Hull RD, Patel KC, et al. Venous thromboembolic disease: comparison of the diagnostic process in blacks and whites. Arch Intern Med 2003;163: Patel RK, Ford E, Thumpston J, Arya R. Risk factors for venous thrombosis in the black population. Thromb Haemost 2003;90: Green D, Twardowski P, Wei R, Rademaker AW. Fatal pulmonary embolism in spinal cord injury. Chest 1994;105: Guihan M, Simmons B, Nelson A, Bosshart HT, Burns SP. Spinal cord injury providers perceptions of barriers to implementing selected clinical practice guideline recommendations. J Spinal Cord Med 2003;26: Seeman TE, Crimmins E, Huang MH, et al. Cumulative biological risk and socio-economic differences in mortality: MacArthur studies of successful aging. Soc Sci Med 2004;58: Gondim FA, Lopes AC Jr, Oliveira GR, et al. Cardiovascular control after spinal cord injury. Curr Vasc Pharmacol 2004;2: Bravo G, Guizar-Sahagun G, Ibarra A, Centurion D, Villalon CM. Cardiovascular alterations after spinal cord injury: an overview. Curr Med Chem Cardiovasc Hematol Agents 2004; 2:

8 VENOUS THROMBOEMBOLISM AFTER SPINAL CORD INJURY, Jones Aito S, Pieri A, D Andrea M, Marcelli F, Cominelli E. Primary prevention of deep venous thrombosis and pulmonary embolism in acute spinal cord injured patients. Spinal Cord 2002;40: Radecki R, Gaebler-Spira D. Deep vein thrombosis in the disabled pediatric population. Arch Phys Med Rehabil 1994;75: Silverstein MD, Heit JA, Mohr DN, Petterson TM, O Fallon WM, Melton LJ III. Trends in the incidence of deep vein thrombosis and pulmonary embolism: a 25-year population-based study. Arch Intern Med 1998;158: Heit J, Silverstein M, Mohr D, Petterson T, O Fallon W, Melton LR. Risk factors for deep vein thrombosis and pulmonary embolism: a population-based case-control study. Arch Intern Med 2000;160: White RH, Murin S, Zhou H, Harvey DJ. Effect of ethnicity and gender on the incidence of venous thromboembolism in a diverse population California in Thrombosis Haemost 2005; 93: Murin S, Romano PS, White RH. Comparison of outcomes after hospitalization for deep venous thrombosis or pulmonary embolism. Thromb Haemost 2002;88: Supplier a. Version 82; StataCorp, 4905 Lakeway Dr, College Station, TX