Under its Evidence-based Practice Program, the Agency for Health Care Policy and Research (AHCPR) is developing scientific information for other agencies and organizations on which to base clinical guidelines, performance measures, and other quality improvement tools. Contractor institutions review all relevant scientific literature on assigned clinical care topics and produce evidence reports and technology assessments, conduct research on methodologies and the effectiveness of their implementation, and participate in technical assistance activities.
Overview / Impact of TBI / Key Questions / Reporting the Evidence / Methodology / Findings / Future Research / Availability of Full Report
Advances in medical technology and improvements in regional trauma services have increased the number of survivors of traumatic brain injury (TBI), producing the social consequences and medical challenges of a growing pool of people with disabilities. Wider awareness of the scope of the problem and its consequences for society has led to rapid growth in the rehabilitation industry. Because of this growth and particularly because clinical rehabilitation strategies vary widely, many groups are interested in the effectiveness of rehabilitation for TBI.
To address this need to identify and assess evidence on TBI rehabilitation, the Agency for Health Care Policy and Research (AHCPR) awarded a contract to Oregon Health Sciences University for a review of published reports and compilation of an evidence report. This summary highlights information presented in the full report.
Injury is the leading cause of mortality among Americans under 45 years of age; TBI is responsible for the majority of these deaths. An estimated 56,000 lives are lost in the United States each year to TBI. Motor vehicle accidents, followed by gunshot injuries and falls, are the leading causes of injuries resulting in death from TBI. Males are 3.4 times as likely as females to die of TBI. About 50 percent of people who sustain TBI are intoxicated at the time of injury.
In a recent analysis based on hospital discharge data and vital statistics, the annual incidence of TBI in the United States was estimated to be 102.8 per 100,000. In males, the incidence peaks between the ages of 15 and 24 (248.3 per 100,000) and again above 75 years of age (243.4 per 100,000). The incidence in females peaks in the same age groups, but the absolute rates are lower (101.6 and 154.9, respectively). These rates underestimate the true incidence of head trauma because patients with milder symptoms at the time of injury are usually not hospitalized.
About three-quarters of traumatic brain injuries that require hospitalization are nonfatal. Each year, about 80,000 survivors of TBI will incur some disability or require increased medical care. Direct medical costs for TBI treatment have been estimated at $48.3 billion per year, including the costs of hospitalization for acute care and various rehabilitation services.
In the years 1988 to 1992, reports of average length of stay (LOS) for the initial admission for inpatient rehabilitation range from 40 to 165 days. In one multicenter study (the Model Systems study), the average rehabilitation LOS was 61 days, and the average charge was $64,648 exclusive of physician fees. Total charges averaged $154,256. In more recent studies performed in the early 1990s, rehabilitation LOS and charges were lower, ranging from 19 days and $24,000 for patients with milder injuries to 27 days and $38,000 for those with severe injuries. In the Medicare population in 1994, mean charges for patients admitted for brain injury (excluding stroke) were $42,056.
To focus attention on important questions, the life of an adult survivor of TBI was characterized by the developers of the report in terms of five phases. The first phase is pre-injury. Medical treatment is divided into two phases: the acute (or immediate) treatment phase and the intensive treatment phase, lasting days to weeks. The rehabilitation phase may last months to years. The survivor phase implies the remaining life of the person with TBI and involves continual development and adjustment. This division into phases clarifies the three challenges to assessing the efficacy of rehabilitation discussed above. For each phase, patient populations, interventions, and outcome measures were identified, and the literature was reviewed to answer key questions identified by technical experts.
The following three questions about the status of brain injury research underlie uncertainty about the effectiveness of rehabilitation services.
Today, a person's path to rehabilitation after sustaining brain injury may be determined by the mechanism of injury, the resources of the community, the person's employment or financial status, the consent of the family, and/or the accuracy of the emergency department diagnosis. While a few metropolitan areas have organized referral systems that connect patients with resources and rehabilitation programs, systematic methods for evaluating the needs of people who have sustained brain injury and referring them to appropriate programs are unusual. Without knowing the efficacy of rehabilitation methods in their specific applications, systematic referral that produces the desired result is not possible.
Two panels of experts worked with the research team to identify key questions in the rehabilitation and survivor phases for adults with TBI. The first panel was composed of two physiatrists, a survivor of TBI, the wife of a survivor of TBI, a State vocational rehabilitation counselor, a neuropsychologist, a psychologist, a clinical coordinator of an outpatient TBI rehabilitation program, and a rehabilitation clinical nurse specialist, all from the Portland, OR, area. The second panel was composed of nationally recognized experts in rehabilitation.
The panels formulated five questions pertaining to the phases of recovery described above. These questions addressed the effectiveness of (1) early rehabilitation in the acute care setting (timing), (2) intensity of rehabilitation, (3) cognitive rehabilitation, (4) supported employment, and (5) care coordination (case management). For each of these questions, members of the research team worked with panelists to write a brief rationale for the question, define key terms, and specify the relevant patient populations, interventions, and outcome measures to be examined in the literature review. The questions were:
Select for Findings.
A MEDLINE® search (1976 to 1997), supplemented by searches of HealthSTAR (1995 to 1997), CINAHL (1982 to 1997), and PsycINFO (1984 to 1997), produced a total of 3,098 references to be considered for inclusion; of these, 569 applied to questions 1 and 2, 600 applied to question 3, 392 applied to question 4, and 975 applied to question 5.
Abstracts of each article retrieved by these searches were reviewed independently by two members of the research team, who applied predefined, broad eligibility criteria. When the two reviewers disagreed, a third reviewer read the abstract and cast the deciding vote on whether it should be included. In the event a reference did not have an abstract, and the title for the reference was not sufficient for determination of status, the article was retrieved and reviewed to determine its eligibility. The two reviewers examined each abstract and indicated whether it met the inclusion criteria and, if not, the reason for exclusion. If the abstract was eligible, or if it did not contain sufficient information to determine eligibility, the full text of the article was retrieved for review in the next phase of the selection process.
Eighty-seven articles pertaining to questions 1 and 2, 114 articles for question 3, 93 articles for question 4, and 69 articles for question 5 passed the eligibility screen. Sixty-seven additional articles were recommended for inclusion by experts or by review of reference lists of review articles. In all, 363 articles were retrieved for review and abstraction.
Additional criteria for inclusion were defined separately for each of the five questions. The criteria varied because the necessary types of studies varied from question to question. Articles that applied to more than one question were maintained as duplicates (or triplicates, etc.) in each question-specific file, so they could be considered for inclusion based on their relevance to each question.
An instrument was designed to record data abstracted from each eligible article. The instrument includes items for patient characteristics, interventions, cointerventions, outcomes, study methods, relevance to the specific research questions, and results of the study. The instrument has two components: the first four pages of the instrument apply to all articles specified for inclusion in the study; the remaining pages are individual instruments that apply to one of the five questions. To abstract an article, a reader used the initial abstraction instrument plus one or more of the five question instruments.
The first few questions of the initial abstraction instrument allowed the reviewer to determine if the article actually met the eligibility criteria for inclusion in the report. If an article was determined to be ineligible, it was passed to a second reader for confirmation. The remaining articles were subjected to the full abstraction protocol.
A three-level system was used to rate individual studies. Well-designed randomized controlled trials (RCTs) were rated as Class I. Studies rated as Class II were RCTs with design flaws; well-done, prospective, quasiexperimental or longitudinal studies; and case-control studies. Case reports, uncontrolled case series, and expert or consensus opinion were generally rated Class III. A well-done, prospective, multicenter or population-based case series can provide valuable information that, in some ways, is more reliable than data from a randomized trial done in a highly selected sample of patients. However, when used to make inferences about effectiveness, an uncontrolled case series is generally classified as Class III, indicating the lowest level of confidence.
A "gray zone" exists between Class II and definite Class III articles. Much of the research in rehabilitation uses quasi-experimental designs. In these observational study designs, control subjects are sometimes identified from a separate patient population. For instance, one group of researchers compared patients undergoing inpatient rehabilitation to a sample of people with TBI who had been treated in a region of the country where formal inpatient TBI rehabilitation was not available. This was an entirely separate patient group, and all the data except outcome measures came from an independent database.
The main difficulty with the quasiexperimental design is lack of control over the constitution of the compared groups. Because there is no randomization and generally no control over the details of the selection process through which the patients received their separate therapies, the groups are likely to differ in the frequency of characteristics that are associated with the outcomes of interest. Even when significant efforts are made to match the experimental and the quasi-control groups, it is likely that significant differences between the groups will remain.
Much of the literature relevant to the five questions addressed in this effort falls into the "gray zone" between Class II and Class III. For this reason, critical appraisal of key studies played a particularly important role in this review. A number of characteristics of these studies were considered relevant to all rehabilitation questions and were recorded in the data abstraction form. Evaluation of the following factors played a major role in critically appraising these articles:
The criteria used to classify articles and the features to be considered in critically appraising them were discussed at the subcommittee, committee, national expert panel, and Aspen Neurobehavioral Conference levels with the goal of maintaining consensus at least on the relative stratification of individual articles.
Evidence tables were constructed to summarize the best evidence about effectiveness pertaining to each question. No randomized trials and only a few quasiexperimental studies were available for questions 1 and 2. There were a large number of relevant observational studies of important relationships (for example, the relation of patient characteristics to outcome); studies that concerned individual causal links or relationships in evidence tables were not summarized. For question 3, addressing cognitive rehabilitation, 15 randomized controlled trials and comparative studies that met specified inclusion criteria were placed into evidence tables. All comparative studies located for the last two questions, which addressed supported employment and care coordination, were included in evidence tables.
For each of the five questions, subcommittees were formed consisting of one or two members of the research team and one or two members of the local technical panel. Each subcommittee was chaired by a member of the research team. Key articles relevant to the assigned question were reviewed in depth by all members of the subcommittees. These reviews were discussed among the various members of the subcommittees, and the results were summarized by the chair. This was an effort to ensure that the summary statements on the research questions reflected the expertise and experience of a variety of technical experts with relevant skills and training. These interpretive efforts addressed the methods and results of individual studies, their rating, and their scientific importance.
All of the critical articles for the five questions were individually read by the principal investigator. Summaries were presented and discussed with national experts at the Aspen Neurobehavioral Conference in April 1998.
One small, retrospective, observational study from a single rehabilitation facility supports an association between the acute institution of formalized, multidisciplinary, physiatrist-driven TBI rehabilitation and decreased length of stay (acute hospital and acute rehabilitation) and some measures of short-term physiologic (noncognitive) patient outcomes. The level of evidence is Class III. This study concerned adult patients with severe brain injury (Glasgow Coma Scale 3-8); there is no evidence from comparative studies for or against early rehabilitation in patients with mild and moderate injury.
When measured as the hours of application of individual or grouped therapies, there is no indication that the intensity of acute, inpatient TBI rehabilitation is related to outcome. Because of methodological weaknesses, however, previous studies are likely to have missed a significant relationship if one exists (a type 2 error). These studies contained insufficient information about severity of injury and baseline function to ensure the comparability of compared groups. Also, these studies did not consider the quality of individual treatments, their lack of autonomy in the cognitive realm, and the delivery milieu. One or more of these factors may affect the outcome of care more than the time spent in each modality. Therefore, future research into efficacy of acute inpatient TBI rehabilitation must more adequately measure such factors and include the factors in their predictive models. Future studies also must employ a wider spectrum of outcome measures, including measurement of outcomes across longer periods of time after discharge.
From a clinical aspect, the evidence does not support equating different TBI rehabilitation delivery systems based on equivalent times of patient exposure to various therapeutic modalities. For example, this analysis would not support predicting that patient benefit would be equal if an equal time spectrum of rehabilitation therapies were delivered at a rehabilitation center as compared with a skilled nursing facility. More detailed analysis of factors involved in predicting response to rehabilitation modalities must be considered in approaching such questions.
Additionally, mandating a minimum number of hours of applied therapy for all TBI patients is not supported by the present state of scientific knowledge. How much of which intervention(s) optimizes recovery in a given type of patient has been inadequately studied. It is certainly reasonable to avoid situations in which patients do not receive potentially beneficial treatment. Based on the above studies, however, defining a minimal rehabilitation program in terms of time of applied therapy is not likely to optimize either the therapists' time or patients' recovery. It is probable that specific basic programs will have to be related to individual patient groups. Developing such algorithms requires further research.
Many patients who suffer TBI do not enter acute inpatient rehabilitation. Only one study of the effectiveness of inpatient rehabilitation included a comparison group of patients who did not undergo inpatient rehabilitation. Future studies should compare acute, inpatient rehabilitation to commonly used alternatives to inpatient rehabilitation, such as care in a well-staffed skilled nursing facility or in less intense variations of acute rehabilitation. Very little is known about the outcomes of TBI in these settings.
There is evidence from two small studies (Class I and Class III) that a personally adapted electronic device, a notebook, and an alarm wristwatch reduce everyday memory failures for people with TBI. There is evidence from one study (Class II[a]) that compensatory cognitive rehabilitation (CCR) reduces anxiety and improves self-concept and relationships for people with TBI. Evidence from two studies (Class I and Class II[b]) supports the use of computer-aided cognitive rehabilitation (CACR) to improve immediate recall on neuropsychological testing, but the clinical importance of this finding has not been validated.
Class II evidence indicates that supported employment can improve the vocational outcomes of TBI survivors. Nearly all information about supported employment comes from two bodies of work, each of which used different experimental designs and different models of supported employment. The findings have not been replicated in other settings or by other centers, so the generalizability of these programs remains untested.
There have been very few studies on the effectiveness of case management, and the results of these studies are mixed. The only outcome for which there were results in the same direction from two or more studies pertained to changes in vocational status. This was associated with the single case-manager and insurance approach, as well as with the combined nurse and vocational case-manager model. There were conflicting results about the effects of case management on disability or functional status, living status, family impact, and other aspects, and some findings were mentioned in only one study. The clinical trial resulted in no functional status changes among case-managed subjects, despite an extended period of rehabilitation. However, when two forms of case management were compared, both the single and multiple case-manager/insurance approaches showed significant functional improvements.
The evidence report identifies the following areas for future research.
The full evidence report from which this summary was taken was prepared for AHCPR by the Oregon Health Sciences University, Portland, OR, under contract No. 290-97-0018. Printed copies may be obtained free of charge from the Publications Clearinghouse by calling 1-800-358-9295. Requesters should ask for Evidence Report/Technology Assessment Number 2, Rehabilitation for Traumatic Brain Injury (AHCPR Publication No. 99-E006). The Evidence Report is available online at the National Library of Medicine Bookshelf.
AHCPR Publication No. 99-E005
Current as of October 1999
Internet Citation:
Rehabilitation for Traumatic Brain Injury. Summary, Evidence Report/Technology Assessment: Number 2, December 1998. Agency for Health Care Policy and Research, Rockville, MD. http://www.ahrq.gov/clinic/epcsums/tbisumm.htm
Return EPC Evidence Reports
Clinical Information
AHRQ Home Page
Department of Health and Human
Services