Introduction
It is important to obtain more insight in the measurement of health-related quality of life (HRQL) of patients with a traumatic brain injury (TBI), since there is a great need to document people’s pathways to recovery and to quantify the impact of TBI on population health over time. Although the mortality of TBI has decreased substantially in recent years, there has not been a proportionate reduction in disability due to TBI [
1]. Disability is increasingly considered an important component of population health in general and more specifically of significance for the field of injury prevention and trauma care [
2,
3]. TBI is a leading cause of long-term impairments and disabilities in functional, physical, emotional, cognitive, and social domains [
4,
5]. Disability is a complex construct and can be measured using functional outcome scales or quality of life instruments. In the field of TBI outcome research, functional measurement scales are often used to assess disability after TBI [
1]. Frequently used measures are the Glasgow Outcome Scale (GOS), GOS Extended (GOSe), Disability Rating Scale (DRS), Functional Independence Measure (FIM), Functional Assessment Measure (FAM), and the Functional Status Examination (FSE) [
6]. Functional measurement scales are useful to portray functional problems but do not capture the patient’s subjective experience of their problems [
6].
A more holistic and complete outcome measure is HRQL [
1]. Quality of life is defined by the WHO as: “the individuals’ perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns. It is a broad-ranging concept affected in a complex way by the persons’ physical health, psychological state, level of independence, social relationships, personal beliefs and their relationship to salient features of their environment” [
7]. From this definition, it becomes clear that the key factor in quality of life is the perception by the individual of his functioning. Since quality of life is a broad concept and may be influenced by numerous factors, the concept HRQL was developed. HRQL reflects an individual’s perception of how an illness and its treatment affect the physical, mental, and social aspects of his or her life [
8]. These three domains (physical, mental, and social functioning) are, however, regularly assessed without evaluating the consequences of impairment on a patient’s life, so without a patient’s evaluation of his functioning. In these cases, only health status is measured. It may be evident that quality of life is often confused with the health status.
Whilst HRQL as an outcome measure in medicine has been used for over 30 years, it is only since the past decade that it is used in patients with a TBI [
9]. Past perception was that TBI survivors would not be able to adequately rate their quality of life. As TBI encompasses multiple transient and permanent types of impairment, HRQL is recognized as an outcome variable that can provide well-standardized information on patient-perceived recovery after onset of the disease. A prerequisite to examining HRQL in patients with TBI is the availability of appropriate measures.
HRQL is usually assessed by questionnaires that will be filled out by the patient. Hence, more recently, these questions are referred to as patient-reported outcome measures (PROMs). HRQL instruments can be generic or disease-specific. Generic instruments do not take a particular condition into account and therefore allow comparisons with healthy individuals along with comparisons across various disease states. Disease-specific instruments take into account a patient’s specific health condition and therefore may be more sensitive to the consequences of the condition and more relevant to patients [
6]. These instruments do allow comparisons with healthy individuals but not with patients with other diseases. A disease-specific HRQL measure for TBI, the Quality of Life after Brain Injury instrument (QOLIBRI) has been recently developed [
10].
Some earlier reviews on the quality of life of patients with a TBI have been performed. Berger et al. [
11] have discussed the literature published before 1999 on quality of life after TBI. They found 16 studies reporting at least two domains of quality of life. Five of these considered all four domains of quality of life (physical, psychological, social, and cognitive). However, most studies identified focused on the use of impairment scales and only one reported HRQL. Di Battista and colleagues [
12] performed a review on quality of life in children and adolescents post-TBI. Eleven studies were included, of which seven used a HRQL measure. Furthermore, reviews on quality of life after TBI have been performed for specific subgroups, as mild TBI [
13], and combat veterans [
14]. In general it is stated that HRQL instruments have not yet been widely used in patients with a TBI [
9,
12,
15] and that little is known about the HRQL of patients with TBI [
9].
Particularly lacking is knowledge about the validity of HRQL instruments in patients with a TBI. Studies of high methodological quality are needed to guarantee appropriate conclusions on measurement properties (e.g. reliability, validity, and responsiveness).
This systematic review and quality assessment was conducted to describe the current state of knowledge in this field, with the aspiration to contribute to further consensus development on preferred methodologies for HRQL measurement within the TBI research field. We aimed: I) to evaluate the methodology of studies that purported to measure HRQL in patients with a TBI; II) to provide a narrative overview and perform a meta-analysis of HRQL of the most frequently used HRQL instrument(s) in patients with a TBI to gain insight into general recovery patterns and residual disability; and III) to evaluate the measurement properties of HRQL instruments used in patients with a TBI using the Consensus-Based Standards for the Selection of Health Status Measurement Instruments (COSMIN) checklist [
16].
Discussion
This systematic review of HRQL measurement in patients with a TBI aimed to provide a comprehensive insight into the methodological quality of the papers, study design, and HRQL of patients with a TBI. There was considerable methodological variation between studies, including different instruments, study population (mix), follow-up periods, and timings of assessment. The SF-36 was the most widely used HRQL instrument to estimate quality of life of patients with TBI. The validity of the SF-36 was evaluated most frequently and showed reasonable internal consistency and interpretability. The meta-analysis of SF-36 studies showed that TBI is a heterogeneous condition that encompasses a broad spectrum of HRQL. Patients with a TBI particularly reported low scores for role limitations-physical and -emotional and social functioning.
Our review is one of the few studies that have considered the measurement of HRQL in TBI. Some earlier reviews on the quality of life of patients with a TBI have been performed. The literature review of Berger [
11] published in 1999 identified 16 studies on quality of life, measured with a functional measurement instrument. Only one of these studies measured HRQL and was included in our review [
51]. Recently, Di Battista and colleagues [
12] performed a review on quality of life in children and adolescents post-TBI, in which 11 studies were found. Only five of these studies were included in our review [
24-
28]. The other six studies were not included, as they did not use a HRQL instrument (n = 5) or because they measured HRQL in caregivers [
52]. In total, our review included 44 new studies assessing HRQL in patients with a TBI.
The review may be limited by the nature of the search strategies and corresponding target words used across databases. Although every attempt was made to ensure that articles relating to the construct of HRQL were included, it is possible that some articles were missed as a result of the breadth of database searches and the vast amount of literature the search yielded. To avoid this we used a variety of literature databases, and keywords were matched to database-specific indexing terms, although some studies might still have been missed. Furthermore, reference lists of review studies and articles included in the review were screened for titles that included key terms.
Decreased HRQL during and after the first year of TBI was a common finding of the studies included. TBI negatively impacts the mental HRQL (MCS) more strongly than the physical HRQL (PCS). In the long-term, patients with a TBI, on average, still showed large deficits from full recovery when measured by population norms. Among the eight domains of the SF-36, we found that role limitations-physical (RP) and -emotional (RE) had the lowest scores for patients with a TBI, compared to the population norm scores. It is notable that Guilfoyle et al. [
44] found marked ceiling and floor effects for both the RP and RE domains. Presumably reflecting the limited number of possible scores in the RP and RE, they are inadequate for detecting change in patients with a TBI.
Our review reveals that there is still a lack of consensus about preferred HRQL instruments and study designs in the TBI field given the wide variety of different approaches used in the included studies. It is remarkable that in the 49 papers reviewed, 18 different HRQL instruments were used. Decisions regarding which HRQL measure to use will be influenced by a range of factors, e.g., availability in own language, availability of normative population values, user fees, and instrument length. Different HRQL instruments assess different domains of health, which make the comparison of study outcomes difficult. Variations in descriptive systems, weightings, and scale ranges between HRQL instruments result in different outcomes (utilities) for similar health states [
53,
54]. This supports the need for guidelines for the conduct of follow-up studies measuring TBI-related HRQL. Before guidelines can be developed for the measurement of HRQL in TBI, several general methodological issues that arise from the incorporation of HRQL measurement into research need to be addressed. Examples of these issues include HRQL instrument selection (Which one? How many?); timing of HRQL measurements (How often? Over what period of time?); and interpretation of results (medical outcomes versus HRQL outcomes, proxy versus self-report) [
9].
Challenges in HRQL assessments in TBI
The assessment of general HRQL in patients with severe TBI, cognitive impairments, and/or very young age is difficult. HRQL measures should be used with caution in these patient groups [
6,
55]. More research is needed into how HRQL measures could be modified to make them more suitable for people with severe brain damage and/or cognitive impairments.
Another major question is whether and under which conditions patients can self-report on their HRQL. Furthermore, in patients who cannot self-report on their HRQL, proxy ratings can be helpful, however taking into account that if the patient could respond, his or her judgment could differ significantly from that of the proxy. Whether a proxy can fill in a questionnaire on someone’s “subjective experience” is under discussion. Self- and proxy-report should thus be considered as complementary sources of information and not as equal replacements [
9].
Generic versus disease-specific instruments
For the TBI research field it is important to make a choice about which HRQL instrument or combination of instruments (generic and disease-specific) can best be used in patients with a TBI. There has been extensive discussion about the advantages and disadvantages of generic versus disease-specific HRQL measures. Berger et al. [
11] claim that generic HRQL measures are not appropriate for research into TBI and its treatment because the measures do not cover domains that are typically significantly affected—especially cognitive functioning. A similar criticism is leveled by Bullinger et al. [
56] who also specified that existing instruments do not cover the “existential domain”. Including that domain would expand the coverage of HRQL measures into the area typically encompassed by “social well-being”. Riemsma et al. [
55] questioned the validity and applicability of most generic quality of life measures to TBI because of self-report problems and the inadequate coverage of salient domains. Von Steinbuechel [
10] states “that the development of a disease-specific HRQL measure for TBI opens the possibility of constructing a composite outcome assessment that covers both functional outcomes and HRQL”. Such a composite assessment would help to complete the picture of outcome after brain injury and potentially give a more sensitive assessment for clinical trials. In contrast, Dijkers [
57] questioned the need to develop disease-specific TBI HRQL measures and suggests that it may be more efficient to explore whether it is possible to develop modules that quantify quality of life in areas typically omitted in generic quality of life measures—for example, cognitive functioning.
We found that the SF-36, a generic instrument, is currently the most widely used instrument to assess HRQoL after TBI. It’s available in several languages, has availability of population norms for many countries, and its track record in other disorders may be the underlying cause of the extensive use of the SF-36 in TBI [
6].
An advantage of the SF-36 is that it can be used to estimate a preference-based single-index measure for health using general population values (and therefore, it can be used in economic evaluations). The validity of the SF-36 was evaluated in three studies and showed positive results for internal consistency and interpretability. However, the SF-36 has some limitations in its application in TBI patients. The SF-36 may not be sensitive enough to detect key issues in patients with a TBI, such as cognitive dysfunction or severe physical restrictions, or patients with psychological problems, such as anxiety, depression, memory, and concentration disturbances. Furthermore, floor and ceiling effects should be kept in mind if selecting individual domains as outcome measures, particularly in the context of clinical trials, since reduced responsiveness to change increases the risk of not detecting a clinically important treatment effect [
44]. According to the current evidence, role limitations due to physical health and emotional problems (RP and RE) are less suitable as outcome measures, since marked ceiling and floor effects were found for both the RP and RE domains. Furthermore, Findler et al. [
42] noted that the SF-36 may be a more sensitive measure of health-related problems in patients with mild TBI than in those with moderate and severe TBI, since the correlations between the SF-36 domains and measures of health problems associated with TBI were weaker and more uniform in the moderate and severe TBI group (compared to the correlations in the mild TBI group).
The recently developed TBI-specific HRQL instrument the QOLIBRI seems promising. An international validation study has recently been published showing that the QOLIBRI is a valid instrument in a TBI population, has good correlation with the GOSe, and provides additional information to that obtained by the SF-36 [
45]. The validation study did not assess the QOLIBRI’s ability to detect change in HRQL over time (that is, its responsiveness). The QOLIBRI is validated in German, Finnish, Italian, French, English, and Dutch [
58]. It does not yet appear to have been used outside of validation studies, and further research is required to determine its responsiveness [
6].
The future of HRQL measurement in TBI
In 2002, an international group of clinicians and quality of life experts made several recommendations for future research of quality of life in patients with a TBI [
56]. With regard to the measurement time point, HRQL assessment was recommended to take place not upon admission to the hospital, but in the early phase of rehabilitation (i.e., within 1 year after trauma) and in the post-rehabilitation phase. Self-report should be preferred to proxy-report. Furthermore, the expert group recommended that the assessment of HRQL includes both a generic and condition-specific instrument [
56]. An instrument or combination of instruments including key problem areas in patients with a TBI is needed to assess the full impact of TBI on HRQL. Therefore, we recommend using a generic measure (SF-36) with a disease specific measure (QOLIBRI).
Fundamental research (as input for guideline development) should be undertaken alongside work on the development and validation of specific instruments. In 1999 the National Institutes of Health (NIH) consensus conference on rehabilitation of persons with TBI made a broad recommendation that generic HRQL assessment instruments must be validated for use with persons with TBI and TBI-specific instruments [
57]. To properly assess the measurement properties of HRQL instruments, studies of high methodological quality are needed. In our study we evaluated the validation studies with the COSMIN checklist [
16]. The COSMIN checklist facilitates a separate judgment of the methodological quality of the included studies and their results. This is in line with the methodology of systematic reviews of clinical trials [
59]. The studies included in this review used 18 different instruments to assess HRQL. Only six of these instruments have been validated in patients with a TBI. Validity assessment of HRQL measurements for people with TBI should be addressed in studies specifically designed for this patient population and should include information on reliability, validity, and responsiveness. When there is need for proxy assessments (e.g., in severe TBI, cognitive impairment, and children) the instrument should also be assessed for patient–proxy agreement and inter-rater agreement.
Conclusions
In conclusion, this review shows that there is considerable variation in study design between studies measuring HRQL in TBI. There are still major gaps in our understanding of how to measure the impact of TBI on personal and population health. The use of the SF-36 in combination with a TBI-specific instrument, e.g., QOLIBRI, seems promising. Development of guidelines for the measurement of HRQL in TBI with validated instruments would facilitate comparability across studies, which would produce improved estimates of TBI disability and recovery patterns.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
SP and JH carried out the search, selected included papers, independently critically appraised the selected papers, and developed the evidence tables. SP wrote the initial draft of the paper. DK performed the meta-analysis. JH, EvB, ES, and DK contributed substantially to the interpretation of study findings and writing the paper. All authors read and approved the final manuscript.