Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY This chapter reviews the scientific literature on the epidemiology of traumatic brain injury (TBI) and on incidence, prevalence, severity, external causes, risk factors, mortality, case- fatality rates, and disability estimates among others. For purposes of this chapter, the papers reviewed were published in 1980 or later, focus on adults only, include incidence reports, and use established methods that resulted in a minimum of sources of bias and misclassification similar to criteria established by the committee for review of studies of long-term health outcomes. Review articles are included only as a source of reference. Scales and Scoring Systems Used to Describe Traumatic Brain Injury There are many classifications of TBI; for example, Knightly and Pulliam (1996) address the various components of TBI incurred in the military. As noted in Chapter 1, there are two basic types of head injury: closed and open. Closed head injuries result from the concussive effects of such exposures as munitions explosions, falls, and deceleration injuries from vehicular crashes; the latter two have also been termed blunt-force injuries. Open head injuries include those caused by penetrating forces, for example, from gunshots or shrapnel, or by impaling forces, such as from knives (see also Chapter 2). Gross Severity of Traumatic Brain Injury Different methods have been used in the last three decades to measure the magnitude of brain damage and to predict the outcome of injuries (see Chapter 2). The mostly widely used tool for measuring severity is the Glasgow Coma Scale (GCS), which was developed in 1974 by Teasdale and Jennett (1974) as a measure of neurologic deficits after TBI and was an important contribution to the standardization of early assessment of TBI. It uses a simple method of scoring three domains: eye opening, verbal response, and motor function (Table 3.1) and yields a total score of from 3 (comatose or nonresponsive) to 15 (no deficits in any of the three domains). The interpretation of scores at the ends of the scale is relatively straightforward, but scores like 8 or 9 or 11 or 12 may be subject to judgment error. Although the GCS is relatively straightforward in its numeric results, the classification of severity has been inconsistent. Many incidence studies have classified severity according to GCS scores of 3â8, severe; 9â12, moderate; and 13â15, mild or minor (e.g., Kraus et al., 1984; Thurman et al., 1996; Langlois et al., 2003). Permutations of that approach are summarized in Table 3.2 (US studies) and Table 3.3 (non-US studies). 59
60 GULF WAR AND HEALTH It was originally intended that the GCS would be applied repeatedly during a patientâs hospital course to monitor improvement or deteriorationâduring emergency transport, in the emergency department (ED), during intensive care, and throughout primary care. Because GCS scores have been reported in almost all recent studies of TBI severity, it is important to compare only readings taken at similar times after injury among studies. The most common time for determining the GCS score is 6 hours after injury, which is generally when the patient is in the ED. The GCS is subject to limitations when used on some patients, such as young children, people with extensive facial injuries that would preclude eye assessment, people subject to cross- language misunderstandings, and people who have been intubated or sedated on arrival at the ED. A major limitation of the GCS is the effect of intoxication. As many as 35â50% of adult civilian patients transported to the ED may be under the influence of alcohol (Jagger et al., 1984a), so its effect on the GCS score and its interpretation cannot be ignored. A study by Sperry et al. (2006), however, suggests that alcohol intoxication had little effect on the GCS. Nell and associates (2000) introduced an extended version of the GCS (GCS-E) to address difficulties of its application to the mild forms of TBI (Table 3.1). A study by Drake and colleagues (2006) showed that the extended GCS is a useful tool for the prediction of symptoms connected with mild TBI. The GCS should not be confused with the Glasgow Outcome Scale (GOS) (Table 3.1). The GCS is a physiologic measure of consciousness and the GOS is a gross measure of complications or residual effects following severe brain injury (Jennett and Bond, 1975). Other methods and instruments have been used as alternatives to the GCS, such as the Abbreviated Injury Scale (AIS) (see Chapter 2) and the International Classification of Diseases (ICD). Clinical measuresâsuch as loss of consciousness (LOC) and duration of posttraumatic amnesia (PTA)âand computed tomography of brain lesions have also been used to assess TBI severity. Table 3.2 shows examples of TBI incidence studies conducted in the United States that used those measures. As can be seen there is no consistency in severity classification systems reflecting available clinical symptoms or evidence from neuroimaging. A review of popular injury scales can be found in the review by MacKenzie (1984). Outcome Scores and Predictors The literature is replete with attempts to predict TBI outcomes on the basis of measurements in the ED or soon after intensive care. One of the most commonly used measures is the GOS (Table 3.1) developed by Jennett and Bond (1975). Although the intent of the GOS was to address severe TBI, it has been applied over the years to less severe TBI. It is acknowledged as a crude measure of medical (neurologic) complications and sequelae but has found favor as a quick and reliable indicator of outcome especially of severe TBI (Teasdale et al., 1998). The GOS is most commonly applied at 3, 6, or 12 months postinjury but can be used at any time after intensive care. Pettigrew and associates (2003) recently showed that the GOS can be successfully applied over the telephone. There are many other measures, but only the GOS is covered in this chapter to assess patient disposition at hospital discharge. Because the gross categories of the GOS have some limitations, an extended version (the GOS-E) was developed (Jennett et al., 1981); the GOS-E adds three categories to the GOS and has good inter- rater agreement.
EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY 61 INCIDENCE OF TRAUMATIC BRAIN INJURY Incidence is the number of newly diagnosed cases occurring in a defined period, usually expressed with reference to a base of 100,000 persons. An incidence study is one in which only newly diagnosed TBI cases in a specified period of time in a population of known size have been enumerated and are included in the study group. Some 30 population-based TBI incidence studies conducted in the United States have been published since 1980 (Table 3.4). Early studies were limited to counties (Kraus et al., 1984), cities (Cooper et al., 1983; Whitman et al., 1984), and states, such as Oklahoma, Massachusetts, Louisiana, and Alaska. National or subnational estimates of the incidence of TBI have recently been published from the Centers for Disease Control and Prevention (CDC) TBI surveillance system (Langlois et al., 2003) or from existing national administrative datasets (Langlois et al., 2006). Methods used for incidence studies have varied. For example, some earlier studies (e.g., Rimel, 1981; Kraus et al., 1984) relied on hospital or coroner records for case findings based on discharge codes, reviewed original institutional records, and abstracted pertinent data. Later studies used hospital discharge records and electronic files; in a few instances, a trauma registry was the source of data on TBI (Warren et al., 1995). On the basis of the data available from those studies, the incidence of hospitalizations for TBI in the United States is about 140/100,000 persons per year. If the highest reported rate (367/100,000) and the lowest reported rate (69/100,000) are excluded, the average rate of hospitalization for TBI (plus cases of immediately fatal TBI) in the United States is about 130/100,000 persons per year. Those estimates do not include ED-based studies, with rates of 444/100,000 (Jager et al., 2000) or 392/100,000 (Guerrero et al., 2000). The rates given in Table 3.4 represent three case-finding methods: hospitalized cases and those identified from medical- examiner records, hospital discharge records only, and trauma-registry files. The differences in case-finding approaches and other methodologic differences result in different rates. Some 36 TBI incidence studies conducted outside the United States have been published since the middle 1970s, most coming from Europe and Australia (Table 3.5). As in the US studies, a wide variety of methods were used in TBI case definition and ascertainment methods. Even when ICD TBI codes were used in existing hospital electronic discharge files, the codes selected were not uniform. About half the incidence studies did not evaluate TBI severity. Therefore, it is difficult to synthesize findings from the non-US studies. Time Trends in Incidence Few incidence studies have collected data beyond a single year or two. MacKenzie and associates (1990) reported an increase in TBI incidence in Maryland from 1979 to 1986. There did not appear to be any remarkable changes in TBI identification procedures in the stateâs database, and only patients admitted to the stateâs 56 acute-care nonfederal hospitals were counted. Using the US National Hospital Discharge Survey, a yearly survey sampled in such a way as to be representative of the US general population, Thurman and Guerrero (1999) reported a 51% decline in incidence from 1980 through 1995. The change over that period was from 199/100,000 to 98/100,000. They noted that the TBI-associated death rate also declined, possibly because of the preventive measures associated with motor-vehicle crash outcomes. The authors noted also that the greatest change in hospitalization rates was in those with mild TBI; that suggested a change in hospital admission practices.
62 GULF WAR AND HEALTH Time-trend studies of incidence are rare in Europe and nonexistent in Asia. Engberg and Teasdale (2001), in an analysis of 1979â1996 data from Denmark, reported an overall decline of 41% in the rate of hospitalization for TBI. The percentage decline varied with ICD code. The authors speculate that the decreases may be explained by changes in hospital admission practices and the possible effect of national prevention programs. Kleiven and associates (2003) observed a varied change in incidence in Sweden from 1987 to 2000: persons over 85 years old appear to have experienced an increase in TBI rates and younger persons a decrease. Mortality The most recent estimates in four US reports indicate an average of about 50,000 deaths each year with TBI-related causes (Table 3.6). The most recent reported TBI mortality in the United States is 17.5/100,000 persons per year (Rutland-Brown et al., 2006). Sixteen incidence reports provide mortality data on subsets of the US population. The years in question are from 1974 through 2003, and the rates vary from 17/100,000 per year to 30/100,000 per year. The large US studies are based on data from the National Center for Health Statistics, and the rates for the latest years are tightly clustered from 17/100,000 per year to 21/100,000 per year. It should be kept in mind that methods of collecting mortality data vary somewhat, but the rates in most studies are based on death-certificate review. TBI mortality in non-US countries varies much more widely than that in the United States (Table 3.7). The lowest rate reportedâ5.2/100,000 of populations in Aquitaine, Franceâ reflects only inpatient deaths (Masson et al., 2001). Low rates have also been reported in northeast Italy, South Australia, and Norway. The highest TBI-death rates on record are in Johannesburg, South Africa (81/100,000), and Hualien County, Taiwan (82/100,000) (Nell and Brown, 1991; Chiu et al., 1997). Prevalence of Traumatic Brain Injury (Disability) Prevalence reflects the total number of cases of TBI at a specified point in time and includes all newly diagnosed patients plus those persons with residual physical and neuropsychologic problems. It should not be confused with incidence. Prevalence is a measure of the cumulative occurrence of TBI in the population at the point or period when measured. It is difficult to determine the exact prevalence of TBI in the United States, but there are estimates of disabilityâphysical, mental, or social impairmentâas a result of TBI. The literature on TBI disability is large and is based on occurrence of disability in a group of persons who have survived and might not be representative of the entire population. For purpose of this chapter, two recent US studies are highlighted because their findings were based on original incidence cohorts with excellent followup methods to ascertain outcomes. From 1996 to 1999, 2,771 Colorado residents 16 years old or older were discharged alive from an acute-care hospital after a diagnosis of moderate or severe TBI (Whiteneck et al., 2004). After multiple attempts at contact, 1,591 were located and responded to structured interviews on a variety of outcomes. Information was sought 1 year after injury on health-service use, the Functional Independence Measure, the Craig Handicap Assessment and Reporting Technique (CHART), and a single question on quality of life. The study authors noted that 65% had problems in cognition (any symptom); 71% used at least one service after injury; 15â37% had activity limitations, depending on the type of activity; and 24% failed to return to work. With
EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY 63 regard to CHART components, 16% were impaired or disabled; and 29% reported less than good quality of life. The authors concluded that âsubstantial percentages of people hospitalized with TBI in a population-based sample reported a variety of problematic outcomes at 1 year postinjury.â It is noteworthy that the problems experienced by members of that injury cohort were in many cases similar in all levels of initial TBI severity. The second study of the incidence of disability was a South Carolina population-based prospective cohort study reported by Pickelsimer et al. (2006). Followup was completed at 1 year after injury with such outcome measures as service needs, psychosocial health, health-related quality of life, functional status, TBI-related symptoms, employment, and life satisfaction. Outcome findings included one or more functional limitations in about 47% of the subjects, unmet service needs in about one-third (35%), and dissatisfaction in quality of life in about one- third (35%). In a second report of that study, Selassie et al. (2008) used the same population sample and outcome measures to estimate the incidence of long-term disability in the United States. The researchers concluded that among the 288,009 survivors hospitalized for TBI in 2003, almost 125,000 (over 43%) had long-term disability. The disability rate varied by age and sex; it was higher in females than in males and was highest in people who fell and in those with self-inflicted injuries. Information on annual disability does not quantify the cumulative prevalence of TBI disability or impairment in the population. If 43% of a hospital-discharged TBI population sustains some form of disability or impairment in 1 year, the question remains, how much of the total population is disabled or impaired from TBI sustained in earlier years? Thurman and colleagues (1999) attempted to estimate that number by using the US National Hospital Discharge Survey data to approximate incidence and then classified the data by TBI severity by applying the computer algorithm known as ICDMAP-90 developed by MacKenzie et al. (1989b). The probability of disability was estimated for each level of severity by using outcome findings on disability from the Colorado state TBI registry and estimation parameters developed by Kraus and McArthur (1996). On the basis of that model, CDC estimated that 5.3 million US citizens (2%) were living with TBI-related disability in 1996. If that proportion is applied to the 2007 US population of over 301 million people, then just over 6 million people are living with the effects of TBI, and 2 million people have unmet health-service needs. BRAIN INJURY SEVERITY As discussed above, LOC is the most common measure used for evaluating brain injury severity and the most widely used tool for LOC is the GCS. Problems that arise in comparing the GCS measured in different places come from differences in timing. For example, intubation and sedation of the brain-injured patient during transport to the ED will obviously affect the person's verbal and motor abilities and eye responses. Differences in timing in the administration of any measurement tool can be critical so Teasdale and Jennett suggested that the GCS be applied at 6 hours post-injury. However, because a patient's injury may require ED procedures like intubation and sedation or acute surgical intervention, repeat measures may be necessary, often minutes or hours apart. Hence there does not appear to be an ideal window that is the best for the GCS, but, if it is to have any predictive quality, it should be applied early in the clinical management of TBI.
64 GULF WAR AND HEALTH Severity Distributions The distribution of severity of brain injury as assessed by the GCS (or other parameters) is shown in Table 3.8. Of the more than 60 population-based incidence studies published worldwide since 1980 only 22 address the degree of TBI severity in the study populations; 10 are from US and 12 from non-US countries. Most studies used the GCS to evaluate brain injury severity but some also used the AIS. The majority of hospital-admitted brain injuries are classified as "mild" (generally, a GCS score of 13 to15 or AIS of 1 or 2). However, the mild category is viewed differently by different researchers some of whom use mild to describe any GCS score above 7 while others refer to GCS scores above 8, above 10, above 13, or 15 only (Kraus and Chu, 2004). Studies published in the 1980s, with the exception of the report from Oklahoma, showed a ratio of mild to moderate to severe of about 8:1:1. With one or two exceptions almost all studies in the United States show less than 20% of patients admitted to a hospital are in the severe TBI range, and mild TBI is diagnosed 60% or more of the time. However, researchers outside the US report severity distribution proportions at even more consistent levels. A study by Hillier et al. (1997) evaluated TBI severity using three different measures: the GCS, LOC, and PTA; results were very similar, which provides support to the acceptance of severity classification when each of those measures is used. Severity distribution findings from non-US studies (Table 3.8) are similar to those from the United States with a ratio of mild to moderate to severe of 7:1:1. The high percentage of severe TBI admissions for Northeast Italy (Baldo et al., 2003) and the Romagna region of Italy (Servadei et al., 2002a) may reflect the referral practice of the acute medical care treatment institutions involved. RISK FACTORS FOR TRAUMATIC BRAIN INJURY Several risk factors have been examined in connection with the incidence of TBI: age, sex, ethnicity, and socioeconomic status. Data on age and sex in TBI can be found in 60 of the 66 papers reviewed (Tables 3.9 and 3.10). Although the papers do not necessarily group ages similarly, findings are remarkably consistent; the age group with the highest incidence of TBI is 15â24 years. In some reports, age groups at highest risk depend on TBI severity. For example, the very young (0â4 years old) and the very old (at least 85 years old) present to an ED with a brain injury most frequently, whereas those 15â24 years old and over the age of 65 years are hospitalized with TBI most frequently. The age-specific rates tend to reflect differences in exposure, particularly to motor-vehicle crashes and falls. Males are at greater risk for TBI than females at all ages in all incidence studies. Every report that gives data on sex-specific incidence shows that males have much higher TBI rates than females, and the ratio of male incidence to female incidence often exceeds 2. In one report (Nell and Brown, 1991), the incidence ratio of males to females exceeded 4 in both blacks and whites in Johannesburg, South Africa. The researchers posit that men in Johannesburg are involved in much higher levels of aggressive activities than women in the same city. The sex-specific mortality ratio is about 3.5:1, strongly indicative of more severe injuries among males (Adekoya et al., 2002). The US TBI death rate in 1989â1998 averaged 27/100,000 in American Indians and Alaskan Natives, 25/100,000 in blacks, and 20/100,000 in whites (Adekoya et al., 2002). The nonfatal-TBI hospitalization rate in 1997 (based on a 14-state surveillance system) was
EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY 65 74/100,000 in blacks, 75/100,000 in American Indians and Alaskan Natives, and 63/100,000 in whites (Langlois et al., 2003). ED incidence studies of TBI show similar results, albeit often lacking complete racial and ethnic categories. For example, the report by Jager and associates (2000) shows the rate of ED-treated TBI in blacks as 582/100,000 and the rates in whites and all others as 429/100,000 and 333/100,000, respectively. Data from the US National Health Interview Survey for 1991 (Sosin et al., 1996) show that whereas ED TBI rates were higher in whites than in blacks or Hispanics, the TBI hospital-admission rates were the opposite, that is, lower in whites than in other race and ethnic groups. Similarly, Nell and Brown (1991), in a 1986 TBI study in Johannesburg, South Africa, reported an incidence 3.3 times higher in blacks, 1.9 times higher in Asians, and 2.7 times higher in coloureds (mixed race) than in whites. Higher risk of TBI is often associated with lower socioeconomic status (SES) because there might be increased exposure to physically demanding or unsafe employment settings, increased exposure to violence, or increased exposure to less well-maintained residences or older vehicles without newer safety features (Hoofien et al., 2003). In the United States, families at the lowest income levels have been shown to incur the highest numbers of injuries of all types on a per capita basis (Collins, 1990). That was found to be true for TBI in a study of San Diego County residents (Kraus et al., 1986), in two Chicago communities (Whitman et al., 1984), and in Rhode Island (Fife et al., 1986). The San Diego County study demonstrated that the link between injury and low SES was not modified when the analysis controlled for race or ethnicity. Two more recent studies have demonstrated the link between the incidence of TBI and race or SES. Selassie et al. (2003, 2004) in a large cohort study of TBI in South Carolina showed that the disposition of TBI patients from the ED might be influenced by insurance status and other factors. Furthermore, black females and the uninsured were less likely to be hospitalized for TBI after adjustment for important confounders. However, Yates et al. (2006) determined hospital TBI âattendanceâ rates in an ED in a large UK population from 1997 to early 2003 by using the Index of Multiple Deprivation and noted that the highest TBI attendance rates were in groups with the lowest SES. Alcohol consumption can disrupt brain activity. Intoxication greatly increases various risks, including risks posed by motor-vehicle operation and the risk of self-inflicted injury and assault (e.g., Waller et al., 1986; Modell and Mountz, 1990). Also, intoxication can complicate diagnosis in the ED by increasing LOC independently of brain-injury severity (e.g., Jagger et al., 1984a). The association between blood alcohol concentration (BAC) and risk of TBI is well established for all external causes, such as motor-vehicle crashes, violence, and even falls. One of the earliest incidence reports on TBI and alcohol involvement was by Rimel (1981), who showed that 72% of patients identified in a central Virginia TBI databank had positive BAC rates on admission and 55% were legally drunk (BAC, 0.10% or higher). Kraus et al. (1989) reported in a TBI incidence study of San Diego County residents in 1981 that 49% of those who were tested for BAC had a BAC of 0.10% or higher (which is either an offense itself or presumptive evidence of driving under the influence). Langlois et al. (2003), reporting on a 14-state TBI surveillance system in 1997, found that 43% of those (including pedestrians) who sustained TBI in motor-vehicle crashes had a BAC of 0.10% or higher; the percentages of those who sustained TBI in falls and assaults were 7% and 28%, respectively. Warren et al. (1995), in a study of TBI in Alaskan residents, reported that almost 67% of those tested had a BAC of 0.10% or higher. Findings like those are not peculiar to the United States. A few non-US TBI incidence studies show evidence of alcohol use. Chiu and colleagues (2007) found that 15% of adults who
66 GULF WAR AND HEALTH sustained TBI in 2001 in Taipei City, Taiwan, used alcohol before the injury incident compared with 42% in Hualein County in the southern part of Taiwan. Researchers in Spain (Vazquez- Barquero et al., 1992) reported that 55% of males and 40% of females with TBI who presented for admission were intoxicated. Similarly, positive BAC rates were reported by Nestvold et al. (1988) and Ingebrigtsen et al. (1998). Simpson and co-workers (1981) reported that among those who died of TBI in New South Wales and were tested, 44% of motor-vehicle drivers, 39% of suicides, and 27% of people who sustained TBI in falls had BAC of 80 mg% or higher. The percentages of people who were tested for BAC in the last four reports were not reported. External Causes of Traumatic Brain Injury Only about half the 66 US and non-US studies reporting the incidence of TBI give details on the exposures that led to it. Data from those studies (Tables 3.11 and 3.12) suggest that the most frequent exposure associated with brain injury is transportation. That category includes automobile and truck occupants, bicycles and motorcycle riders, and pedestrians hit by vehicles and, less frequently, aircraft, watercraft, and road farm equipment. One precaution in discussing reported external causes is that the specific components of each of the general categories are not always uniform. For example, TBI stemming from bicycleâmotor-vehicle collisions may be classified as âmotor vehicleâ or âsports or recreation,â depending on the inclination of the researcher. As can be seen from Tables 3.11 and 3.12, the distributions of those gross external causes can vary widely among studies, but they do illustrate vast differences within a general cause. For example, in the two US ED-based studies (Guerrero et al., 2000; Jager et al., 2000), the most important exposure reported is falls, compared with hospital-based studies, in which transportation is the most frequent cause of brain injury. But in two US studies and the study in South Africa, the most frequently reported external cause is violence (which includes the use of firearms and self-inflicted injury); in these studies, incidence was determined on the basis of inner-city populations. An analysis by Adekoya et al. (2002) in the United States reported that the leading cause of TBI deaths was violence, especially related to firearms. Falls are also an important cause of TBI in the United States. Recent studies reported by Rutland-Brown et al. (2006) in the United States and Ingebrigtsen et al. (1998) in Norway show falls as the leading cause of TBI. Additional important exposures involve sports and recreational activities. Misclassifications are likely, however; for example, sports-related events may account for up to 10% of TBI deaths but might be reported as falls or as being struck by an object (Whitman et al., 1984). Reports from Alaska (Warren et al., 1995) and Australia (Tate et al., 1998) show sports and recreation activities account for one-fifth to one-fourth of TBI hospital admissions. Military Exposures Although there is ample literature on injury in military populations (e.g., Smith et al., 2000), only three population-based TBI incidence studies could be located. McCarroll and Gunderson (1990) published a report on TBI hospitalization rates in the US Army. The database used was the US Army Patient Administration Systems and Biostatistics Activity for fiscal years 1983â1987. The number of active-duty personnel was obtained from the Defense Manpower Data Center. ICD-9 codes 800, 801, 850, 851, 852, 853, and 854 were used to identify hospital- admitted TBI patients. Incidences per 100,000 persons were derived by age group, sex, and race. About 2,500 patients were admitted each year over the 5-year study period. Rates of concussion and intracranial injury were somewhat higher in males than in females, but the reverse was
EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY 67 observed in some years. The investigators found that 10% of the TBI patients had alcohol or drug involvement, and 97% of the alcohol related-TBIs were in males. Ommaya and associates (1996) evaluated TBI incidence in the US military medical system. Records of discharges from military facilities and private facilities reimbursed by the military for fiscal year 1992 were reviewed for TBI admissions. Medical records of persons with a head-injury diagnosisâICD-9-CM codes 800â801, 803â804, and 850â854âwere identified. ICDMAP (MacKenzie et al., 1989b) was used to convert ICD codes to AIS values. The investigators reported an incidence of 21 (female beneficiaries) to 231 (male active-duty) per 100,000 of population by age group. TBI admission rates were higher in active-duty males 15â17 years old and 18â24 years old. The most common diagnosis was intracranial injury in military hospital admissions and frequently involved firearms. Falls and motor-vehicle crashes accounted for over 62% of the admissions, and fighting 10%. Case-fatality rates (CFRs) ranged from 0% for parachuting to 41% for firearms-related injury. Details on injury severity were not highlighted. Ivins and associates (2006) studied rates of TBI hospitalization of active-duty US Army personnel in 1990â1999. The data source was the Standard Inpatient Data Record database. TBI was identified on the basis of at least one IDC-9-CM code of 800â801, 803â804, and 850â854 in the medical record. ICDMAP-90 was used to assign AIS severity scores for each diagnosis. When there was a lack of information, such as LOC, severity of the TBI was assigned by using criteria developed by the American Congress of Rehabilitation Medicine. Rate ratios were used to compare the incidence of TBI hospitalizations in the Army with the incidence in US civilians 17â49 years old. The overall TBI hospital admission rate in fiscal year 1990 was 248/100,000 active-duty personnel. The rate in 1999 was 62/100,000, 75% lower. TBI incidence declined in each of the three severity classes, but the largest decline in admission rates was in those who had a diagnosis of mild TBI. Overall admission rates declined equally in males and females and in all age groups. There was little change in rates of TBI hospitalization of military active-duty personnel treated in civilian hospitals during the same period. The researchers concluded that the basis of the dramatic decline in rates was effective injury-prevention measures, such as stricter drug- and alcohol-abuse policies, and changes in the Army population; and that changes in hospital admission policies most likely contributed to the decrease in rates of hospitalization for mild TBI. RECURRENT TRAUMATIC BRAIN INJURY Researchers at the Mayo Clinic (Annegers et al., 1980) were among the first to measure the relative risk (RR) of TBI in those with a previous brain injury. They estimated that the risk of a second TBI in those with an earlier TBI was about 3 times the risk of TBI in the general population without such a history. The RR of recurrent TBI given any initial head injury increased with age, and the RR of a third TBI given a second head injury was 8â9 times that of an initial head injury. Jagger et al. (1984b) observed that 31% of their TBI patients reported a previous hospitalization for a head injury. Nestvold and associates (1988) reported that 17% of 465 patients admitted to three hospitals in Akershus County, Norway, had reported an earlier head injury, and about one-fourth of those reported more than one previous TBI. Salcido and Costich (1992) called attention to some possible effects of repeat TBI, including psychosocial aspects and the course of a second rehabilitation. Ruff and co-workers (1990), Kreutzer and co-
68 GULF WAR AND HEALTH workers (1990), and Corrigan et al. (1995) reviewed the literature on TBI and recurrent injury and showed a strong association with alcohol abuse. Closely related to repeat TBI is what has been called the âsecond-impact syndrome,â in which a repeat mild TBI was catastrophic or even fatal (Kelly et al., 1991). Gronwall and Wrightson (1975) concluded that the effects of concussion might be cumulative especially in sports, in which populations may be easily monitored. Recurrent head injury in sports has been the subject of several case reports and case-series studies (e.g., Kelly et al., 1991; Cantu and Voy, 1995). Their findings of risks posed by recurrent TBI have prompted recommendations on when players can return to games in the event of even a minor concussion (CDC, 1997). TRAUMATIC BRAIN INJURY AND SHORT-TERM OUTCOMES One outcome of TBI is death. Whereas mortality is an ideal measure of the magnitude of severity of TBI in the general population, the CFR after hospital admission is a measure of the immediate gross consequences of brain injury. The CFR has been used for decades as an indicator of hospital quality of care, but its use is subject to biases as described below. Case-Fatality Rates CFR data are available from 15 US population-based incidence studies (Table 3.13). They range from 4.4/100 hospitalized patients in Maryland (MacKenzie et al., 1989a) to about 25/100 in the Bronx, New York (Cooper et al., 1983), and 23/100 in Oklahoma (Oklahoma State Department of Health, 1991). The range in rates may reflect gross differences in hospital patient- admission practices. That is, hospitals that admit a high proportion of patients with severe brain injury would be expected to have higher CFRs than hospitals that admit a large proportion of patients with mild brain injury, who are less likely to die. CFRs in the most recent reports in the United States show the effect of changes in hospital admission practices of the last decade: fewer of the mildly head-injured persons were admitted. Table 3.14 summarizes CFR data from outside the United States. The rates in the 15 studies range from 0.8/100 hospitalized patients in a report from South Australia (Badcock, 1988) to 30/100 in a county in Denmark (Engberg and Teasdale, 2001); the latter CFR represented only hospital-admitted patients with ICD-9-CM codes 850â854. The very low rate in South Australia may reflect the fact that over 90% of the patient cohort admitted to the hospitals in the study region had mild TBI. The CFR in severe-TBI patients in the study was 55%, which is comparable with rates in other studies that focused on severe-TBI patients. Discounting the single high CFR from Denmark, all remaining rates are less than 10/100 admitted patients. Occasionally, a total or general CFR appears in the literature (e.g., Servadei et al., 2002a). Such a rate would reflect both in-hospital and prehospital deaths and express the risk of death from the moment of injury to hospital discharge. It is often 2 or 3 times the in-hospital CFR. Examples are found in Kraus et al. (1984), Vazquez-Barquero et al. (1992), and Tiret et al. (1990).
EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY 69 Disposition at the End of Acute Care As previously noted, one of the scales used to assess early outcome after hospitalization for TBI is the GOS. The GOS is a crude indicator of medical (neurologic) complications or of residual effects at the time of discharge from a primary treatment center. The major difficulty with the GOS is its inability to classify patients properly because of the lack of specific criteria that separate severe from moderate and moderate from the good-recovery categories. Good recovery does not mean, nor was it ever intended to mean, complete recovery, and, as noted above, Jennett and Teasdale (1981) devised an extended version of the GOS (GOS-E) to account for the insensitivity of the scale to some changes in functional ability, especially in the moderate and severe categories. The large number of population-based TBI incidence studies might suggest the availability of much more information on the GOS as an early hospital-discharge tool, but only seven of the 66 studies (US and non-US) reported on the scale. Rimel (1981) observed that 69% of TBI patients had a âgood recoveryâ at the time of discharge. The highest percentage of persistent vegetative state was also reported in that study. Almost all other studies in Table 3.15 had a rate of good recovery of 75% or higher. The one exception is the study by Masson et al. (2003), in which only 18% of patients were discharged with a good recovery; their study population, however, consisted of only patients admitted to the hospital with severe TBI. SUMMARY Almost all the incidence studies had shortcomings, and that should be considered in drawing conclusions. No two published studies are identical in methods. However, many studies have used reasonable methods to identify patient cases, defined and measured the populations that gave rise to the patients, used acceptable methods in identifying patients in treatment facilities or in administrative datasets, defined TBI (and severity levels) in reasonable ways, classified exposures that gave rise to the injuries in ways that make sense, recorded basic descriptive information about patients in uniform formats, and, in longitudinal studies, followed patients for outcomes by using acceptable methods to reduce losses and used accepted outcome instruments. Thus, we can learn a great deal about the epidemiology of TBI and use that knowledge to help in designing prevention strategies.
70 TABLE 3.1 Glasgow Coma Scales and Glasgow Outcome Scales Glasgow Coma Scalea Glasgow Coma Scale-Extendedb Glasgow Outcome Scale Glasgow Outcome Scale-Extendedc Ability Assessed Points Memory Assessed Points Condition Points Condition Points Eye opening Amnesia Dead 1 Dead 1 Not open 1 >3 mo 0 Vegetative state 2 Vegetative state 2 To pain 2 31â90 days 1 Severely disabled 3 Lower severe disabled 3 To speech 3 8â30 days 2 Moderately disabled 4 Upper severe disabled 4 Spontaneous 4 1â7 days 3 Good recovery 5 Lower moderate disabled 5 Verbal response 3â24 h 4 Upper moderate disabled 6 Silence 1 0.5â3 h 5 Lower good recovery 7 Sounds 2 <30 min 6 Upper good recovery 8 Nonsense 3 No amnesia 7 Confused 4 Motor response to pain No response 1 Arm extension 2 Arm flexion 3 Withdrawal 4 Localizing 5 To command 6 a SOURCE: Jennett and Teasdale, 1981. b SOURCE: Nell et al., 2000. c SOURCE: Jennett et al., 1981.
TABLE 3.2 US TBI Incidence Studies: Case Identification, Data Source, and TBI Severity Scoring Year(s) Reference of Data Location Case Definition and Data Source TBI Severity Criteria and Scoring Annegers et al., 1935 to Olmstead Record linkage with head injury, with concussion, with LOC, PTA, Fatal: (< 28 days); Severe: intracranial 1980 1974 County, MN neurological signs of brain injury or skull fracture concussion, with hematoma, contusion or LOC > 24 LOC, PTA, neurologicsigns of TBI hours, or PTA > 24 hours; Moderate: LOC or PTA 30 minutes to 24 hours, skull fracture, or both; Mild: LOC or PTA < 30 minutes without skull fracture Klauber et al., 1978 San Diego ICDA-8 Codes 800, 801, 804 806, and 850â854 with hospital GCS of 3, 4â5, 6â7, 8â15 1981 County, CA admission diagnosis or cause of death with skull fracture, LOC, PTA neurological Rimel, 1981 1977 to Central Virginia CNS referral patients with significant head injury admitted to GCS (3â5, 6â8, 9â11, 12â15); severe 1979 neurosurgical service unit. Prehospital deaths from medical examiner = < 8; moderate = 9â11; mild = 12â15 Jagger et al., 1978 North Central Patients within defined service area with overnight stay, and Not reported 1984b Virginia documented head injuries Kraus et al., 1984 1981 San Diego Physician-diagnosed physical damage from acute mechanical energy Modified GCS: severe 8; moderate County, CA exchange resulting in concussion, hemorrhage, contusion, or = 9â11; plus hospital stay of 4â8 laceration of brain hours and brain surgery, or abnormal CT, or GCS 9â12; mild = all others, GCS 13â15 Whitman et al., 1979 to Inner city Any hospital discharge diagnosis of ICD-9-CM 800â804, 830, 850â (1) Fatal; (2) Severe = intracranial 1984 1980 Chicago and 854, 873, 920, 959. Injury within 7 days prior to hospital visit and hematoma, LOC /PTA > 24 hours Evanston, IL blow to head/face with LOC, or laceration of scalp or forehead contusion; (3) Moderate + LOC or PTA 30 minutes to < 24 hours; (4) Mild + LOC or PTA < 30 minutes; (5) Trivial + remainder Fife et al., 1986 1979 to Rhode Island All admissions to Rhode Island hospitals Professional Activities Severity not evaluated 1980 Study (PAS) using ICD-9 codes 800â801.9, 803â804.9, 850â854.9 Fife, 1987 1977 to US US National Health Interview Survey translated rates ICD codes 800â Severity not evaluated 1981 801.9, 803â803.9, 850â854.9 71
72 Year(s) Reference of Data Location Case Definition and Data Source TBI Severity Criteria and Scoring MacKenzie et al., 1986 Maryland ICD-9-CM codes 800, 801, 803, 804, 850â854 ICDMAPâconverts ICD codes to 1989a Abbreviated Injury Severity Scores MacKenzie et al., 1979 to Maryland ICD-9-CM codes 800, 801, 803, 804, 850â854 ICDMAPâconverts ICD codes to 1990 1986 Abbreviated Injury Severity Scores Fuortes et al., 1984 to Iowa State central head injury registry of hospital discharge abstracts Not reported 1990 1986 Oklahoma State 1979 to Oklahoma Hospital discharge codes ICD-9-CM 800â800.9, 801â801.9, 803â AIS 1 = minor AIS 2 = moderate Department of 1986 803.9, 804â804.9, 850â850.9, 851â851.9, 852â852.9, 853â853.9, AIS 3 = 3â5 = severe Health, 1991 854â854.9, 905, 907. Excluded ED visits, ME probable cause of death for TBI Cooper et al., 1980 to Bronx, NY Hospital/ED logs and ICD-9-CM codes 800â801, 803â804, 850â854 Not reported 1983 1981 Schuster, 1994 1989 to Massachusetts State vital statistics mortality file ICD-9; codes 800â802, 803â804, Not reported 1991 850â854, 873 State uniform hospital discharge data set ICD-9 CM codes 800â 801,803â804, 850â851 Warren et al., 1991 to Alaska State Trauma Registry ICD-9-CM codes 800â804, 850â854, 950â954 Not reported 1995 1993 Thurman et al., 1990 to Utah Discharge date from all Utah acute care hospitals and state vital (1) Initial GCS: Severe = < 8; 1996 1992 records using ICD-9-CM codes 800â801.9, 803â804.9, and 850â Moderate = 9â12; mild = 13â15; (2) 854.1 in any primary or secondary data fields Demonstrated intracranial traumatic lesions; (3) Focal abnormalities on neurological examination Diamond, 1996 1988 to Virginia All ED treated patients from Virginia Brain Injury Central Registry Severity not evaluated 1993 including hospital admitted ICD-9-CM codes 850â854.1, 800â804.9, 348.1, 900â900.9, 950â951.9 Gabella et al., 1990 to Colorado, Hospital discharge data for all state hospitals or healthcare providers No severity data reported 1997a 1993 Missouri, Oklahoma, Utah
Year(s) Reference of Data Location Case Definition and Data Source TBI Severity Criteria and Scoring Gabella et al., 1991 to Colorado Colorado surveillance system of hospitalized and fatal TBI using ICDMAP using as many as 5 ICD 1997b 1992 ICD-9-CM codes 800, 801, 803, 804, 850â854 discharge diagnoses. Severe TBI = died or ISS > 9 Sosin et al., 1996 1991 US Self-reported data from US National Health Interview Survey Injury Severity not evaluated Supplement. Mild and moderate brain injury defined as loss of consciousness in previous 2 months Thurman and 1980 to US All hospital discharge records with one or more ICD-9-CM code(s) of ICDMAP used to convert ICD codes Guerrero, 1999 1995 800â801.9, 803â804.9 or 850â854.1 from the National Hospital to approximate Abbreviated Injury Discharge Survey Scale Scores. 1â2 = mild; 3 = moderate; 4â6 = severe Jager et al., 2000 1992 to US Same ICD codes as Thurman et al., 1996; identified from US National Severity not evaluated 1994 Hospital Ambulatory Medical Care Survey Guerrero et al., 1995 to US All visits to emergency departments with same ICD codes as Severity not evaluated 2000 1996 Thurman et al. 1996; identified from US National Hospital Ambulatory Medical Care Survey Schootman et al., 1993 Iowa Hospital discharge data ICD-9 codes 800-801, 803-804, 850-854 No severity data reported 2000 [capture - recapture method] plus death certificates Langlois et al., 1997 14 US states State TBI surveillance projects. Deaths excluded, cases identified as GCS < 8 = severe; 9â12 = moderate 2003 ICD-9-CM 800â801.9, 803â804.9, 850â854.1, 959.1 plus evidence of > 12 no brain lesions LOC, PTA, skull fracture, etc. > 12 with brain lesion > 12 no cat. done Langlois et al., 1995 to US ED visits from National Ambulatory Care Survey ICD-9-CM codes Not evaluated 2006 2001 800â801, 803â804, 850â854, 959 Hospitalizations: National Hospital Discharge Survey, same as ICD codes as above Deaths multiple cause of death taken from US National Vital Statistics System [some double counting was probable] Selassie et al., 2004 1996 to South Carolina Statewide surveillance of TBI related hospitalizations. Used ICD-9- Mild = AIS 1â2, Moderate = AIS 3, 2001 CM codes as in Langlois et al., 2003 Severe = AIS 4â5 73
74 Year(s) Reference of Data Location Case Definition and Data Source TBI Severity Criteria and Scoring Texas Department 1998 Texas Texas Trauma Registry and Bureau of Vital Statistics ICD-9 codes GCS used but not reported of Health, 2004 800â801, 803â804, 850â854 Rutland-Brown et 1995 to US Update from Langlois et al., 2003, see this for case ID See Langlois et al., 2003 al., 2006 2001 NOTE: AIS = Abbreviated Injury Scale, CA = California, CNS = central nervous system, CT = computed tomography, ED = emergency department, GCS = Glasgow Coma Scale, ICD = International Classification of Diseases, ICDA-8 = International Classification of Diseases, Eighth Revision, ICD-9-CM = International Classification of Diseases, Ninth Revision, Clinical Modification, ICDMAP = computer algorithm; ID = identification, ISS = Injury Severity Score, LOC = loss of consciousness, MN = Minnesota, PAS = Professional Activities Study, PTA = posttraumatic amnesia, TBI = traumatic brain injury, US = United States.
TABLE 3.3 Non-US Incidence Studies: Case Identification, Data Source, and TBI Severity Score Year(s) Reference of Data Location Case Definition and Data Source TBI Severity Criteria and Scoring Jennett and 1974 England, Wales Death records, hospital admission records with ICD 800, Not reported MacMillan, and Scotland 801, 803, 804, 850â854 1981 Selecki et al., 1977 New South Wales Hospital inpatient statistics of Health Commission ICD-8 Not reported 1981 and South for principal diagnosis Australia Servadei et al., 1981 to San Marino Medical record review with those with skull fracture or Evaluated by GCS but not reported 1985 1982 Republic LOC hospital admitted Wang et al., 1983 Urban areas of Survey of 6 cities with door-to-door interviews and Survey included only a survival population. 1986 China medical record followup Severity not evaluated Nestvold et al., 1974 Central Norway, Prospective identification by surgeons on duty case Survey ranked by length of PTA: None = 1, < 0.5 1988 Akershus County inclusion with neurological symptoms hr = 2, 0.5â6 hr = 3, 6â24 hr = 4, 1â2 days = 5, 3â7 (Oslo) days = 6, > 7 days = 7 Servadei et al., 1981 to Ravenna, Italy ED identification plus hospital admission and record GCS; 3â5, 6â8, 9-12, 13â15 1988 1982 review Badcock, 1984 South Australia Prospective study of all ED visits, hospital admissions and Length of PTA: none, < 5 min, 5â60 min, 1â24 1988 prehospital deaths hrs, 1â7 days, 1â4 wks, > 4 wks Tiret et al., 1986 Aquitaine, Prehospital deaths and hospital admissions survey by Severity by 3 classes based on PTA of coma > 6 1990 France medical staff using 180 possible head injury codes using hrs = severe, PTA 15 min to 6 hrs = moderate, AIS and ISS PTA, 15 min = mild Levi et al., 1984 to Northern Israel Prospective patient identification from referral to GCS used but not recorded 1990 1988 neurological service records Nell and Brown, 1986 Johannesburg, Inpatient admission with screening ICD-9 codes 800â804, GCS, mild = 13â15, moderate = 7â12 and severe = 1991 South Africa 850â854, 293, 294, 310, 870â873, 950â951, 958, 345, 3â6 347, 348, 253.9 75
76 Year(s) Reference of Data Location Case Definition and Data Source TBI Severity Criteria and Scoring Johansson et al., 1984 to Northern Sweden Hospital admissions with ICD 850â854 Severity not evaluated 1991 1985 Annoni et al., 1987 Canton St. Hospitalized patients with intracranial lesions on Severe brain injury only GCS < 7, 7â9, 10â12, > 1992 Gallen, admission CT 12 Switzerland Vazquez- 1988 Cantabria, Spain Hospital admissions with objective neurological findings GCS, minor = 13â15, moderate 9â12, severe 3â8 Barquero et al., such as LOC, skull fracture 1992 Engberg, 1995 1988 Frederiksborg ED and hospital ICUs in 4 hospitals using hospital Severity by PTA: 24 hrs-7 days = severe, very County, Denmark records, Danish Hospital Register and National Register severe 7 days Chiu et al., 1988 to Taiwan Hospital admission with LOC, skull fracture, neurological GCS: mild = 13â5, moderate = 9â12 (or CT pos), 1997 1994 deficit or CT intracranial hemorrhage severe 8 Hillier et al., 1987 South Australia All public and private hospitals with admission ICD-9 GCS: mild = 13â5, moderate = 9â12 (or CT pos), 1997 codes of 348, 800, 803, 804, 850-854 severe = 3â 8; PTA < 30 min = mild, 30â60 min = moderate, > 60 min = severe, PTA < 60 min = mild, 60 min = moderate, 24 hrs = severe Ingebrigtsen et 1993 Northern Norway All patient referral medical records includes ED visits GCS: minimal = 15 no LOC, mild = 14 or 15 plus al., excludes scalp, facial injuries PTA or brief LOC or impaired alertness, moderate 1998 = 9â13 or LOC > 5 min or focal neurological deficit, severe = 5â8, critical = 3â4 Tate et al., 1988 New South Admission to region hospital with ICD-9 codes 310, 800, Severe = PTA > 24 hrs, or GCS of < 9, moderate 1998 Wales, Australia 801, 803, 804, 850â854, 905.0, 907 = PTA 1â24 hrs or GCS 9â12, mild = PTA or LOC < 1 hr Alaranta et al., 1991 to Finland Hospital discharge or register using ICD-9 codes: 800, Severity not evaluated 2000 1995 801, 803, 850â854 (first-time patients only) Pickett et al., 1988 Greater Kingston Computerized ED injury records from the CHIRPP system Severity not reported 2001 Area of Canada Engberg and 1979 to Denmark Danish National Hospital Register using 8th ICD codes Severity not evaluated Teasdale, 1996 800, 801, 803, 850â854, mortality data from National 2001 Death Register using ICD 8th and 10th codes
Year(s) Reference of Data Location Case Definition and Data Source TBI Severity Criteria and Scoring Masson et al., 1996 Aquitaine, France Persons hospital admitted through emergency service with AIS score of 4 or 5 or LOC 6â24 hrs GCS < 9 2001 of any one of 19 hospitals, data from treating hospital Firsching and 1996 Germany Head injury hospital admitted patients including Severity scoring not reported Woischneck, concussion; deaths from Federal Bureau of Statistics 2001 Gururaj, 1999 Bangalore, India Case definitions from the Neurotrauma Registry of GCS used by categories of severity not defined 2002 National Institute of Mental Health and Neuroscience, Bangalore India including LOC or PTA neurological changes, skull fracture, death due to TBI Servadei et al., 1998 Romagna and Hospital admissions with ICD-9 codes 800â800.3, 801â Severity not evaluated 2002b Trentino, Italy 801.3, 803â803.3, 850; 851â851.1, 852â852.1 853â853.1, 854â854.1 Servadei et al., 1998 Romagna, Italy All patients admitted to hospital care with a discharge Mild TBI as defined by Duckin using ICD codes 2002a diagnosis of ICD-9 800â803.0, 801â801.3, 803â804.3, GCS of 14â15 = mild, 9â13 = moderate, < 9 = 850â854. In hospital and prehospital deaths identified severe from hospital records or death certificates Masson et al., 1996 Aquitaine, France Persons admitted to anyone of 19 public hospitals with Severe TBI by GCS of < 9 for at least 24 hrs 2003 prolonged coma determined by LOC > 24 hrs or GCS of < 9 before sedation Kleiven et al., 1987 to Sweden National hospital discharge register using ICD codes 800â Severity not evaluated 2003 2000 804, 850â854, (ICD-9) and S2.0âS2.9, S6.0âS6.9 (ICD- 10) Andersson et al., 1992 to Western Sweden Persons identified from hospitals ED unit, discharge Mix of symptoms defined by American Congress 2003 1993 register, regional neurological clinic and coronerâs records of Rehabilitation Medicine ICD-9, 850â854, 800â804 Baldo et al., 1966 to Northeast Italy Hospital discharges with ICD-9 codes 800, 801.9, 803, ICDMAP-90 used to convert ICD codes to AIS: 2003 2000 804.9, 850â854.1 located on data base for region 1/2 = mild, 3 = moderate, 4/5 = severe Santos et al., 1994, Portugal From National Institute of Statistics using ICD-9 codes Severity not evaluated 2003 1996, 800, 801, 803, 804, 850â854, 907 for hospital discharge 1997 and mortality data 77
78 Year(s) Reference of Data Location Case Definition and Data Source TBI Severity Criteria and Scoring Steudel et al., 1972 to Germany Federal Bureau of Statistics using ICD-9 codes 800-804 Focus of study is on fatal head injury 2005 1998 and 850â854 and ICD-10 S02âS02.9 and S06âS06.9 Tennant, 2005 2001 to England Hospital Episodes Statistics using ICD-10 codes S00â Severity not evaluated 2003 S09.9 for hospital inpatient care plus Primary Care Trusts Chiu et al., 2007 1991, Taipei City and Prospective TBI registry data. Excludes prehospital deaths GCS: severe 9, moderate = 9â15 plus hospital 2001 Hualien County, in 2001 stay at least 48 hrs and had brain surgery or Taiwan abnormal CT scan, mild = all others Yates et al., 1997 to Royal Devon and ED database from one hospital. ICD codes used but not Based on ICD-10 but not defined 2008 2003 Exeter Hospital, stated UK Wu et al., 2008 2004 6 Providences of Hospital admitted patients with data from attending GCS: severe 9, moderate = 9â13, mild = 14, 15 Eastern China physician NOTE: AIS = Abbreviated Injury Scale, CHIRPP = Canadian Hospitals Injury Reporting and Prevention Program, CT = computed tomography, ED = emergency department, GCS = Glasgow Coma Scale, ICD = International Classification of Diseases, ICDMAP = computer algorithm, ICU = intensive care unit, ISS = Injury Severity Score, LOC = loss of consciousness, PTA = posttraumatic amnesia, TBI = traumatic brain injury, UK = United Kingdom.
TABLE 3.4 US TBI Incidence Studies Base Year(s) of Number of Population Rate / 105 Reference Data Location Patients (x1000) per year Comments Annegers et al., 1980 1965 to 1974 Olmstead County, MN 3,587 NS 193 Age adjusted to 1970 US population, rate averaged from men only and women only rates Fuortes et al., 1990 1984 to 1986 Iowa NS NS 159 in 1984 Hospital admissions only 133 in 1985 117 in 1986 Rimel, 1981 1977 to 1979 Central Virginia 1,330 NS NS Hospital patients and prehospital deaths Klauber et al., 1981 1978 San Diego, CA 5,055 NS 294 Includes some nonresidents, excludes a few external causes Cooper et al., 1983 1980 to 1981 Bronx, NY 1,209 NS 249 Rate based on sample, age adjusted to 1980 US population Jagger et al., 1984b 1978 North Central Virginia 735 354 208 Rate includes residents and nonresidents; no ED cases or prehospital deaths Kraus et al., 1984 1981 San Diego County, CA 3,358 1862 180 Population based, not age adjusted Whitman et al., 1984 1979 to 1980 Inner city Chicago and 782 213 331 Composite rate from data in publication, Evanston, IL average across race and gender Fife et al., 1986 1979 to 1980 Rhode Island 2,870 947 152 Hospital patients only Fife, 1987 1977 to 1981 US 307,000 226,545 136 Hospital patients only; 1.87 million 1.87 million 805 All injured patients MacKenzie et al., 1989a 1986 Maryland 5,838 NS 132 Hospital patients only MacKenzie et al., 1990 1979 to 1986 Maryland NS NS 114-134 Hospital patients only, range in rates Oklahoma State 1989 Oklahoma 3,672 NS 121 Hospital and fatal cases Department of Health, 1991 Schuster, 1994 1990 Massachusetts 27,819 6,016 10 Mortality rate 86 Hospital admissions 366 ED only Warren et al., 1995 1991 to 1993 Alaska 2,178 457 130 Hospital patients only Diamond, 1996 1988 to 1993 Virginia 46,680 NS NS Only age-specific rates reported Sosin et al., 1996 1991 US 1.54 million NS 618 Total rate 158 Hospitalized 79
80 Base Year(s) of Number of Population Rate / 105 Reference Data Location Patients (x1000) per year Comments 307 ED only 153 No care Thurman et al., 1996 1990 to 1992 Utah 5,782 NS 106 Age adjusted rate to 1990 US population Gabella et al., 1997a 1990 to 1993 Colorado, Missouri, 13,978 13,687 103 Age adjusted rate to 1990 US population Oklahoma, Utah Gabella et al., 1997b 1991 to 1992 Colorado 7056 NS 101 Hospitalized and deaths, age adjusted to US Thurman and Guerrero, 1994 to 1995 US NS NS 98 Hospitalized patients only 1999 Jager et al., 2000 1992 to 1994 US 1.144 million NS 444 ED patients only Schootman et al., 2000 1993 Iowa 2,559 NS 91 Severe TBI rate based capture- recapture; age adjusted rate to 1990 US population Guerrero et al., 2000 1995 to 1996 US 1.027 million NS 392 ED patients only Louisiana Office of 1996 to 1999 Louisiana 16,203 NS 90 Hospitalized patients and prehospital Public Health Injury and deaths Research Prevention Section, 2004 Langlois et al., 2003 1997 14 US states 62,771 NS 70 Live hospital discharges only Langlois et al., 2006 1995 to 2001 US 1.396 million NS 505 Total rate, age adjusted to 2000 US population 235 86 Hospitalized patients only 1.111 million 401 ED visits only Selassie et al., 2004 1996 to 2001 South Carolina 70,671 NS 68 Hospital patients only 220 ED patients only Texas Department of 1998 Texas 20,000 NS NS Hospitalized patients only Health, 2004 Rutland-Brown et al., 2003 US 1.565 million NS 538 Total 2006 421 ED visits only 100 Hospitalization NOTE: CA = California, ED = emergency department, IL = Illinois, MN = Minnesota, NS = not stated, NY = New York, TBI = traumatic brain injury, US = United States.
TABLE 3.5 Non-US TBI Incidence Data Base Year(s) of Number of Population Rate / 105 Reference Data Location Patients (x1000) per year Comments Jennett and MacMillan, 1974 England, Wales and NS NS 270 in Annual rates based on sample weeks, 1981 Scotland England and rates not age adjusted Wales 313 in Scotland Selecki et al., 1981 1977 New South Wales, 18,678 4,960 377 Hospital admissions only excludes Australia prehospitalized deaths Servadei et al., 1985 1981 to 1982 San Marino 327 23.5 468 Hospital admissions Wang et al., 1986 1982 Urban areas of China 35 63 56 Rates based on samples of households in city communities Nestvold et al., 1988 1974 Central Norway 488 350 236 Hospital admissions Badcock, 1988 1984 South Australia 1,698 NS 520 Includes ED visits, admissions and prehospital deaths Servadei et al., 1988 1984 Ravenna, Italy 644 172 372 Hospitalized cases only excludes ED treated and released Levi et al., 1990 1984 to 1988 Northern Israel 1,370 1,200 25 Rate in person-years Tiret et al., 1990 1986 Aquitaine, France 8,940 2,700 281 Hospital admissions and deaths Johansson et al., 1991 1984 to 1985 Northern Sweden 242 70 242 Ages 16-60 only, hospital admissions Nell and Brown, 1991 1986 Johannesburg, South 5,106 NS 316 Rate based on population estimates Africa Annoni et al., 1992 1987 St. Gallon Canton of 80 410 20 Rate based on sample of hospital Switzerland weeks of data collection Vazquez-Barquero et al., 1988 Cantabria, Spain 477 523 91 Hospital admissions only 1992 Engberg, 1995 1988 Frederiksborg County, NS 340 22.6 (7.1) ICD 851-4 only (PTA > 7 days) Denmark Chiu et al., 1997 1988 to 1994 Taiwan 58,563 NS Taipei: 220 Number of patients excludes Hualien Co: prehospital deaths; rates include 30 nonhospital deaths Hillier et al., 1997 1987 South Australia 4,486 1,393 322 Rate for persons 16+, excludes prehospital deaths 81
82 Base Year(s) of Number of Population Rate / 105 Reference Data Location Patients (x1000) per year Comments Ingebrigtsen et al., 1998 1993 Northern Norway 247 108 229 Hospital referred patients Tate et al., 1998 1988 New South Wales 1,259 NS 100 Hospital admitted patients excludes prehospital deaths Alaranta et al., 2000 1991 to 1995 Finland 24,497 5,100 95â100 Hospital discharge first-time TBI patients only, rate range over 5 years Engberg and Teasdale, 1979 to 1996 Denmark NS NS 265, 224, Hospitalized patients trend from 1979â 2001 157 1981, 1985â1987, 1991â1993, excludes prehospital deaths Firsching and Woischneck, 1996 Germany 279,000 82,000 350 Hospital admitted patients only 2001 Masson et al., 2001 1996 Aquitaine, France 497 2,800 17.3 Total AIS 4 and 5, Severe TBI (AIS4 = 7.2 AIS5 = 10.1) Pickett et al., 2001 1998 Kingston, Canada 760 176 431 Rate calculated from published data (115) (potential head injuries) Gururaj, 2002 1999 Bangalore, India NS NS 160 Hospital admitted patients only Servadei et al., 2002b 1998 Romagna and Trentino, 3,554 1,439 314 Total for residence (Romagna, Italy (2,421, (969, 470) (297, Trentino) hospitalization rates 1,133) 332) Servadei et al., 2002a 1998 Romagna, Italy 2,430 971 250 Hospital admitted patients only Andersson et al., 2003 1992 to 1993 Western Sweden 753 138 546 Include ED, hospital admitted and coroner records Baldo et al., 2003 1996 to 2000 Northeast Italy 55,368 NS 301â212 Hospital admissions 1996â2000 Kleiven et al., 2003 1987 to 2000 Sweden 22,000 8,400â8,900 259 Average over 14 years hospital per yr discharged patients Masson et al., 2003 1996 Aquitaine, France 248 2,800 8.5 Coma patients only Santos et al., 2003 1994, 1996, Portugal 40,633 9,500 151 (1994), All hospital discharges for 1994â1997, 1997 137 (1996, plus TBI deaths 1997) Steudel et al., 2005 1998 Germany 276,584 82,000 337 Hospitalized cases and incidence rate; ICD 9 codes 800â804, 850â854 Tennant, 2005 2001 to 2002 England 112,718 NS 229 Hospitalized incidence rate
Base Year(s) of Number of Population Rate / 105 Reference Data Location Patients (x1000) per year Comments Chiu et al., 2007 2001 Taipei and Hualien 5,754, 2,634 218, 417 Excludes prehospital deaths County, Taiwan 1,474 353 Wu et al., 2008 2004 6 Provinces of Eastern 14,948 NS NS Hospital admitted cases China Yates et al., 2006 1997 to 2003 England NS 345 453 From an ED database including hospital admitted NOTE: AIS = Abbreviated Injury Scale, Co = county, ED = emergency department, ICD = International Classification of Diseases, PTA = posttraumatic amnesia, TBI = traumatic brain injury. 83
84 TABLE 3.6 US TBI Deaths and Mortality Rates Number of Reference Year(s) of Data Location Deaths Rate / 105 per year Comments Annegers et al., 1980 1965 to 1974 Olmsted County, Minnesota 446 Male: 32 Average rates per year Female: 9 Klauber et al., 1981 1978 San Diego County, California 381 22 Excludes gunshot deaths Cooper et al., 1983 1980 to 1981 Bronx, New York NS 28 50% from violence, 75% before hospital admission Kraus et al., 1984 1981 San Diego County, California 562 30 Includes inhospital and prehospital deaths Whitman et al., 1984 1979 to 1980 Inner city Chicago and 54 19 Evanston blacks Average rates per year Evanston, IL 11 Evanston whites 32 Inner city Fife et al., 1986 1979 to 1980 Rhode Island 248 26/year Rate derived from data in text Cowan et al., 1990 1990 Delaware 122 18 59% from motor vehicle crashes Sosin et al., 1996 1979 to 1986 US 39,416 per yr 17 Death certificate review, average over 8 yrs Sosin et al., 1996 1979 to 1992 US 52,000 25 in 1979 Average number of deaths per yr 19 in 1992 Thurman et al., 1996 1990 to 1992 Utah 1,067 20 Average rate per yr Gabella et al., 1997b 1991 to 1992 Colorado 1,312 18 in urban regions 34 in rural regions Gabella et al., 1997a 1990 to 1992 Colorado, Missouri, 3,172 23 Average rate per yr Oklahoma, Utah Thurman et al., 1999 1994 US 51,350 20 Data source is state TBI registry Adekoya et al., 2002 1989 to 1998 US 53,288 per yr 21 22 in 1989 19 in 1998 Langlois et al., 2006 1995 to 2001 US 49,900 per yr 18 Average rate over 7 years Rutland-Brown et al., 2003 US 51,000 18 2006 NOTE: IL = Illinois, TBI = traumatic brain injury, US = United States.
TABLE 3.7 Non-US TBI Deaths and Mortality Rates Year(s) of Number of Rate / 105 per Reference Data Location Deaths year Comments Jennett and MacMillan, 1972 to 1976 England, Wales and Scotland NS 9 Average over 5 years 1981 Simpson et al., 1981 1977 New South Wales and South 1,727 28 Average for both regions Australia Badcock, 1988 1984 South Australia NS 6 Nestvold et al., 1988 1974 Central Norway 23 7a Prehospital plus inhospital deaths Servadei et al., 1988 1984 Ravenna, Italy 42 24 Source of mortality data not given Levi et al., 1990 1984 to 1988 Northern Israel 59b 3 Source of mortality not given Tiret et al., 1990 1986 Aquitaine, France 391 22 Source was mortality statistics for Aquitaine Johansson et al., 1991 1984 to 1985 Northern Sweden 14 12 For ages 16â60 years only Nell and Brown, 1991 1986 Johannesburg, South Africa 1,303 80 Source is coronal file. Rate over 190/100,000 for blacks age 25â44 Vazquez-Barquero et al., 1988 Cantabria, Spain 103 20 Death rates higher for males 1992 Engberg, 1995 1988 Frederiksborg County, Denmark 45 13 Computerized search of death records Chiu et al., 1997 1988 to 1994 Taipei, 2,621 20 Hualien County, Taiwan 82 Engberg and Teasdale, 2001 1979 to 1996 Denmark NS 11 1994 to 1996 average Firsching and Woischneck, 1996 Germany 9,415 12 2001 Masson et al., 2001 1996 Aquitaine, France 149 5 Limited to severe TBI only Gururaj, 2002 1999 Bangalore, India NS 20 Servadei et al., 2002b 1998 Romagna and Trentino, Italy 85 5 In hospital deaths (75, 10) (8, 2) Servadei et al., 2002a 1998 Romagna, Italy 225 18 Highest rates for 5â24 and 75+ Andersson et al., 2003 1992 to 1993 Western Sweden 5 4 Few coma patients Baldo et al., 2003 1996 to 2000 Northeast Italy NS 7 Average rate over 4 years Masson et al., 2003 1996 Aquitaine, France 128 5 Coma patients only, rate derived from text Santos et al., 2003 1994, 1996, Portugal 5,425 17 Highest rate for those age 20â29, 80+ 1997 85
86 Year(s) of Number of Rate / 105 per Reference Data Location Deaths year Comments Steudel et al., 2005 2000 Germany 7567 9 Represents 67% decrease from 1972 a Rate derived from data in narrative. b Includes work and other. NOTE: NS = not stated, TBI = traumatic brain injury.
TABLE 3.8 Percent Severity Distributions of Hospitalized Patients in US and Non-US Incidence Studies Year(s) of Percent Percent Percent Reference Data Location Mild Moderate Severe Comments Annegers et al., 1980 1965 to 1974 Olmsted County, MN 63 29 7 See table 3 for criteria Rimel, 1981 1977 to 1979 Central Virginia 49 26 25 Based on GCS Klauber et al., 1981 1978 San Diego County, 91 25 91% includes mild and moderate. Mild = 12â15 California Whitman et al., 1984 1979 to 1980 Innercity Chicago and 86 9 5 Severity criteria same as Annegers, see table 3 Evanston, IL Kraus et al., 1984 1981 San Diego County, 82 9 9 Using slightly modified GCS California Badcock, 1988 1984 South Australia 90 8 2 Based on PTA Tiret et al., 1990 1986 Aquitaine, France 80 11 9 Based on LOC and AIS Nell and Brown, 1986 Johannesburg 87 8 5 GCS moderate = 7â12, severe = 3â6 1991 Vazquez-Barquero et 1988 Cantabria, Spain 88 7 5 Based on GCS al., 1992 Thurman et al., 1996 1990 to 1992 Utah 39 42 19 Using the GCS Chiu et al., 1997 1988 to 1994 Taiwan 79 9 12 Based on GCS over 7 years Hillier et al., 1997 1987 South Australia 75 9 16 Based on GCS, ICDMAP, 82 3 15 PTA, respectively 82 9 9 Tate et al., 1998 1988 New South Wales 65 21 14 Based on PTA Thurman et al., 1999 1980 to 1995 US 78 16 6 1980â1981 (Used ICDMAP to create AIS 1994â 23 1995) Gururaj, 2002 1999 Bangalore, India 70 14 16 Based on GCS Servadei et al., 2002a 1998 Romagna, Italy 66 7 27 Mild = 14â15, Moderate = 9â13 Baldo et al., 2003 1991 to 2000 Northeast Italy 69 22 9 1996 63 16 21 2000; Used GCS Chiu et al., 2007 1991 vs. 2001 Taipei and Hualien 78 10 12 Taipei 1991 County, Taiwan 77 9 14 Taipei 2001 87 6 7 Hualien County 1991 83 9 8 Hualien County 2001 Based on GCS 87
88 Wu et al., 2008 2004 6 Provinces of Eastern 62 18 20 Used GCS China NOTE: AIS = Abbreviated Injury Scale, GCS = Glasgow Coma Scale, ICDMAP = computer algorithm, IL = Illinois, LOC = loss of consciousness, MN = Minnesota, PTA = posttraumatic amnesia.
EPIDEMIOLOGY OF TRAUMATIC BRAIN INJURY 89 TABLE 3.9 Highest Age-Specific TBI Rates and Gender Rate Ratios: US Studies References Highest Age Ranges Male / Female Rate Ratio Annegers et al., 1980 15â24 2.3 Klauber et al., 1981 15â19, 20â29, 70+ 1.3â1.8 Cooper et al., 1983 20â39 2.8 Jagger et al., 1984b 15â19, 20â24, 25â29 2.4 Whitman et al., 1984 Varied by race/community 2.5 Kraus et al., 1984 15â24, 70+ 2.2 Fife et al., 1986 15â25, < 10, 75+ 2 MacKenzie et al., 1989a 15â24, 75+ 2.1 MacKenzie et al., 1990 15â24, 75+ 2.1 Fuortes et al., 1990 15â19 NS Oklahoma State Department of Health, 1991 15â19, 20â24, 60+ 2 Schuster, 1994 15â24, 0â4 1.5 Warren et al., 1995 85+, 15â24 2.4 Thurman et al., 1996 75, 15â24 2.1 Diamond, 1996 0â5, 11â19, 20â29 1.4 Sosin et al., 1996 15â24 1.6a Gabella et al., 1997b 15â24, 65+ 2 Gabella et al., 1997a 15â24, 75+ 2.1 Thurman and Guerrero, 1999 1980 to 1981 = 15â24, 0â4 2 1994 to 1995 = 65+, 15â24 Jager et al., 2000 0â4, 85+, 15â24 1.7 Guerrero et al., 2000 0â14, 15â24 1.6b Schootman et al., 2000 15â24, 75â84 1.9 Louisiana Office of Public Health Injury and 85+, 75â84, 15â24 2 Research Prevention Section, 2004 Langlois et al., 2003 65+, 15â19 1.9 Langlois et al., 2006 0â4, 15â19, 75+ 1.5 Texas Department of Health, 2004 20â24, 75+ 2.1 Rutland-Brown et al., 2006 0â4, 5â24 1.5 a Mild and moderate only. b ED patients only. NOTE: ED = emergency department, NS = not stated.
90 GULF WAR AND HEALTH TABLE 3.10 Highest Age-Specific TBI Rates and Gender Rate Ratios: Non-US Studies Reference Highest Age Ranges Male / Female Rate Ratio Jennett and MacMillan, 1981 15â19, 0â4 NS Wang et al., 1986 40â49 1.6 Nestvold et al., 1988 10â19, 20â29 1.9 Servadei et al., 1988 NS 1.6 Badcock, 1988 15â19, 0â4 2 Levi et al., 1990 0â4, 75+ 2.7 Tiret et al., 1990 < 5, 15â24, 75+ 2.1 Johansson et al., 1991 15â24 3.2a Nell and Brown, 1991 Africans: 25â44, 15â24 4.6 Whites: 15â24, 65+ 9.3 Annoni et al., 1992 NS 3.1 Vazquez-Barquero et al., 15â25, < 15 2.9 1992 Engberg, 1995 65+ 2.1 Chiu et al., 1997 20â29, 10â19, 30â39 2.2 Hillier et al., 1997 16â25 2.3b Ingebrigtsen et al., 1998 0â9, 10â24, 85+ 1.7 Tate et al., 1998 15â24 2.7c Alaranta et al., 2000 Males: 0â9 1.5 Females: 70+ Engberg and Teasdale, 2001 Varied by ICD diagnosis NS Masson et al., 2001 75+, 15â29 2.5d Pickett et al., 2001 10â19, < 9 1.7 Servadei et al., 2002a 1â4, 15â24, 75+ 1.6 Andersson et al., 2003 0â9, 10â19 1.5 Baldo et al., 2003 NS 1.6 Kleiven et al., 2003 Males: 15â19, 85+ 2.1 Females: 0â4, 85+ Masson et al., 2003 75+ 3.1e Santos et al., 2003 20â29, 80+ 1.8c Steudel et al., 2005 75â90 NS Chiu et al., 2007 20â29 (2001), 70+ (1991) 2.1 Yates et al., 2006 15â19, 80â84, 0â5 NS a Ages 16â60. b Ages 16+. c Cases, not rates. d Severe brain injury. e Patients in coma. NOTE: NS = not stated, TBI = traumatic brain injury.
TABLE 3.11 Percent Distributions of TBI Incidence Cases by External Cause: US Studies Firearms / Reference Transport Falls Violencea Sport / Recreation Otherb Annegers et al., 1980 40 29 6 16 8 Klauber et al., 1981 53 NS NS 6 41 Rimel, 1981 55 18 12 17 NS Cooper et al., 1983 27 32 34 NS NS Jagger et al., 1984b 55 20 11 NS 14 Kraus et al., 1984 48 21 18 10 4 Whitman et al., 1984 32 21 35 5 6 Fife et al., 1986 39 35 9 NS 17 MacKenzie et al., 1989a 49 26 11 NS 14 Oklahoma State Department of Health, 1991 42 25 19 8 6 Cowan et al., 1990c 59 3 37 NS 2 Schuster, 1994 33 30 18 NS 20 Warren et al., 1995 31 20 21 20 7 Thurman et al., 1996 54 21 15 6 4 Sosin et al., 1996d 28 NS 9 20 43 Jager et al., 2000e 29 39 NS NS 32 Gabella et al., 1997a 48 23 18 3 9 Guerrero et al., 2000e 22 31 35 NS 11 Adekoya et al., 2002c 34 10 40 NS 16 Louisiana Office of Public Health Injury and Research Prevention Section, 46 34 12 NS 8 2004 Langlois et al., 2006f 39 28 11 NS 22 Texas Department of Health, 2004 48 20 12 NS 20 Rutland-Brown et al., 2006 19 32 10 NS 39 a Includes self-inflicted. b Includes work and other. c Mortality data. d Mild and moderate. e ED visits only. f Hospital admits only. NOTE: NS = not stated, TBI = traumatic brain injury. 91
92 TABLE 3.12 Percent Distributions of TBI Incidence Cases by External Cause: Non-US Studies Firearms / Reference Transport Falls Violencea Sport / Recreation Otherb Tiret et al., 1990 60 33 <1 NS 7 Nell and Brown, 1991c Females: 40 3 39 NS 16 Males: 36 4 46 15 Vazquez-Barquero et al., 1992 60 24 NS NS 16 Chiu et al., 1997 69 20 7 1 3 Hillier et al., 1997 57 29 10 NS 4 Servadei et al., 1988 69 26 1 NS 5 Ingebrigtsen et al., 1998 21 62 7 NS 10 Tate et al., 1998 40 20 8 25 6 Alaranta et al., 2000 26 61 5 NS 8 Firsching and Woischneck, 56 31 12 NS NS 2001 Masson et al., 2001 48 42 3 NS 7 Servadei et al., 2002a 48 33 1 1 17 Andersson et al., 2003 16 58 NS NS 26 Kleiven et al., 2003 26 54 15 NS 7 Masson et al., 2003 59 16 10 NS 14 Chiu et al., 2007 Taipei: 45 34 11 NS 10 Hualien Co:55 28 13 NS 4 a Includes self-inflicted injury. b Includes work and all other causes. c Both races, nonfatal TBI. NOTE: Co = county, NS = not stated.
TABLE 3.13 TBI In-Hospital Case Fatality Rates (CFR) from US Population-Based Studies Reference Location Source of Data and Study Population N = Sample Size CFR % Rimel, 1981 Central Virginia Hospital medical records N = 1330 7.0 Cooper et al., 1983 Bronx, New York Hospital medical and medical exam records N = 1209 24.9 Kraus et al., 1984 San Diego County, Hospital medical records and medical exam records N = 3358 5.2 California Whitman et al., 1984 Chicago area Hospital medical records and medical exam records N = 782 6.9 Jagger, 1984b North Central Virginia Hospital records N = 735 6.5 Fife et al., 1986 Rhode Island Hospital records N = 2870 4.9 MacKenzie et al., 1989a Maryland Hospital records N = 5838 4.4 Oklahoma State Department of Oklahoma Hospital/medical examiner records N = 3672 23 Health, 1991 Schuster, 1994 Massachusetts Hospital/medical examiner records N = 5778 10.1 Warren et al., 1995 Alaska Trauma registry N = 2178 5.6 Gabella et al., 1997b Colorado TBI surveillance N = 6863 7.6 Gabella et al., 1997a Colorado, Missouri, Hospital discharge data N = 11,611 6.9 Oklahoma, Utah Langlois et al., 2003 US 14 state TBI surveillance system N = 67,309 6.9 Rutland-Brown et al., 2006 US National Hospital Discharge Survey and Multiple Cause of Death tape N = 340,757 14.9 Langlois et al., 2006 US Same as Rutland-Brown, 2003 N = 284,900 17.5 NOTE: CFR = case fatality rate, TBI = traumatic brain injury. 93
94 TABLE 3.14 TBI In-Hospital Case Fatality Rates (CFR) from Non-US Population-Based Studies Reference Location Source of Data and Study Population Group Size CFR % Badcock, 1988 South Australia Medical record 1698 0.8 Nestvold et al., 1988 Norway Medical record 488 3.3a Servadei et al., 1988 Ravenna, Italy Hospital record 578 1.9a Tiret et al., 1990 Aquitaine, France Medical record review 281 4.4 Vazquez-Barquero et al., 1992 Cantabria, Spain Hospital record review 477 1.7 Engberg, 1995 Frederisksborg County, Denmark National hospital discharge registry 95 16a Tate et al., 1998 New South Wales Hospital record review 1259 3.9 Engberg and Teasdale, 2001 Denmark National hospital discharge registry NS 30b Gururaj, 2002 Bangalore, India Neurotrauma registry 2814 9 Servadei et al., 2002b Romagna and Trentino, Italy Hospital discharge records 4442 1.0a Servadei et al., 2002a Romagna, Italy Hospital discharge records 2430 2.8c Baldo et al., 2003 Northeast Italy Hospital discharge records 11,074 3.2d Santos et al., 2003 Portugal National Institute of Statistics 39,042 9.8a Steudel et al., 2005 Germany Federal Bureau of Statistics 276,758 1 Chiu et al., 2007 Taipei and Hualien County, Taiwan TBI registry Taipei: 5754 Taipei: 5.4 Hualein County: 1474 Hualein County: 6.7 a CFR derived from text. b CFR for ICD 851â854. c For residents and nonresidents. d From severity rates reported. NOTE: CFR = case fatality rates, TBI = traumatic brain injury.
EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY 95 TABLE 3.15 Percent Distribution of GOS Outcome Categories at Hospital Discharge Rate for US and Non-US Studies Persistent Vegetative Reference Good Recovery Moderate Disability Severe Disability State Death Rimel, 1981 69 12 8 4 7 Kraus et al., 1984 90 3 1 0.5 6 Chiu et al., 1997 87 4 3 1 5 Masson et al., 2003 18 9 16 3 52 Langlois et al., 2003 74 10 6 0.6 NS Chiu et al., 1997 87 6 4 0.3 3 (Taipei City only) Wu et al., 2008 77 7 2 3 11 NOTE: GOS = Glasgow Outcome Score, NS = not stated.
96 GULF WAR AND HEALTH REFERENCES Adekoya, N., D. J. Thurman, D. D. White, and K. W. Webb. 2002. Surveillance for traumatic brain injury deathsâUnited States, 1989â1998. Morbidity and Mortality Weekly Report Surveillance Summaries 51(10):1â14. Alaranta, H., S. Koskinen, L. Leppanen, and H. Palomaki. 2000. Nationwide epidemiology of hospitalized patients with first-time traumatic brain injury with special reference to prevention. Wiener Medizinische Wochenschrift 150(22):444â448. Andersson, E. H., R. Bjorklund, I. Emanuelson, and D. Stalhammar. 2003. Epidemiology of traumatic brain injury: A population based study in western Sweden. Acta Neurologica Scandinavica 107(4):256â259. Annegers, J. F., J. D. Grabow, L. T. Kurland, and E. R. Laws, Jr. 1980. The incidence, causes, and secular trends of head trauma in Olmsted County, Minnesota, 1935â1974. Neurology 30(9):912â919. Annoni, J. M., S. Beer, and J. Kesselring. 1992. Severe traumatic brain injuryâepidemiology and outcome after 3 years. Disability and Rehabilitation 14(1):23â26. Badcock, K. A. 1988. Head injury in South Australia: Incidence of hospital attendance and disability based on a one-year sample. Community Health Studies 12(4):428â436. Baldo, V., A. Marcolongo, A. Floreani, S. Majori, M. Cristofolettil, A. Dal Zotto, G. Vazzoler, and R. Trivello. 2003. Epidemiological aspect of traumatic brain injury in northeast Italy. European Journal of Epidemiology 18(11):1059â1063. Cantu, R. C., and R. Voy. 1995. Second impact syndrome: A risk in any sport. Physician Sports Medicine 23:27â36. CDC (Centers for Disease Control and Prevention). 1997. Sports-related recurrent brain injuriesâUnited States. Morbidity and Mortality Weekly Report 46(10):224â227. Chiu, W. T., S. J. Huang, S. H. Tsai, J. W. Lin, M. D. Tsai, T. J. Lin, and W. C. Huang. 2007. The impact of time, legislation, and geography on the epidemiology of traumatic brain injury. Journal of Clinical Neuroscience 14(10):930â935. Chiu, W. T., K. H. Yeh, Y. C. Li, Y. H. Gan, H. Y. Chen, and C. C. Hung. 1997. Traumatic brain injury registry in Taiwan. Neurological Research 19(3):261â264. Collins, J. G. 1990. Types of injuries by selected characteristics. Vital and Health Statistics 10(175):1â68. Cooper, K. D., K. Tabaddor, and A. W. Hauser. 1983. The epidemiology of head injury in the Bronx. Neuroepidemiology 2:69â88. Corrigan, J. D., E. Rust, and G. L. Lamb-Hart. 1995. The nature and extent of substance abuse problems in persons with traumatic brain injury. Journal of Head Trauma Rehabilitation 10(3):29â46. Cowan, L. S., M. E. Cannor, and A. L. Hathcock. 1990. Fatal injury surveillance report, Delaware, 1990 head and spinal cord injuries. Delaware Medical Journal 65(7):435â448. Diamond, P. T. 1996. Brain injury in the Commonwealth of Virginia: An analysis of central registry data, 1988â1993. Brain Injury 10(6):413â419.
EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY 97 Drake, A. I., E. C. McDonald, N. E. Magnus, N. Gray, and K. Gottshall. 2006. Utility of Glasgow Coma Scale-Extended in symptom prediction following mild traumatic brain injury. Brain Injury 20(5):469â475. Engberg, A. 1995. Severe traumatic brain injuryâepidemiology, external causes, prevention, and rehabilitation of mental and physical sequelae. Acta Neurologica Scandinavica. Supplementum 164:1â151. Engberg, A., and T. Teasdale. 2001. Traumatic brain injury in Denmark 1979â1996. European Journal of Epidemiology 17:437â442. Fife, D. 1987. Head injury with and without hospital admission: Comparisons of incidence and short-term disability. American Journal of Public Health 77(7):810â812. Fife, D., G. Faich, W. Hollinshead, and W. Boynton. 1986. Incidence and outcome of hospital- treated head injury in Rhode Island. American Journal of Public Health 76(7):773â778. Firsching, R., and D. Woischneck. 2001. Present status of neurosurgical trauma in Germany. World Journal of Surgery 25(9):1221â1223. Fuortes, L., K. Phillips, and J. Muldoon. 1990. Traumatic head and spinal cord injury in Iowa. Iowa Medicine 80(12):560â562. Gabella, B., R. Hoffman, and G. Land. 1997a. Traumatic brain injuryâColorado, Missouri, Oklahoma, and Utah 1990â1993. Morbidity and Mortality Weekly Report 46:8â11. Gabella, B., R. E. Hoffman, W. W. Marine, and L. Stallones. 1997b. Urban and rural traumatic brain injuries in Colorado. Annals of Epidemiology 7(3):207â212. Gronwall, D., and P. Wrightson. 1975. Cumulative effect of concussion. Lancet 2(7943):995â 997. Guerrero, J. L., D. J. Thurman, and J. E. Sniezek. 2000. Emergency department visits associated with traumatic brain injury: United States, 1995â1996. Brain Injury 14(2):181â186. Gururaj, G. 2002. Epidemiology of traumatic brain injuries: Indian scenario. Neurological Research 24(1):24â28. Hillier, S. L., J. E. Hiller, and J. Metzer. 1997. Epidemiology of traumatic brain injury in South Australia. Brain Injury 11(9):649â659. Hoofien, D., E. Vakil, A. Gilboa, and P. Donovick. 2003. Comparison of the predictive power of socioeconomic variables, severity of injury and age on long-term outcome of traumatic brain injury: Sample specific variables versus factors as predictors. Brain Injury 16:9â27. Ingebrigtsen, T., K. Mortensen, and B. Romner. 1998. The epidemiology of hospital-referred head injury in northern Norway. Neuroepidemiology 17(3):139â146. Ivins, B. J., K. A. Schwab, G. Baker, and D. L. Warden. 2006. Hospital admissions associated with traumatic brain injury in the US Army during peacetime: 1990s trends. Neuroepidemiology 27(3):154â163. Jager, T. E., H. B. Weiss, J. H. Coben, and P. E. Pepe. 2000. Traumatic brain injuries evaluated in US emergency departments, 1992â1994. Academic Emergency Medicine 7(2):134â140. Jagger, J., D. Fife, K. Vernberg, and J. A. Jane. 1984a. Effect of alcohol intoxication on the diagnosis and apparent severity of brain injury. Neurosurgery 15(3):303â306. Jagger, J., J. I. Levine, J. A. Jane, and R. W. Rimel. 1984b. Epidemiologic features of head injury in a predominantly rural population. Journal of Trauma-Injury Infection and Critical Care 24(1):40â44.
98 GULF WAR AND HEALTH Jennett, B., and M. Bond. 1975. Assessment of outcome after severe brain damage. Lancet 1(7905):480â484. Jennett, B., and R. MacMillan. 1981. Epidemiology of head injury. British Medical Journal Clinical Research Edition. 282(6258):101â104. Jennett, B., and G. Teasdale. 1981. Management of Head Injuries. Philadelphia: FA Davis. Jennett, B., J. Snoek, M. R. Bond, and N. Brooks. 1981. Disability after severe head injury: Observations on the use of the Glasgow Outcome Scale. Journal of Neurology, Neurosurgery and Psychiatry 44(4):285â293. Johansson, E., M. Ronnkvist, and A. R. Fugl-Meyer. 1991. Traumatic brain injury in northern Sweden. Incidence and prevalence of long-standing impairments and disabilities. Scandinavian Journal of Rehabilitation Medicine 23(4):179â185. Kelly, J. P., J. S. Nichols, C. M. Filley, K. O. Lillehei, D. Rubinstein, and B. K. Kleinschmidt- DeMasters. 1991. Concussion in sports. Guidelines for the prevention of catastrophic outcome. JAMA 266(20):2867â2869. Klauber, M. R., E. Barrett-Connor, L. F. Marshall, and S. A. Bowers. 1981. The epidemiology of head injury: A prospective study of an entire community-San Diego County, California, 1978. American Journal of Epidemiology 113(5):500â509. Kleiven, S., P. M. Peloso, and H. von Holst. 2003. The epidemiology of head injuries in Sweden from 1987 to 2000. Injury Control and Safety Promotion 10(3):173â180. Knightly, J. J., and M. W. Pulliam. 1996. Military head injuries. In Neurotrauma, edited by R. K. Narayan. New York: McGraw-Hill. Kraus J., and L. Chu. 2004. Epidemiology of traumatic brain injury. In: Neuropsychiatry of Traumatic Brain Injury. 2nd ed, edited by J. M. Silver and T. W. McAllister. Washington, DC: American Psychiatric Press. Kraus, J. F., and D. L. McArthur. 1996. Epidemiologic aspects of brain injury. Neurologic Clinics 14(2):435â450. Kraus, J. F., M. A. Black, N. Hessol, P. Ley, W. Rokaw, C. Sullivan, S. Bowers, S. Knowlton, and L. Marshall. 1984. The incidence of acute brain injury and serious impairment in a defined population. American Journal of Epidemiology 119(2):186â201. Kraus, J. F., D. Fife, K. Ramstein, C. Conroy, and P. Cox. 1986. The relationship of family income to the incidence, external causes, and outcomes of serious brain injury, San Diego County, California. American Journal of Public Health 76(11):1345â1347. Kraus, J. F., H. Morgenstern, D. Fife, C. Conroy, and P. Nourjah. 1989. Blood alcohol tests, prevalence of involvement, and outcomes following brain injury. American Journal of Public Health 79(3):294â299. Kreutzer, J. S., K. R. Doherty, J. A. Harris, and N. D. Zasler. 1990. Alcohol use among persons with traumatic brain injury. Journal of Head Trauma Rehabilitation 5(3):9â20. Langlois, J. A., S. R. Kegler, J. A. Butler, K. E. Gotsch, R. L. Johnson, A. A. Reichard, K. W. Webb, V. G. Coronado, A. W. Selassie, and D. J. Thurman. 2003. Traumatic brain injury- related hospital discharges. Results from a 14-state surveillance system, 1997. Morbidity and Mortality Weekly Report Surveillance Summaries 52(4):1â20.
EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY 99 Langlois, J. A., W. Rutland-Brown, and K. E. Thomas. 2006. Traumatic Brain Injury in the United States: Emergency Department Visits, Hospitalizations, and Deaths. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Levi, L., S. Linn, M. Revach, and M. Feinsod. 1990. Head trauma in northern Israel: Incidence and types. Neuroepidemiology 9(5):278â284. Louisiana Office of Public Health Injury and Research Prevention Section. 2004. Traumatic Brain and Spinal Cord Injury in Louisiana: 1996â1999. Baton Rouge, LA: Department of Health and Hospitals, Office of Public Health. MacKenzie, E. J. 1984. Injury severity scales: Overview and directions for future research. American Journal of Emergency Medicine 2(6):537â549. MacKenzie, E. J., S. L. Edelstein, and J. P. Flynn. 1989a. Hospitalized head-injured patients in Maryland: Incidence and severity of injuries. Maryland Medical Journal 38(9):725â732. âââ. 1990. Trends in hospitalized discharge rates for head injury in Maryland, 1979â86. American Journal of Public Health 80(2):217â219. MacKenzie, E. J., D. M. Steinwachs, and B. Shankar. 1989b. Classifying trauma severity based on hospital discharge diagnoses. Validation of an ICD-9-CM to AIS-85 conversion table. Medical Care 27(4):412â422. Masson, F., M. Thicoipe, P. Aye, T. Mokni, P. Senjean, V. Schmitt, P. H. Dessalles, M. Cazaugade, and P. Labadens. 2001. Epidemiology of severe brain injuries: A prospective population-based study. Journal of Trauma-Injury Infection and Critical Care 51(3):481â 489. Masson, F., M. Thicoipe, T. Mokni, P. Aye, P. Erny, and P. Dabadie. 2003. Epidemiology of traumatic comas: A prospective population-based study. Brain Injury 17(4):279â293. McCarroll, J. E., and C. Gunderson. 1990. 5-year study of incidence rates of hospitalized cases of head injuries in the US Army. Neuroepidemiology 9(6):296â305. Modell, J. G., and J. M. Mountz. 1990. Drinking and flyingâthe problem of alcohol use by pilots. New England Journal of Medicine 323(7):455â461. Nell, V., and D. S. Brown. 1991. Epidemiology of traumatic brain injury in JohannesburgâII. Morbidity, mortality and etiology. Social Science and Medicine 33(3):289â296. Nell, V., D. W. Yates, and J. Kruger. 2000. An extended Glasgow Coma Scale (GCS-E) with enhanced sensitivity to mild brain injury. Archives of Physical Medicine and Rehabilitation 81(5):614â617. Nestvold, K., T. Lundar, G. Blikra, and A. Lonnum. 1988. Head injuries during one year in a central hospital in Norway: A prospective study. Epidemiologic features. Neuroepidemiology 7(3):134â144. Oklahoma State Department of Health. 1991. Traumatic Brain Injuries Oklahoma, 1989. Oklahoma City, OK. Ommaya, A. K., A. K. Ommaya, A. L. Dannenberg, and A. M. Salazar. 1996. Causation, incidence, and costs of traumatic brain injury in the US military medical system. Journal of Trauma-Injury Infection and Critical Care 40(2):211â217.
100 GULF WAR AND HEALTH Pettigrew, L. E., J. T. Wilson, and G. M. Teasdale. 2003. Reliability of ratings on the Glasgow Outcome Scales from in-person and telephone structured interviews. Journal of Head Trauma Rehabilitation 18(3):252â258. Pickelsimer, E. E., A. W. Selassie, J. K. Gu, and J. A. Langlois. 2006. A population-based outcomes study of persons hospitalized with traumatic brain injury: Operations of the South Carolina traumatic brain injury follow-up registry. Journal of Head Trauma Rehabilitation 21(6):491â504. Pickett, W., C. Ardern, and R. J. Brison. 2001. A population-based study of potential brain injuries requiring emergency care. Canadian Medical Association Journal 165(3):288â292. Rimel, R. W. 1981. A prospective study of patients with central nervous system trauma. Journal of Neurosurgical Nursing 13(3):132â141. Ruff, R., F. Marshall, and M. R. Klauber. 1990. Alcohol abuse and neurologic outcomes of the severely head injured. Journal of Head Trauma Rehabilitation 5:21â31. Rutland-Brown, W., J. A. Langlois, K. E. Thomas, and Y. L. Xi. 2006. Incidence of traumatic brain injury in the United States, 2003. Journal of Head Trauma Rehabilitation 21(6):544â 548. Salcido, R., and J. F. Costich. 1992. Recurrent traumatic brain injury. Brain Injury 6(3):293â298. Santos, M. E., L. De Sousa, and A. Castro-Caldas. 2003. [Epidemiology of craniocerebral trauma in Portugal]. Acta MÃ©dica Portuguesa 16(2):71â76. Schootman, M., M. Harlan, and L. Fuortes. 2000. Use of the capture-recapture method to estimate severe traumatic brain injury rates. Journal of Trauma-Injury Infection and Critical Care 48(1):70â75. Schuster, M. 1994. Traumatic Brain Injury in Massachusetts. Massachusetts Department of Public Health, Bureau of Family and Community Health, Injury and Prevention Control Program. Selassie, A. W., M. L. McCarthy, and E. E. Pickelsimer. 2003. The influence of insurance, race, and gender on emergency department disposition. Academic Emergency Medicine 10(11):1260â1270. Selassie, A. W., E. E. Pickelsimer, L. Frazier, Jr., and P. L. Ferguson. 2004. The effect of insurance status, race, and gender on ED disposition of persons with traumatic brain injury. American Journal of Emergency Medicine 22(6):465â473. Selassie, A. W., E. Zaloshnja, J. A. Langlois, T. Miller, P. Jones, and C. Steiner. 2008. Incidence of long-term disability following traumatic brain injury hospitalization, United States, 2003. Journal of Head Trauma Rehabilitation 23(2):123â131. Selecki, B. R., I. T. Ring, and D. Simpson. 1981. Injuries to the Head, Spine, and Peripheral Nerves. Neurological Society of Australia, Truama Subcommittee and Division of Health Services Research, Health Commission of New South Wales. Servadei, F., V. Antonelli, L. Betti, A. Chieregato, E. Fainardi, E. Gardini, G. Giuliani, L. Salizzato, and J. F. Kraus. 2002a. Regional brain injury epidemiology as the basis for planning brain injury treatment. The Romagna (Italy) experience. Journal of Neurosurgical Sciences 46(3â4):111â119.
EPIDEMIOLOGY OF ADULT TRAUMATIC BRAIN INJURY 101 Servadei, F., S. Bastianelli, G. Naccarato, G. Staffa, G. Morganti, and G. Gaist. 1985. Epidemiology and sequelae of head injury in San Marino Republic. Journal of Neurosurgical Sciences 29(4):297â303. Servadei, F., G. Ciucci, G. Piazza, G. Bianchedi, G. Rebucci, G. Gaist, and F. Taggi. 1988. A prospective clinical and epidemiological study of head injuries in northern Italy: The Comune of Ravenna. Italian Journal of Neurological Sciences 9(5):449â457. Servadei, F., A. Verlicchi, F. Soldano, B. Zanotti, and S. Piffer. 2002b. Descriptive epidemiology of head injury in Romagna and Trentino. Comparison between two geographically different Italian regions. Neuroepidemiology 21(6):297â304. Simpson, D., J. D. Antonio, J. B. North, I. T. Ring, B. R. Selecki, and M. F. Sewell. 1981. Fatal injuries of the head and spine: Epidemiological studies in New South Wales and South Australia. Medical Journal of Australia 2(12â13):660â664. Smith, G. S., A. L. Dannenberg, and P. J. Amoroso. 2000. Hospitalization due to injuries in the military. Evaluation of current data and recommendations on their use for injury prevention. American Journal of Preventive Medicine 18(3 Suppl):41â53. Sosin, D. M., J. E. Sniezek, and D. J. Thurman. 1996. Incidence of mild and moderate brain injury in the United States, 1991. Brain Injury 10(1):47â54. Sperry, J. L., L. M. Gentilello, J. P. Minei, R. R. Diaz-Arrastia, R. S. Friese, and S. Shafi. 2006. Waiting for the patient to "sober up": Effect of alcohol intoxication on Glasgow Coma Scale score of brain injured patients. Journal of Trauma-Injury Infection and Critical Care 61(6):1305â1311. Steudel, W. I., F. Cortbus, and K. Schwerdtfeger. 2005. Epidemiology and prevention of fatal head injuries in Germanyâtrends and the impact of the reunification. Acta Neurochirurgica 147(3):231â242. Tate, R., S. McDonald, and J. Lulham. 1998. Incidence of hospital-treated traumatic brain injury in an Australian community. Australia and New Zealand Journal of Public Health 22:419â 423. Teasdale, G., and B. Jennett. 1974. Assessment of coma and impaired consciousness. A practical scale. Lancet 2(7872):81â84. Teasdale, G. M., L. E. Pettigrew, J. T. Wilson, G. Murray, and B. Jennett. 1998. Analyzing outcome of treatment of severe head injury: A review and update on advancing the use of the Glasgow Outcome Scale. Journal of Neurotrauma 15(8):587â597. Tennant, A. 2005. Admission to hospital following head injury in England: Incidence and socio- economic associations. BMC Public Health 5(1):21. Texas Department of Health. 2004. Traumatic Brain Injury Report 1998 Data Analysis. Thurman, D., and J. Guerrero. 1999. Trends in hospitalization associated with traumatic brain injury. JAMA 282(10):954â957. Thurman, D. J., C. Alverson, K. A. Dunn, J. Guerrero, and J. E. Sniezek. 1999. Traumatic brain injury in the United States: A public health perspective. Journal of Head Trauma Rehabilitation 14(6):602â615. Thurman, D. J., L. Jeppson, C. L. Burnett, D. E. Beaudoin, M. M. Rheinberger, and J. E. Sniezek. 1996. Surveillance of traumatic brain injuries in Utah. Western Journal of Medicine 165(4):192â196.
102 GULF WAR AND HEALTH Tiret, L., E. Hausherr, M. Thicoipe, B. Garros, P. Maurette, J. P. Castel, and F. Hatton. 1990. The epidemiology of head trauma in Aquitaine (France), 1986: A community-based study of hospital admissions and deaths. International Journal of Epidemiology 19(1):133â140. Vazquez-Barquero, A., J. L. Vazquez-Barquero, O. Austin, J. Pascual, L. Gaite, and S. Herrera. 1992. The epidemiology of head injury in Cantabria. European Journal of Epidemiology 8(6):832â837. Waller, P. F., J. R. Stewart, A. R. Hansen, J. C. Stutts, C. L. Popkin, and E. A. Rodgman. 1986. The potentiating effects of alcohol on driver injury. JAMA 256(11):1461â1466. Wang, C. C., B. S. Schoenberg, S. C. Li, Y. C. Yang, X. M. Cheng, and C. L. Bolis. 1986. Brain injury due to head trauma. Epidemiology in urban areas of the People's Republic of China. Archives of Neurology 43(6):570â572. Warren, S., M. Moore, and M. S. Johnson. 1995. Traumatic head and spinal cord injuries in Alaska (1991â1993). Alaska Medicine 37(1):11â19. Whiteneck, G., C. A. Brooks, D. Mellick, C. Harrison-Felix, M. S. Terrill, and K. Noble. 2004. Population-based estimates of outcomes after hospitalization for traumatic brain injury in Colorado. Archives of Physical Medicine and Rehabilitation 85(4 Suppl 2):S73âS81. Whitman, S., R. Coonley-Hoganson, and B. T. Desai. 1984. Comparative head trauma experiences in two socioeconomically different Chicago-area communities. American Journal of Epidemiology 119:569â590. Wu, X., J. Hu, L. Zhuo, C. Fu, G. Hui, Y. Wang, W. Yang, L. Teng, S. Lu, and G. Xu. 2008. Epidemiology of traumatic brain injury in eastern China, 2004: A prospective large case study. Journal of Trauma-Injury Infection and Critical Care 64(5):1313â1319. Yates, P. J., W. H. Williams, A. Harris, A. Round, and R. Jenkins. 2006. An epidemiological study of head injuries in a UK population attending an emergency department. Journal of Neurology, Neurosurgery and Psychiatry 77(5):699â701.