In moderate to severe TBI indicators of severity, such as length of loss of consciousness (LOC), depth of coma measured with the Glasgow Coma Scale (GCS; Jennett, Teasdale, & Braakman, 1976), and duration of post-traumatic amnesia (PTA) have been found to be predictive of psychosocial, cognitive and functional outcomes (Dikmen, Machamer, Winn, & Temkin, 1995a; Dikmen, Ross, Machamer, & Temkin, 1995b; Gordon, et al., 2006; Ownsworth & McKenna, 2004). A prospective study of Dikemen and colleagues (1995a), comparing 436 hospitalised head injured individuals and 121 general trauma control participants, found that the TBI group performed significantly worse on a range of neuropsychological measures, and the magnitude and the pervasiveness of impairments was dependent on injury-severity in a dose-response relationship. In this study injury-severity was measured by length of coma (time to follow commands) on the GCS. The group with <1 hour coma (n = 121) were comparable with the trauma controls, with no significant difference on any of the measures used (motor functioning, attention, executive functioning, memory and general intellectual abilities). Selective impairments on measures of attention and memory were found in the group with coma of more than 1 hour but less than 24 hours,
and with further increase in injury-severity all measures were affected. However it is important to note that in this study there was variability in performance within severity levels. So although there was a dose-response relationship between severity and outcome, there was considerable overlap in outcome across severity groups. This finding suggests that there are no ranges of impairment that are specific to a TBI of a given severity, with the exception of extreme groups such as those with several weeks of coma.
Injury-severity has not been found to be a predictor of persisting PCS in mild injuries. The majority of mTBI patients return to normal occupational social and independent functioning within days or weeks following their injury (Alexander, 1995; Binder, Rohling, & Larrabee, 1997). However a small but significant minority continue to experience persistent cognitive, somatic and behavioural complaints usually referred to as post-concussion symptoms (PCS). While a large number of studies have
attempted to identify reasons for this, injury-severity has not been implicated in the majority of them. For example a systematic review of the literature published between 1980 and 2002 on mTBI (Carroll, et al., 2004b) failed to find a study that reported severity of injury as an independent predictor of persistent PCS. Outcome studies have also indicated that disability is as common after mild injuries as after more serious ones (Thornhill et al., 2000; Whitnall, McMillan, Murray, & Teasdale, 2006).
3.2.1.1 Measures of injury-severity.
Loss of consciousness (LOC): A number of studies have found a relationship between loss of consciousness (LOC) and poor outcome (Cifu et al., 1997; Kreutzer et al., 2003) following TBI, and a longer period of coma has been shown to be associated with skull fracture and intracranial complication (Teasdale et al., 1990). However LOC may not always correlate with injury-severity. For example Hanlon, Demery,
Martinovich and Kelly (1999) examined the roles of various acute neurological variables in relation to neuropsychological status and vocational outcome in 100 cases of mTBI, prospectively collected. They found no difference between patients who had suffered brief loss of consciousness and those without LOC. The use of LOC as a predictor of outcome also has some practical difficulties, in that individuals with mild injuries may not experience LOC, and those that do may not know its duration. For example, in a three-centre study approximately half of the sample was unable to report the duration of their LOC (Levin, et al., 1987). This may be further confounded in those who have consumed alcohol prior to their TBI, as the effects of alcohol may be indistinguishable from LOC.
Depth of coma: Depth of coma is usually measured with the Glasgow Coma Scale (GCS; Jennett, et al., 1976) an observational instrument consisting of three scales which evaluate and assign a numerical value to eye opening, verbal responses and motor response The GCS has a total score ranging from 3 to 15 and a convention has emerged (Jennett, 1998) that patients with a coma score of 8 or less are classed as severe, and those with a score of 13 or more as mild, with the others being classed as moderate.
Numerous studies have reported a significant relationship between depth of coma, as measured by the GCS and outcome. (Dikmen, et al., 1994; Wagner,
Hammond, Sasser, Wiercisiewski, & Norton, 2000). However in mild injuries its utility has been questioned (Culotta, Sementilli, Gerold, & Watts, 1996), because the range of GCS scores from 13-15, by which mTBI is conventionally defined, encompasses a group which shows significant heterogeneity. Measurement of depth of coma using the GCS also has some practical limitations, particularly in mild injuries. It is a time- dependent assessment tool which must be administered quickly, as the symptoms it catches may resolve in hours. For example a GCS score of 10 at the scene of an
accident, which would typically be classified as a moderate injury, may improve to 13, a mild injury, by the initial evaluation in the ED. Also in some injuries there may be unreliable responses in any one of the three areas that the scale assesses. For example facial swelling may restrict eye opening and drugs administered to reduce intracranial swelling may also affect responses. As GCS has to be measured at the time of injury its use is limited in studies where data is collected retrospectively, especially for mild injuries where a GCS score may not have been recorded, either in the field or in the ED.
Post traumatic amnesia (PTA): For hospitalised patients PTA may be measured prospectively using scales such as the Westmead Post Traumatic Amnesia Scale (Shores, Marosszeky, Sandanam, & Batchelor, 1986) and the Galveston Orientation and Amnesia Test (GOAT; Levin, O'Donnell, & Grossman, 1979). The GOAT assesses orientation to person, place, and time, and measures memory for events preceding and following the injury. A validity study of 52 closed head-injured patients found that the duration of impaired GOAT scores was strongly related to the acute neurosurgical ratings on the GCS (Levin, et al., 1979).
The Westmead Post Traumatic Amnesia Scale is a validated procedure for the measurement of PTA, containing twelve items assessing orientation and memory. The patient is asked seven questions, assessing orientation to time and place, and requested to remember the face and name of the assessor and three pictures of common objects. The next day the same questionnaire is repeated and the recall of face, name and pictures is checked using a standard procedure. This is repeated daily until a perfect score of 12 is recorded on three consecutive days, indicating that PTA has ended. A revised Westmead Post Traumatic Amnesia Scale (RWPTAS), designed to assess PTA on an hourly basis in mTBI patients in the ED, has also been developed and tested on 147 mTBI patients and 109 controls (RWPTAS; Ponsford et al., 2004). It was found to
be a valid measure of PTA duration in adults for use in ED with mTBI. PTA may also be measured retrospectively using structured questionnaires, and these two methods of measuring PTA have been shown to be highly correlated (McMillan, Jongen, & Greenwood, 1996).
PTA duration has been shown to be predictive of long-term outcome. For example, a study of the predictive value of acute injury characteristics in mild to moderate head injuries (van der Naalt, van Zomeren, Sluiter, & Minderhoud, 1999) found that outcome and return to work were determined more by duration of PTA than by admission GCS. They also found that 100% of their patients with PTA shorter than 24 hours had a good outcome as measured on the Glasgow Outcome Scale (GOS; Jennett & Bond, 1975) at follow-up a year later. There is evidence that, in severe TBI, duration of PTA may contribute information that is dissociable from the duration of coma, and is related to the extent of multi-focal brain lesions. Wilson, Teasdale, Hadley, Wiedmann and Lang (1994) found that patients may have short or negligible coma, but prolonged PTA. They collected information for 38 TBI patients and found that a comparison of eight patients with periods of coma less than six hours, but with PTA greater than seven days, had significantly more extensive hemispheric damage compared with the rest of the group. Ellenberg, Levin and Saydari (1996) extended these findings in a sample of 314 severely-injured patients, by showing that PTA provides incremental information, apart from coma duration, in predicting outcome at six months post-injury. Fleming, Tooth, Hassell and Chan (1999) investigated predictors of community integration and vocational outcome two to five years after rehabilitation in a population of 449 individuals with severe TBI and PTA emerged as the most useful of the injury-severity indicators.
PTA therefore has a number of advantages over LOC and depth of coma as a measure of injury-severity: it has been shown to provide information dissociable from duration and depth of coma in severe injuries, it is better able to differentiate injuries in the mild range than GCS scores, and it can be measured retrospectively, enabling classification of injuries in which LOC is not known, did not occur and/or a GCS score was not recorded in the medical files.