Life Expectancy Following Rehabilitation: A NIDRR Traumatic Brain Injury Model Systems Study

Harrison-Felix, Cynthia PhD; Kreider, Scott E. D. MS; Arango-Lasprilla, Juan C. PhD; Brown, Allen W. MD; Dijkers, Marcel P. PhD; Hammond, Flora M. MD; Kolakowsky-Hayner, Stephanie A. PhD, CBIST; Hirshson, Chari PhD; Whiteneck, Gale PhD; Zasler, Nathan D. MD, FAAPMR, FAADEP, DAAPM, CBIST

Journal of Head Trauma Rehabilitation:
doi: 10.1097/HTR.0b013e3182738010
Original Articles

Objective: To characterize overall and cause-specific mortality and life expectancy among persons who have completed inpatient traumatic brain injury rehabilitation and to assess risk factors for mortality.

Design: Prospective cohort study.

Setting: The Traumatic Brain Injury Model Systems.

Participants: A total of 8573 individuals injured between 1988 and 2009, with survival status per December 31, 2009, determined.

Interventions: Not applicable.

Main Outcome Measures: Standardized mortality ratio (SMR), life expectancy, cause of death.

Results: SMR was 2.25 overall and was significantly elevated for all age groups, both sexes, all race/ethnic groups (except Native Americans), and all injury severity groups. SMR decreased as survival time increased but remained elevated even after 10 years postinjury. SMR was elevated for all cause-of-death categories but especially so for seizures, aspiration pneumonia, sepsis, accidental poisonings, and falls. Life expectancy was shortened an average of 6.7 years. Multivariate Cox regression showed age at injury, sex, race/ethnic group, marital status and employment status at the time of injury year of injury, preinjury drug use, days unconscious, functional independence and disability on rehabilitation discharge, and comorbid spinal cord injury to be independent risk factors for death.

Conclusion: There is an increased risk of death after moderate or severe traumatic brain injury. Risk factors and causes of death have been identified that may be amenable to intervention.

Author Information

Craig Hospital, Englewood, Colorado (Drs Harrison-Felix and Whiteneck and Mr Kreider); University of Deusto. Ikerbasque, Basque Foundation for Science, Bilbao, Spain (Dr Arango-Lasprilla); Mayo Clinic, Rochester, Minnesota (Dr Brown); Mount Sinai School of Medicine, New York, New York (Drs Dijkers and Hirshson); Carolinas Rehabilitation, Carolinas HealthCare System, Charlotte, North Carolina, and Indiana University School of Medicine, Indianapolis (Dr Hammond); Santa Clara Valley Medical Center, San Jose, California (Dr Kolakowsky-Hayner); and Concussion Care Centre of Virginia, Ltd and Tree of Life Services, Inc, and Virginia Commonwealth University, Richmond (Dr Zasler).

Corresponding Author: Cynthia Harrison-Felix, PhD, Research Department, Craig Hospital, 3425 South Clarkson St, Englewood, CO 80113 (

This work was supported, in part, by grants from the National Institute on Disability and Rehabilitation Research, Office of Special Education Services, Department of Education to Craig Hospital (H133A110006, H133A070022), Mount Sinai School of Medicine (H133A070033), Virginia Commonwealth University (H133A070036), Mayo Clinic (H133A070013), Santa Clara Valley Medical Center (H133A070038), and Carolinas HealthCare System (H133A070042). However, the contents of this manuscript do not necessarily represent the policy of the Department of Education and endorsement by the Federal Government should not be assumed.

The authors declare no conflicts of interest.

Article Outline

TRAUMATIC BRAIN INJURY (TBI) is a leading cause of death and disability in young people in the United States, with an estimated 52 000 deaths each year.1 It is important to accurately determine mortality rates and causes of death after TBI in order to appropriately estimate life expectancy and need for long-term care, monitor trends, and evaluate the impact of interventions.

An analysis of mortality and life expectancy using the prospectively acquired longitudinal TBI Model System National Database (TBIMS NDB) cohort in 2004 indicated that—compared with individuals in the general population of similar age, sex, and race—individuals with TBI were twice as likely to die and had an estimated average life expectancy reduction of 7 years.2 Recent studies examining life expectancy after TBI show high early postinjury fatality rates that appear to decrease over time (ie, after 6–12 months postinjury).3,4 Certain causes of death, such as seizures, septicemia, pneumonia, and respiratory conditions, are more common for persons with greater TBI severity.5 Other studies show a strong positive correlation between greater levels of immobility and functional dependency and risk of both morbidity and mortality.6 Factors that affect life expectancy to a lesser degree include mechanism of injury, comorbid injuries, sex, ethnicity, a history of substance use/abuse, and medical conditions, including poor general health, postinjury lifestyle, and insurance status.711

Variations in sample makeup and study methodologies often lead to contradictory findings. For instance, most of the literature indicates that older age at the time of injury and greater TBI severity are factors that affect life expectancy.3,7,1113 However, some studies have shown higher mortality in younger, not older, patients,4 whereas others have shown injury severity to be a relevant predictor only during the early postinjury phase.3,4 Standardized mortality ratios (SMRs) are quite variable across studies, also likely due to methodological differences.

The TBIMS NDB provides a unique resource to study individuals with TBI over time using prospectively acquired data. The goal of this analysis was to report mortality and cause of death in this ever-growing cohort since the initial report in 2004.2 The research questions for this analysis were as follows: (1) Does TBI increase mortality and decrease life expectancy in individuals with TBI who receive inpatient rehabilitation, relative to the general population? (2) In this TBI cohort, what are the characteristics that affect the risk of death? (3) Are individuals with TBI at greater risk of death due to certain causes, relative to the general population?

Back to Top | Article Outline


The HCA–HealthONE Institutional Review Board for Craig Hospital approved this study. Since 1987, the US Department of Education, National Institute on Disability and Rehabilitation Research, has funded the TBIMSs. One of the major components of the TBIMS program is a standardized National Database (NDB) for innovative analyses of TBI treatment and outcomes.14 The TBIMS NDB contains information on cases treated within the 20 TBIMS centers funded between 1988 and 2012, which are located around the United States. For the purposes of the NDB, TBI is defined as damage to brain tissue caused by an external mechanical force as evidenced by medically documented loss of consciousness or posttraumatic amnesia (PTA) due to brain trauma or by objective neurological findings that can be reasonably attributed to TBI on physical examination or mental status examination. Subjects included in the database, in addition, must (1) meet at least 1 of the following criteria for moderate to severe TBI: PTA greater than 24 hours, trauma-related intracranial neuroimaging abnormalities, loss of consciousness exceeding 30 minutes, or a Glasgow Coma Scale score in the emergency department of less than 13 (unless due to intubation, sedation, or intoxication); (2) be at least 16 years of age at the time of injury; (3) present to the Model System's acute care hospital within 72 hours of injury; (4) receive both acute hospital care and comprehensive rehabilitation in a designated brain injury inpatient unit within the Model System; and (5) provide informed consent to participate or have a proxy provide consent.15

For purposes of this study, the NDB as of March 10, 2011, was limited to 8573 individuals treated within 20 TBIMSs, with injury dates from October 25, 1988, through December 31, 2008, and follow-up dates or death dates through December 31, 2009. Information in the database was collected during acute care and rehabilitation hospitalization and at 1 year, 2 years, 5 years, and every fifth year thereafter at about the anniversary of injury. The rate of successful follow-up over all years of follow-up is around 80%.15

TBIMS staff attempted to contact individuals for interview; if they were told that the person had died, the death date was entered in the NDB. Center staff then attempted to obtain the death certificate, which was sent to the TBIMS National Data and Statistical Center, where causes of death listed on the death certificate were coded using the International Classification of Diseases, Ninth Revision (ICD-9) codes. If between follow-up interviews center staff learned that a participant had expired, the date of death was recorded in the NDB at that time point. If contact with the participant or a knowledgeable proxy at a routine follow-up was not made, center staff attempted to verify the vital status of participants using the Social Security Death Index (SSDI). If persons were not listed in the SSDI as deceased, and they could not be found, they were reported as lost to follow-up at that time point. Even if a person was reported as lost to follow-up at one scheduled follow-up interview, the staff would attempt to contact the individual at the next scheduled follow-up time point, often with success. For cases reported as lost as of their most recent follow-up (n = 1158 of 8573), the TBIMS staff for purposes of this study were asked to again look up cases in the SSDI to determine whether they were alive or had expired. This was done in an attempt not to bias the study due to loss to follow-up. For those previously lost cases that were found in the SSDI as expired, the vital status was updated in the study database and the center attempted to obtain the death certificate. Those previously lost cases that were not listed in the SSDI were presumed to be alive at the end of the study period (December 31, 2009).

The data analysis was performed using SAS software, Version 9.3 for Windows (SAS software, Version 9.3 of the SAS System for Windows; SAS Institute Inc, Cary, NC). Descriptive statistics (means for continuous variables and counts and proportions for categorical variables) were used to characterize the study sample. Survival was measured as indicated previously; only deaths occurring after discharge from initial inpatient rehabilitation and before January 1, 2010, were included.

The SMRs were calculated as the observed number of deaths in the study cohort divided by the expected number of deaths in similar individuals from the general population. The number of expected deaths was calculated by assigning the US mortality rate for a given calendar year for each year postinjury on the basis of sex, race, and age at each annual anniversary of injury. Once rates were assigned for each postinjury anniversary year for each case, these rates were summed for each case. Finally, the cumulated rate was summed over all cases in a population study cohort. For the purposes of this study, the US mortality rates published by the federal government16 for the calendar year 1999 (the median person-years of follow-up in the study) were used. Statistical significance of the SMR was determined by calculating its 95% confidence interval, which was considered significant if it did not contain 1.0.17

A cumulative conditional probability of survival curve was produced. Cumulative conditional probability of survival is based on the idea that the current probability of survival is conditioned on the fact that a person must have survived each of the previous time periods. That is, the survival curve is not simply a curve showing the proportion of people who remain alive; it includes multiple conditional probabilities of survival. For example, an individual's current probability of surviving is dependent on the fact that he or she survived up until now (similar to how the probability of flipping 3 heads in a row is 0.50 × 0.50 × 0.50 = 0.125 and not simply 0.50).

SMRs were also calculated for specific causes of death. The expected number of deaths in the absence of TBI was calculated using the methodology described previously but using the age-/sex-/race-specific mortality rates for each cause of death category.16 In 1999, the federal government transitioned from using ICD-9 to International Classification of Diseases, Tenth Revision (ICD-10), for coding causes of death. The cause of death mortality rates based on ICD-10 code categories were used to calculate expected deaths for the SMRs, and these were then matched to observed causes of death by ICD-9 code categories, because the National Data and Statistical Center had used the ICD-9 to code the death certificates (see Table 5 for a ICD-9/ICD-10 code category crosswalk).

Life expectancy with TBI by age, sex, and race was estimated by applying the overall SMR (for those with TBI) to the latest age-/sex-/race-specific mortality rates published by the federal government for the most recent year available at the time of this study (calendar year 2007), using the methodology described by DeVivo.17

Finally, to assess the independent impact of each potential mortality risk factor, Cox proportional hazards regression analysis was conducted taking into account survival time. A backward selection procedure was used to narrow the list of risk factors to only those that exhibited a significant relationship (P < 0.05) with the outcome of interest. Demographic factors assessed in the analyses were age at injury, sex, race/ethnic group, level of education, marital status, and employment status at injury. Preinjury drug use was also used as a factor. Injury-related risk factors included calendar year and cause of injury, Glasgow Coma Scale18 score at emergency department admission, days of PTA, days of unconsciousness, and whether the person also incurred a spinal cord injury (SCI) at the same time as the TBI. Other potential risk factors included acute care and rehabilitation length of stay, third-party sponsor (for acute and for rehabilitation care), place of discharge after rehabilitation, who the patient lived with at discharge from rehabilitation, Functional Independence Measure (FIM) scores (Motor, Cognitive, and Total),19 and Disability Rating Scale (DRS)20 scores at rehabilitation discharge. For the purposes of this study, DRS had a score range of 0 indicating no disability to 29 indicating extreme vegetative state.

Back to Top | Article Outline


For the 8573 individuals included in the study from the 20 participating TBIMSs, a total of 41 662 person-years of life with TBI were included for analysis. The length of follow-up in the study ranged from 1 day to 20.3 years beyond discharge from inpatient rehabilitation. There were 844 deaths, giving a mortality rate of 9.8%, with an average of 3.5 years from injury to death. Of the individuals who died, 193 (23%) did so between the time of rehabilitation discharge and 1 year postinjury anniversary.

In Table 1, the individuals who died are compared with those who were alive at the end of the study period. The average age at injury of study participants was 39 years; 74% were men and 67% were white. The majority (55%) were injured as a result of a motor vehicle crash, with 21% of injuries resulting from falls and 14% from acts of violence. Individuals in the study generally had moderate to severe TBI, with a median length of unconsciousness of 2 days and a median duration of PTA of 25 days. Individuals who died tended to be older, more likely to have had TBIs resulting from falls, and had longer hospitalizations but had a less severe injury, as indicated by higher average Glasgow Coma Scale scores. However, these comparisons may be misleading, in that the risk factors are not independent from one another—for instance, it may be that older people who incur a TBI in a fall have a less severe injury but die because of their age. The Cox regression analysis addresses these associations.

Table 2 contains the SMRs by study participant characteristics. Based on age-/sex-/race-specific mortality rates for the US general population, the expected number of deaths in the absence of TBI and given the length of time each person was followed was 376 for all individuals included in the study. Because 844 deaths were observed, the SMR was 2.25, indicating that individuals with TBI were more than twice as likely to die as individuals of similar age, sex, and race in the general population. As survival time increased, mortality ratios decreased but still remained elevated even after 5 and 10 years postinjury. As age at injury increased, SMRs decreased but still were elevated. Males had a greater excess mortality than females, and Asians had the highest excess mortality of any race/ethnicity category. Finally, the risk of death increased as injury severity increased. Figure 1 shows the overall cumulative conditional probability of survival. This graph indicates that individuals with TBI have a decreasing probability of survival as they progress further out from their injury date.

The estimated life expectancy in years (calculated under the assumption of a constant SMR of 2.25 in individuals with TBI) for various age, sex, and race groups with and without TBI is given in Table 3. Life expectancy was shortened between 3 and 11 years, depending on age at injury, race, and sex. For example, individuals injured at 15 to 19 years of age had their life expectancy shortened by 9 to 11 years, whereas individuals injured at 85+ years of age had their life expectancy shortened by 3 to 4 years. Within a given age-at-injury category, the years by which life expectancy was shortened (relative to persons without TBI) rarely vary by more than 1 or 2 years, depending on sex and race. On average, TBI reduced life expectancy in this cohort by 6.7 years.

The results of the multivariate Cox regression analysis to identify the independent risk factors for death after TBI are provided in Table 4. There was a 4% increase in risk of death for each additional year of age at injury. Females had a 60% lower risk of death than males. Hispanics had a 50% lower risk of death than whites. Those who were divorced or widowed at the time of TBI had a 33% greater risk of death than those who were married. Persons who were competitively employed at the time of injury had a lower risk of death than those who were unemployed, retired, or who reported another type of primary productive activity such as homemaking or volunteering. Individuals who used illicit or nonprescription drugs preinjury had a 33% greater risk of death than individuals who did not use drugs. There was a 6% increase in risk of death for every later calendar year of injury. Individuals whose TBI was the result of a fall or violent act were at greater risk of death than individuals injured as a result of a motor vehicle crash. Those who incurred an SCI at the same time as the TBI had a 52% lower risk of death than those without an SCI. There was a 1% lower risk of death for each additional day that individuals were unconsciousness after their TBI, and a 2% lower risk of death for every one-point increase in the discharge FIM motor subscale score (ie, higher motor functional independence). Finally, there was a 6% increase in risk of death for every 1-point increase in the DRS score (ie, greater disability).

Primary causes of death are provided in Table 5, as are the cause of death-specific SMRs for the causes accounting for the greatest proportion of deaths, or those identified in the literature as potentially being greater than expected for individuals with TBI. The largest number of deaths was secondary to circulatory conditions (22%), with the majority of those being due to ischemic or other heart disease. The next largest number of deaths was attributable to an external cause of injury (15%), with the greatest number being accidental poisonings (n = 32), followed by vehicular injuries (n = 25). The third largest group was secondary to respiratory conditions (13%), with half of those being due to pneumonia. In addition, 10% of all deaths were due to a neoplasm (predominantly lung cancer) and 7% were caused by infectious diseases (most due to sepsis). It is also of note that 2% of all deaths were seizure related and 2% were the result of suicide.

For all individual causes that were examined, the cause-specific SMR was greater than expected and statistically significant. Individuals with TBI were about 33 times more likely to die of seizures than comparable persons in the general population; 13 times more likely to die of aspiration pneumonia; 10 times more likely to die of sepsis, accidental poisoning, or falls; 6 times more likely to die of pneumonia; almost 4 times more likely to die of all external causes, all respiratory causes combined or homicide; almost 3 times more likely to die of mental/behavioral conditions or nervous system conditions; 2 times more likely to die of suicide, digestive conditions, or vehicular causes; and 1.3 times more likely to die of circulatory conditions.

Back to Top | Article Outline


Standardized mortality ratios overall and by subject characteristics

Consistent with previous studies,2,13,2123 our results indicate that individuals with TBI in this sample were more than twice as likely to die than individuals of similar age, sex, and race in the general population. For example, Ratcliff et al10 found that when compared with the Pennsylvania Vital Statistics Table, their sample of subjects injured 8 to 24 years before were 2.78 times more likely to die than the general population of Pennsylvania. Although slightly higher than the rate of 1.5 found by Harrison-Felix et al,24 these findings are significantly lower than those of Baguley et al,25 who found that individuals who are on average 10.5 years postinjury and functionally dependent after TBI are 13 times more likely to die than the general population. In 2012, Baguley et al21again found dramatically higher rates of mortality than we did in this study, 12 times greater than the general population of New South Wales.

In our sample, as age increased, SMRs decreased, with individuals aged 15 to 24 years nearly 4 times (SMR = 3.79) as likely to die as individuals of comparable age, sex, and race in the general population. This finding was similar to that of a previous study that reported that SMRs stratified by age increased until the age of 60 years and then decreased with successive age.13 Furthermore, although both sexes had a greater than expected risk of death, males fared worse than females, which is consistent with previous work suggesting that males are at risk for increased mortality rate post-TBI.2427 Similar to other findings,28 we found greater excess deaths in individuals of Asian descent with TBI than in the general population of similar age, sex, and race. However, among individuals with TBI, Asian descent was not a significant independent risk factor for death. Thus, it is the relatively low rates of death in the Asian general population that produce a high SMR for Asians with TBI, but the latter group's death rate is not significantly different from that of other ethnic/race categories within the TBI sample.

The accuracy of SMR estimates is increased with larger sample sizes and greater numbers of person-years. Among studies of mortality after TBI attempting to estimate SMRs, the current one (n = 8573) has more than double the sample size of any other investigation of rehabilitation inpatients, but it is eclipsed by the statewide population-based Colorado study of patients with TBI hospitalized in acute care facilities (n = 18 998).13 The current study with 41 662 person-years has also more than double the follow-up period of any other TBI inpatient rehabilitation study but still is smaller than the 83 268 person-years of the Colorado acute care study. The Craig Hospital study,24 with a maximum of more than 4 decades of follow-up (resulting in 16 599 person-years of data on 1678 rehabilitation inpatients), followed patients twice as long as the current study and 8 times longer than the Colorado study. Among these 3 relatively large (and presumably comparative) TBI mortality investigations, the SMRs are quite similar. The large Colorado study13 SMR for first-year survivors (1.71) is very close to the current study (1.79) and within the current study's SMR 95% confidence interval (1.65–1.93). Although the Craig Hospital Study24 SMR of 1.51 for first-year survivors is below that of the current study, that study had longer follow-up and its SMR was consistent with the current study's SMR for 5-year survivors (1.48).

As time passes, the rate of death relative to the general population (SMR) becomes smaller, not because the actual rate of death drops among people with TBI (it continues to increase with age) but because the rate of death also increases in the general population, and relatively faster. Also as time passes, the causes of death for those with TBI begin to approximate the causes of death in the general population.

The overall average SMR of this study (2.25, 95% CI, 2.10–2.40) is somewhat lower than the overall 2.47 SMR of the Colorado study.13 (The Craig Hospital Study did not report an overall SMR.) The difference in these SMRs can be explained primarily by the nature of the study populations and their inherent differences in brain injury severity.

Both the current and the Craig Hospital studies followed people with TBI initially identified as rehabilitation inpatients, whereas in the Colorado study, the initial identification occurred during acute care hospitalization. Rehabilitation inpatients are a subset of cases hospitalized in acute facilities. Most cases referred to inpatient rehabilitation are somewhat more severely injured, justifying rehabilitation. However, some cases hospitalized for acute TBI care are too severely injured to benefit from rehabilitation; it is these that have substantially higher mortality rates, which presumably accounts for the slightly higher SMR in the Colorado study. The difference between overall SMRs and those of first-year survivors suggests that a more intensive analysis of mortality in just the first year postinjury might yield insights into how risk factors and causes of death differ in the first year from later periods. In addition, the mild or nonhospitalized TBI population needs further investigation, because we know little about how these less severe TBIs affect mortality. The SMRs of this study (2.25 overall and 1.79 for first-year survivors) are somewhat higher than the SMRs in the study of the same TBI Model System population conducted 8 years earlier (2.00 overall and 1.91 for first-year survivors).2 A potential explanation may lie in the finding from this study that more recent calendar years of injury are associated with greater risk of death. Future monitoring of this finding is recommended to determine whether this might be an early indication of poorer outcomes associated with a rapidly changing service delivery system.

Back to Top | Article Outline
Life expectancy estimates

TBI reduced life expectancy rates by an average of 6.7 years in this sample, which is consistent with some previous research2,13 but still higher than the estimated rates of 3 to 5 years in other published studies.23,24,2934 The difference in life expectancy estimates across studies may also be explained by differences in the sample characteristics and follow-up time periods as indicated previously for differences in SMRs.

Back to Top | Article Outline
Causes of death

The most frequent causes of death were cardiovascular disease, external causes (accidental poisonings, vehicular injuries, homicide, and fall), respiratory disease (pneumonia, aspiration pneumonia), neoplasm, and infectious disease. This pattern resembles the prior findings in the literature. The high proportion of cardiovascular- and neoplasm-related deaths is similar to the findings of the Colorado population-based study,13 although that investigation found a higher percentage of cardiovascular deaths (35% compared with 22% in this study). In contrast with the Colorado study, this study revealed a much higher proportion of deaths that were due to external causes of injury, respiratory disease (including aspiration pneumonia and other types of pneumonia), infectious disease, and seizure. The proportion of deaths due to seizure, aspiration pneumonia, pneumonia, and sepsis was similar to the findings of the prior analysis of the TBI Model System and Craig Hospital cohorts.5,24 Death due to a digestive cause was less common (3%) in this study than in prior studies, with a report of 5% to 7%.5,24

Back to Top | Article Outline
Standardized mortality ratios by cause of death

The distribution of cases by cause of death only tells part of the story. For example, although circulatory disease was the most common cause of death (22%), the likelihood of dying due to circulatory etiology was not much higher than in the general population. This paradox is consistent with a prior study.13 Consideration of the proportion of cases by cause needs to include the SMR for cause of death in order to uncover the important areas deserving attention in clinical care and research. Causes of death following TBI that were much higher than expected included seizure, aspiration pneumonia, sepsis, pneumonia, external causes of injury (particularly, fall, suicide, accidental poisoning), respiratory disease, mental/behavioral disorder, nervous system, digestive disorder, and circulatory disease. These same causes were similarly more frequent than in the general population in prior studies.5,13

Seizure, found in this study to be 33 times more likely among individuals with TBI than in the general population, has been a consistent issue; prior studies found that individuals with TBI were 15 to 37 times more likely to die of seizure.5,13 Despite this consistent finding, there is still little known about the circumstances of seizure as a cause of death and what might be done to prevent these deaths. Early seizure was not assessed as an independent risk factor for death in this study. It is not known whether late seizures preceded the occurrence of the seizure that caused death. This information would have important implications for prevention and treatment directions. The current literature does not support the use of anticonvulsants in the prevention of seizures.35,36

Studies have consistently pointed to infections causing death following TBI, with higher than expected rates of septicemia, aspiration pneumonia, pneumonia, and other infectious causes. Aspiration pneumonia, at 13 times greater than in the general population, is an issue needing attention in TBI management. It is not known, in this study and prior studies demonstrating higher than expected deaths due to aspiration pneumonia,5,13 whether the individuals had experienced prior episodes of aspiration pneumonia, whether they were on a modified diet, or whether they had experienced neurologic decline putting them at new risk for aspiration. In prior work,5,13 individuals with TBI were 3 to 49 times more likely to die of aspiration pneumonia. In one study, it was noted that those who died of aspiration pneumonia were classified as in vegetative state or having severe disability at the time of rehabilitation discharge, and the average time from injury to death was 18 years. Pneumonia and respiratory diseases have also been previously reported as more common after TBI.5,32,37 This study as well as prior studies5 also found a higher than expected number of deaths resulting from sepsis.

External causes were found to be a more frequent cause of death in our TBI sample than in the general population, especially fall, accidental poisoning, homicide, vehicular injuries, and suicide. However, these areas have not received much clinical or research attention. External causes of death previously reported as higher after TBI than in the general population include drowning (20 times more likely)32 and choking and suffocation (30 times more likely).37 The current leading cause of TBI is falling,38 and after TBI one may be particularly at risk for a subsequent fall due to impairments of vision, sensation, awareness, attention, reaction time, judgment, impulse control, balance, muscle tone, or strength. Accidental poisonings are a worrisome cause of death and a potentially preventable cause that may point to a need for ongoing education and preventive measures. Further detail regarding the circumstances of death would be needed to understand these fatalities.

Death due to vehicular cause was 2.4 times more likely after TBI than in the general population, which may point to needs of additional education, assessment, and training. As with the other external causes, further information about the vehicular crashes would likely help identify the factors contributing to the injury, and such information is important for the design of preventive and intervention strategies. Behavioral and cognitive impairments following TBI (such as high-risk behavior, aggression, impaired judgment, altered cognition and awareness, and substance use/abuse) may place some individuals with TBI at greater risk for homicide.

Digestive conditions as a cause of death were almost twice as high as in the general population, which is consistent with other studies that found those with TBI to be 3 times more likely to die of a digestive condition.5,13 Further information about the digestive causes and the individual's medical history would be needed to use this finding to prevent such deaths. Important information for clinicians and researchers to consider would include the presence of alcohol use or abuse, liver disease, and medication effects.

Back to Top | Article Outline
Risk factors for death

Although the risk factors for death revealed in this study had some similarities to those of prior research, this study had some important differences. Earlier calendar year of injury was a risk factor in the Craig Hospital study, which went much further back in time5; however, the present study found those having a later year of injury to have greater risk. It is possible that with recent improvements in emergency and early acute care, those with TBI who previously might not have survived their injuries are now surviving long enough to receive inpatient rehabilitation, but some still succumb to the late effects of their injuries not long thereafter. In addition, over time, this study cohort is trending toward being older at injury and having lower independent functioning (lower FIM scores) and greater disability (higher DRS scores) at rehabilitation discharge, both of which are also risk factors for death.

Similar as in prior studies,2,5 those of older age were found to be at greater risk for death. This is also true for the general population; however, the SMRs by age provide an important insight. At all ages, individuals with TBI are at greater risk of death than the general population, though it is the younger age groups that were found to be at comparatively higher risk. Similar to the finding of prior studies,2,5 greater disability (DRS) increased risk and greater functional independence (FIM) decreased risk of death. Illicit and nonprescription drug use as a risk factor is a new important finding that may point to preventive and treatment needs that may decrease risk.

The findings of more days of unconsciousness and a co-occurring SCI both being protective factors for death after TBI seem counterintuitive. Further inspection of the data did not provide much insight. SCI was a rare co-occurring condition in the study cohort (4.5% of 8573 cases) and only 16 individuals with SCI died. The distribution of days of unconsciousness was highly skewed, with 29% of cases having less than 1 day of unconsciousness and a few cases having extremely long periods of unconsciousness (up to 235 days). Pure speculation may lead one to think that perhaps more intense medical care provided, at least initially after TBI, may prevent or postpone early death. However, given the nonnormal distribution of these variables in the cohort, this result should be considered with caution.

Back to Top | Article Outline


The strengths of this study were that it was a multicenter study of individuals receiving inpatient rehabilitation following TBI, with a relatively large sample size, a fairly long average period of follow-up, a high rate of vital status follow-up, using the TBIMS NDB including an extensive amount of demographic, injury characteristic, and outcome data. Furthermore, due to the magnitude and comprehensive nature of the TBIMS NDB, this study was able to include a large number of important risk factors for death after TBI. Given that the individuals included in the study were all treated at TBIMSs, 1 study limitation is that the outcomes may not be generalizable to individuals with TBI receiving inpatient rehabilitation treatment elsewhere. However, Corrigan et al39 found that TBIMS NDB is representative of patients receiving inpatient rehabilitation for TBI in the United States, when one takes into account the lower percentage of patients older than 65 years or those with stays fewer than 10 days. Another limitation to consider is that these findings represent a select group of individuals who survived injury and acute care hospitalization and received inpatient rehabilitation. This study did not include those who did not require hospitalization or did not receive inpatient rehabilitation. Further research should be conducted on the additional risk factors for death among individuals with TBI who do not seek or receive treatment past the emergency department or qualify for inpatient rehabilitation services. In addition, it is common for individuals with TBI for whom inpatient rehabilitation is indicated to have acquired injuries to other organ systems at the time of TBI that may affect survival. Using the SMR for comparison does not allow taking this into account, and this may affect the accuracy of these results. Fifteen percent of the deaths identified in this study were missing the cause of death information, which could result in an underestimation of the prevalence of some causes. The risk factors examined in this study were those available only up to discharge from inpatient rehabilitation. As such, future studies should include an examination of risk factors closer to the date of death. In addition, future studies should examine factors such as the general lifestyle indicators that have been found to be highly associated with risk of death in the general population (eg, smoking, alcohol, diet, exercise).

Back to Top | Article Outline


This study has shown that moderate and severe TBI puts individuals at risk for early death due to both typical and unique causes. Examining patient characteristics by cause of death may shed further light on how, for a number of TBI-associated disorders, prevention, surveillance, and treatment may need to depart from usual care of the general population. Increasingly, TBI is being recognized as a chronic medical condition.40,41 This concept is underscored by the findings of this and prior studies on mortality following TBI. Similar to the general population, individuals with TBI are in need of routine medical care. However, individuals with TBI experience medical, emotional, behavioral, and cognitive challenges that pose unique health demands. Continued research is needed to learn what unique challenges exist and how to best prevent or treat them. This study adds to the literature on mortality, life expectancy, and risk factors for death and causes of death after moderate to severe TBI. The results can aid in long-term life planning, resource allocation, and prevention of untimely death in individuals after TBI.

Back to Top | Article Outline


1. Coronado VG, Xu L, Basavaraju SV, et al. Surveillance for traumatic brain injury-related deaths—United States, 1997–2007. MMWR Surveill Summ. 2011;60(5):1–32.
2. Harrison-Felix C, Whiteneck G, DeVivo M, Hammond FM, Jha A. Mortality following rehabilitation in the Traumatic Brain Injury Model Systems of Care. NeuroRehabilitation. 2004;19(1):45–54.
3. McMillan TM, Teasdale GM, Weir CJ, Stewart E. Death after head injury: the 13 year outcome of a case control study. J Neurol Neurosurg Psychiatry. 2011;82(8):931–935.
4. McMillan TM, Teasdale GM. Death rate is increased for at least 7 years after head injury: a prospective study. Brain. 2007;130(pt 10):2520–2527.
5. Harrison-Felix C, Whiteneck G, Devivo MJ, Hammond FM, Jha A. Causes of death following 1 year postinjury among individuals with traumatic brain injury. J Head Trauma Rehabil. 2006;21(1):22–33.
6. Baguley IJ, Nott MT, Howle AA, et al. Late mortality after severe traumatic brain injury in New South Wales: a multicentre study. Med J Aust. 2012;196(1):40–45.
7. Colantonio A, Escobar MD, Chipman M, et al. Predictors of postacute mortality following traumatic brain injury in a seriously injured population. J Trauma. 2008;64(4):876–882.
8. Egede LE, Dismuke C, Echols C. Racial/ethnic disparities in mortality risk among US veterans with traumatic brain injury. Am J Public Health. 2012;102(suppl 2):S266–S271.
9. Pentland B, Hutton LS, Jones PA. Late mortality after head injury. J Neurol Neurosurg Psychiatry. 2005;76(3):395–400.
10. Ratcliff G, Colantonio A, Escobar M, Chase S, Vernich L. Long-term survival following traumatic brain injury. Disabil Rehabil. 2005;27(6):305–314.
11. Selassie AW, McCarthy ML, Ferguson PL, Tian J, Langlois JA. Risk of posthospitalization mortality among persons with traumatic brain injury, South Carolina 1999–2001. J Head Trauma Rehabil. 2005;20(3):257–269.
12. Cameron CM, Purdie DM, Kliewer EV, McClure RJ. Ten-year outcomes following traumatic brain injury: a population-based cohort. Brain Inj. 2008;22(6):437–449.
13. Ventura T, Harrison-Felix C, Carlson N, et al. Mortality after discharge from acute care hospitalization with traumatic brain injury: a population-based study. Arch Phys Med Rehabil. 2010;91(1):20–29.
14. Dijkers MP, Harrison-Felix C, Marwitz JH. The Traumatic Brain Injury Model Systems: history and contributions to clinical service and research. J Head Trauma Rehabil. 2010;25(2):81–91.
15. Traumatic Brain Injury Model Systems National Data and Statistical Center (TBINDSC). Accessed May 15, 2012.
16. CDC WONDER. Accessed May 15, 2012.
17. DeVivo MJ. Estimating life expectancy for use in determining lifetime costs of care. Top Spinal Cord Inj Rehabil. 2002;7(4):49–58.
18. Teasdale G, Jennett B. Assessment and prognosis of coma after head injury. Acta Neurochir (Wien). 1976;34(1–4):45–55.
19. Guide for the Uniform Data Set for Medical Rehabilitation (Including the FIM Instrument). Version 5.1 ed. Buffalo, NY: State University of New York at Buffalo; 1997.
20. Rappaport M, Hall KM, Hopkins K, Belleza T, Cope DN. Disability Rating Scale for severe head trauma: coma to community. Arch Phys Med Rehabil. 1982;63(3):118–123.
21. Baguley I, Slewa-Younan S, Lazarus R, Green A. Long-term mortality trends in patients with traumatic brain injury. Brain Inj. 2000;14(6):505–512.
22. Shavelle R, Strauss D. Comparative mortality of adults with traumatic brain injury in California, 1988–97. J Insur Med. 2000;32(3):163–166.
23. Strauss DJ, Shavelle RM, Anderson TW. Long-term survival of children and adolescents after traumatic brain injury. Arch Phys Med Rehabil. 1998;79(9):1095–1100.
24. Harrison-Felix CL, Whiteneck GG, Jha A, DeVivo MJ, Hammond FM, Hart DM. Mortality over four decades after traumatic brain injury rehabilitation: a retrospective cohort study. Arch Phys Med Rehabil. 2009;90(9):1506–1513.
25. Baguley IJ, Nott MT, Slewa-Younan S. Long-term mortality trends in functionally-dependent adults following severe traumatic-brain injury. Brain Inj. 2008;22(12):919–925.
26. Koskinen S, Alaranta H. Traumatic brain injury in Finland 1991–2005: a nationwide register study of hospitalized and fatal TBI. Brain Inj. 2008;22(3):205–214.
27. Conroy C, Kraus JF. Survival after brain injury. Cause of death, length of survival, and prognostic variables in a cohort of brain-injured people. Neuroepidemiology. 1988;7(1):13–22.
28. Berry C, Ley EJ, Mirocha J, Salim A. Race affects mortality after moderate to severe traumatic brain injury. J Surg Res. 2010;163(2):303–308.
29. Corkin S, Sullivan EV, Carr FA. Prognostic factors for life expectancy after penetrating head injury. Arch Neurol. 1984;41(9):975–977.
30. Lewin W, Marshall TF, Roberts AH. Long-term outcome after severe head injury. Br Med J. 1979;2(6204):1533–1538.
31. Rish BL, Dillon JD, Weiss GH. Mortality following penetrating craniocerebral injuries. an analysis of the deaths in the Vietnam Head Injury Registry Population. J Neurosurg. 1983;59(5):775–780.
32. Roberts A. Life expectancy and causes of death. In: Roberts A, ed. Severe Accidental Head Injury: An Assessment of Long-Term Prognosis. London: Macmillian Press Ltd; 1979:140–179.
33. Shavelle RM, Strauss D, Whyte J, Day SM, Yu YL. Long-term causes of death after traumatic brain injury. Am J Phys Med Rehabil. 2001;80(7):510–516; quiz 517–519.
34. Strauss DJ, Shavelle RM, Ashwal S. Life expectancy and median survival time in the permanent vegetative state. Pediatr Neurol. 1999;21(3):626–631.
35. Temkin NR, Dikmen SS, Wilensky AJ, Keihm J, Chabal S, Winn HR. A randomized, double-blind study of phenytoin for the prevention of post-traumatic seizures. N Engl J Med. 1990;323(8):497–502.
36. Chang BS, Lowenstein DH. Quality Standards Subcommittee of the American Academy of Neurology. Practice parameter: antiepileptic drug prophylaxis in severe traumatic brain injury: report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology. 2003;60(1):10–16.
37. Shavelle RM, Strauss DJ, Day SM, Ojdana KA. Life expectancy. In: Zasler ND, Katz D, Zafonte R, eds. Brain Injury Medicine. Principles and Practice. New York, NY: Demos Publishers; 2007:247–261.
38. CDC—leading causes and risk groups—traumatic brain injury—injury center. Accessed May 21, 2012.
39. Corrigan JD, Cuthbert JP, Whiteneck GG, et al. Representativeness of the Traumatic Brain Injury Model Systems National Database. [published online ahead of print September 2, 2011]. J Head Trauma Rehabil. doi: 10.1097/HTR.0b013e3182238cdd
40. Kobeissy FH, Guingab-Cagmat JD, Razafsha M, et al. Leveraging biomarker platforms and systems biology for rehabilomics and biologics effectiveness research. PM R. 2011;3(6 suppl 1):S139–S147.
41. Masel BE, DeWitt DS. Traumatic brain injury: a disease process, not an event. J Neurotrauma. 2010;27(8):1529–1540.

brain injury; chronic; epidemiology; life expectancy; mortality; rehabilitation; vital statistics

© 2012 Lippincott Williams & Wilkins, Inc.