Mortality and longevity after traumatic spinal cord injury in Switzerland from 1990 to 2011: a 21-year longitudinal study

Introduction Traumatic spinal cord injuries (TSCIs) have life-long implications on the health and mobility of the individual's affected. In Switzerland, TSCIs are due most often to sports- and leisure-related activities, transportation-related accidents, and falls; with the highest incidence rates among the young (16–30 years) and the old (76 years and older). Although limited, extant literature suggests a reduced life expectancy up to 50% in comparison to the general population, following TSCI. However, these estimates vary across settings and according to SCI-characteristics. To adequately inform targeted intervention strategies aimed at reducing premature mortality, appropriately stratified, country-specific estimates of risk factors for mortality and reduced survival are imperative. Presently, Switzerland has no estimates on mortality and life expectancy after TSCI. Methods We used data from the Swiss Spinal Cord Injury (SwiSCI) cohort study to model mortality in relation to age, sex and lesion characteristics. Poisson regression using an offset of follow-up time was used to estimate mortality rate ratios and adjusted mortality rates for each pre-determined follow-up period. Hazard ratios (HRs) were estimated using flexible parametric survival models of time since discharge from first rehabilitation to death or September 30, 2011, whichever came first. Skew-normal regression was used to estimate marginally adjusted life expectancy (LE) at age 30 years. Results Two thousand two hundred and fifteen persons were included that incurred a new TSCI between 1990-2011, contributing a total time at risk of 18‘846 person-years (median 7.8 years) and 380 deaths. A total of 73.4% were male, 59.1% were paraplegic, and the average age at injury was 43.4 years (SD = 18.8; IQR = 31). The mortality rate within the first six months post-injury was 36.8 per 1′000 person-years (95% CI = 26.6–47.1), after which the mortality rate reduced dramatically to 9.0 per 1′000 person-years (95% CI = 6.1–11.9) between six months and two years, and remained stable thereafter. Controlling for age and sex, there was a 58% increased risk of mortality for high compared to low lesions (HR = 1.58; 95% CI = 1.23–2.02) and a 48% higher risk for complete versus incomplete lesions (HR = 1.48; 95% CI = 1.16–1.88). At an attained age of 30 years, comparing the highest, more severe lesions (C1–C4: AIS A, B, or C) to less severe lesions at any level (AIS D/E), estimated LE reduced by 17 years. Conclusion This study provides the first estimates of mortality and life expectancy following TSCI in Switzerland. Main risk factors identified were: Age at injury, severity and lesion level. The observed discrepancies in mortality estimates point towards disparities in mortality and survival outcomes according to SCI-specific characteristics. Future research is thus needed that identifies risk of cause-specific mortality for targeted interventions to improve LE and reduce avoidable mortality contributing to large SCI-specific disparities.