Injury, Int. J. Care Injured 46 (2015) 616–624
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Health-related quality of life after mild, moderate and severe traumatic brain injury: Patterns and predictors of suboptimal functioning during the first year after injury A.C. Scholten a,*, J.A. Haagsma a, T.M.J.C. Andriessen b, P.E. Vos c, E.W. Steyerberg a, E.F. van Beeck a, S. Polinder a a
Department of Public Health, Erasmus MC, PO Box 2040, 3000 CA Rotterdam, The Netherlands Department of Neurology, Radboud University Medical Center, PO Box 9101, 6500 HB Nijmegen, The Netherlands c Department of Neurology, Slingeland Hospital, PO Box 169, 7000 AD Doetinchem, The Netherlands b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 10 June 2014 Received in revised form 15 October 2014 Accepted 26 October 2014
Background: The Glasgow Outcome Scale Extended (GOSE) is the established functional outcome scale to assess disability following traumatic brain injury (TBI), however does not capture the patient’s subjective perspective. Health-related quality of life (HRQL) does capture the individual’s perception of disability after TBI, and has therefore been recognized as an important outcome in TBI. In contrast to GOSE, HRQL enables comparison of health outcome across various disease states and with healthy individuals. We aimed to assess functional outcome, HRQL, recovery, and predictors of 6 and 12-month outcome in a comprehensive sample of patients with mild, moderate or severe TBI, and to examine the relationship between functional impairment (GOSE) and HRQL. Methods: A prospective cohort study was conducted among a sample of 2066 adult TBI patients who attended the emergency department (ED). GOSE was determined through questionnaires or structured interviews. Questionnaires 6 and 12 months after ED treatment included socio-demographic information and HRQL measured with Short-Form Health Survey (SF-36; reflecting physical, mental and social functioning) and Perceived Quality of Life Scale (PQoL; measuring degree of satisfaction with functioning). Results: 996 TBI survivors with mild, moderate or severe TBI completed the 6-month questionnaire. Functional outcome and HRQL after moderate or severe TBI was significantly lower than after mild TBI. Patients with moderate TBI showed greatest improvement. After one year, the mild TBI group reached outcomes comparable to population norms. TBI of all severities highly affected SF-36 domains physical and social functioning, and physical and emotional role functioning. GOSE scores were highly related to all SF-36 domains and PQoL scores. Female gender, older age, co-morbidity and high ISS were strongest independent predictors of decreased HRQL at 6 and 12 months after TBI. Conclusions: HRQL and recovery patterns differ for mild, moderate and severe TBI. This study indicates that GOSE, although clinically relevant, fails to capture the subjective perspective of TBI patients, which endorses the use of HRQL as valuable addition to established instruments in assessing disability following TBI. Influence of TBI severity on recovery, together with female gender, older age, co-morbidity and high ISS should be considered in long-term follow-up and intervention programs. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Traumatic brain injury Follow-up studies Glasgow outcome scale-extended Health-related quality of life SF-36
Background Traumatic brain injury (TBI) is a leading cause of death and long-term disability, particularly in young adults. TBI can cause
* Corresponding author. Tel.: +31 10 7038465. E-mail address: [email protected] (A.C. Scholten). http://dx.doi.org/10.1016/j.injury.2014.10.064 0020–1383/ß 2014 Elsevier Ltd. All rights reserved.
assorted impairments and disabilities in functional, physical, emotional, cognitive, and social domains which drastically reduce health-related quality of life (HRQL) [1,2]. Because of major improvements in trauma care, the number of survivors of severe TBI has rapidly grown [3]. However, the disability due to TBI has not appreciably reduced [4]. This has resulted in a shift in attention from mortality towards disability of TBI patients.
A.C. Scholten et al. / Injury, Int. J. Care Injured 46 (2015) 616–624
Disability following TBI is often assessed by functional measurement scales that have been designed for TBI specifically, e.g. the Glasgow Outcome Scale (GOS) and the GOS Extended (GOSE) [4,5]. The GOS is a descriptive outcome scale with 5 categories. Five categories are believed to be too few to represent the wide range of mental and physical disability a patient can suffer following TBI [6,7]. Therefore, the GOS was extended to 8 categories, by dividing 3 categories into a lower and upper one. The GOSE is more sensitive to change than the GOS [6,8], is quick to administer, can be applied to all cases, and has clinically relevant categories. These practical advantages have led to its widespread adoption in early management studies and clinical trials. However, one criticism of scales such as the GOSE is that they fail to capture the subjective perspective (e.g. HRQL) of TBI patients [8]. HRQL reflects an individual’s perception of how an illness and its treatment affect physical, mental and social aspects of his/her life [9]. HRQL has been recognized as an important outcome in TBI, because it provides well-standardized information on recovery patterns and frequency, nature, and predictors of disabilities [10]. In contrast to the GOSE, HRQL measures enable comparison of health outcome after TBI with other diseases and the general population, and their outcome in terms of an health status on a scale from 0 (death) to 1 (perfect health) scale can be used in economic evaluations. Research has shown that even years after injury, many TBI patients still report significantly lower HRQL than the general population [1,2,11–14]. Most studies however, focus on recovery after mild [15–17] or moderate and severe [1,12,13,18,19] TBI. HRQL and recovery pattern differences between mild, moderate and severe TBI are not often studied. Large variation exists in the use of HRQL instruments to quantify the impact of TBI on population health over time. The most widely used instrument to estimate HRQL after TBI is the 36Item Short-Form (SF-36) Health Survey [20]; a multidimensional questionnaire, reflecting features of health including physical, mental, and social functioning. Another HRQL instrument that has previously been used in TBI [21,22], is the Perceived Quality of Life Scale (PQoL); a measure of the degree to which the individual is satisfied with his/her functioning, or global life satisfaction [23]. Findings from earlier studies suggest similar SF-36 and PQoL patterns after TBI [7,21,24]. Due to the heterogeneity of TBI patients and their wide array of short- and long-term recovery patterns, accurate measurement of HRQL and the impact of all severities of TBI over time is needed. Furthermore, more insight is needed in the assessment of HRQL following TBI as a potential addition to established instruments, such as the GOSE. Therefore, the current study focused on HRQL after mild, moderate and severe TBI, and on the relationship between functional outcome measured with GOSE and HRQL measured with the SF-36 (including all domains) and PQoL. The objectives of the present prospective cohort study were to (1) assess the functional outcome (GOSE), HRQL (SF-36 and PQoL), and recovery patterns at 6 and 12 months after mild, moderate and severe TBI, (2) assess the relationship and discrepancies between GOSE and HRQL for all TBI severity levels, and (3) test sociodemographic and injury-related characteristics as predictors for suboptimal functioning after TBI.
617
prospective observational cohort study on the association between demographic and clinical variables, posttraumatic complaints, and functional outcome of patients with brain injury. This study encompassed multiple outcome measures (GOSE, SF-36 and PQoL) of patients 6 and 12 months after mild, moderate and severe TBI. Between 1998 and 2010, patients admitted to the emergency department (ED) of the Radboud University Nijmegen Medical Centre (RUNMC), a level I trauma centre, with a diagnosis of mild, moderate or severe TBI were included in the RUBICS database. TBI was defined as an acute insult to the brain caused by an external physical force [28]. Mild and moderate TBI were defined by an ED Glasgow Coma Scale (GCS) score of, respectively, 13–15 [29] and 9–12 [30] after initial resuscitation at the ED or an admission GCS of, respectively, 13–15 and 9–12 followed by sedation and intubation during resuscitation for a non-neurological cause. Severe TBI was characterized by an ED GCS 8 [31] after resuscitation. Clinical data registered by a neurologist and/or neurosurgeon in the ED were collected by a research nurse and entered into the RUBICS database. The RUBICS database comprised demographic data, trauma mechanism, hospitalization, clinical injury variables, and comorbidities. Co-morbidity was defined as the presence of any co-existing medical diseases or disease processes additional to the injury that the injury patients sustained. The following diseases were assessed as co-morbid disease: asthma, chronic bronchitis, chronic non-specific lung disease (not questioned), heart disease, diabetes, back hernia or chronic backache, osteoarthritis, rheumatoid arthritis, and cancer. Further, Abbreviated Injury Scale of the Head (AISH) revised 1990 (AIS-90) [32], Injury Severity Score (ISS), and GOSE were recorded. Study participants In the current study, all patients, aged 16 years and older, with mild, moderate and severe TBI, admitted to the ED of RUNMC, between June 2003 and June 2010, who completed the 6 month questionnaire, were selected from the RUBICS database. Exclusion criteria were no informed consent, alcohol or drug abuse or dementia, unknown address, and inability to speak or write Dutch. Furthermore, patients who died within 6 months were excluded. Written informed consent was obtained from all participating patients. Functional outcome measure The GOSE scores functional outcome with eight questions covering consciousness, independence at home, major social roles (work, social and leisure activities, family and friendships), and return to normal life [33]. It results in an 8-point scale classifying functional outcome from 1 (dead) to 8 (complete recovery). GOSE scores were determined using a structured interview during regular visits to the outpatient clinic or during consultation by telephone [34]. Patients not visiting the outpatient clinic were sent a GOSE questionnaire by regular mail, and when not returned a reminder was sent [35]. Finally, we attempted to reach all nonresponding patients by telephone to acquire an outcome score. Assessment often took place at 6 (70%) and 12 (66%) months postinjury. Outcomes obtained within a 2 months range were also accepted if no outcome at exactly 6 or 12 months was available. Patients with a GOSE score of 1 (dead) were excluded from this study.
Methods Health-related quality of life measures Study design Data for the present study were obtained from the Radboud University Brain Injury Cohort Study (RUBICS) [25–27]. RUBICS is a
HRQL was determined using the SF-36 (Version 1) and PQoL. Patients were asked to fill in a questionnaire, which included the HRQL measurements at 6 and 12 months post-injury.
618
A.C. Scholten et al. / Injury, Int. J. Care Injured 46 (2015) 616–624
The SF-36 is the most frequently used HRQL instrument in TBI and showed positive results for internal consistency and validity in a TBI population [36,37]. It is a 36-item questionnaire that covers eight domains of health status: physical functioning (PF), role limitations related to physical health problems (RP), bodily pain (BP), general health perception (GH), vitality (VT), social functioning (SF), role limitations related to emotional problems (RE), and mental health (MH) [38]. For each domain, a summation of item responses is linearly transformed into a score ranging from 0 to 100. Physical (PCS) and mental summary scores (MCS) are calculated by standardizing patients’ scores, by subtracting Dutch subscale means from each individual’s subscale scores and dividing the result by Dutch standard deviation to generate Z-scores [39]. In order to facilitate international comparison [40], Z-scores are multiplied by United State (US) subscale factor coefficients for PCS and MCS and summed over all eight subscales into PCS and MCS sums. Both sums were re-scaled into T-scores, with a mean of 50 and standard deviation of 10 for the US norm population [40]. Missing values at 6 and 12 months of 10.8% and 10.7% of the respondents were replaced by the mean value of the respondents’ completed items in the same scale, provided that at least 50% of the items within that scale had been completed [38]. The PQoL was initially developed as a cognitive appraisal of life satisfaction for patients after intensive medical care [41]. It has been used for adults with chronic neurologic disability [21,42] and showed good internal reliability in a TBI population [22]. PQoL measures the degree to which the individual is satisfied with his/ her functioning on an 11-point scale ranging from 0 (extremely dissatisfied) to 10 (extremely satisfied). It consists of 19 items in 3 domains (physical, cognitive and social), assessing 10 areas of functioning including physical health, thinking and remembering, family relationships, community participation and leisure, work and income, and meaning and purpose of life. PQoL scores may be considered a measure of global life satisfaction, with PQoL < 7.5 ‘‘Dissatisfied’’ and PQoL > 7.5 ‘‘Satisfied’’ [23]. A previous study showed that the PQoL scores in adults without chronic conditions, range between 8.3 and 8.5 [42]. We used the mean score (range of 0 to 10) in our analyses. Because PQoL scores can only be computed in case of complete information on all items, missing values of 11.4% respondents at both 6 and 12 months were estimated by hot deck imputation per domain if at least 50% of the items within that domain had been completed, using the reported values of respondents with similar scores on the items that were reported in that domain [43].
Socio-demographic and injury-related characteristics were tested as predictors of HRQL measured with the SF-36 domains and PQoL score 6 and 12 months after TBI in a simple linear regression analysis. We included the socio-demographic variables gender (male/female), age (continuous), a dummy-coded variable for primary/secondary (reference), higher and academic education, and co-morbidity (continuous). The injury-related variables were a dummy-coded variable for GCS 13–15 (mild TBI, reference), GCS 9– 12 (moderate TBI), and GCS 3–8 (severe TBI), ISS (continuous), and AISH (continuous). Variables associated with outcome (p < 0.20 in the univariable analysis) were included in stepwise multivariable linear regression analyses [44,45]. Analysis of the variance inflation factor (VIF) showed low VIF-values (all VIFs < 1.9), indicating that higher order collinearity was not problematic in this study. None of the variables showed high correlations between each other (all variables r < 0.7). All statistical analyses were carried out using the statistical package SPSS for Windows, version 21 (IBM SPSS Statistics, SPSS Inc, Chicago, IL). Results Patient characteristics Between June 2003 and June 2010, 4576 patients with TBI of 16 years old and older were admitted to the ED of the Radboud University Nijmegen Medical Centre (Fig. 1). Of these patients, 688 were excluded from the database due to inability to speak Dutch, or death. Mortality for mild, moderate and severe TBI was, respectively, 4.1%, 17.2%, and 35.7%. Of the 3888 eligible patients, 1602 did not receive a first questionnaire due to various reasons (e.g. dementia, unknown address). Of the 2286 TBI patients that received a 6 month questionnaire, 996 (44%) completed the 6 month questionnaire, of whom 386 (39%) filled in the 12 month questionnaire. Of the 996 respondents, 797 had mild (80%), 50 moderate (5%) and 149 severe (15%) TBI (Table 1). The median age of the respondents was 44 years and 63% were male. Respondents with severe TBI were significantly younger than respondents with mild or moderate TBI (median age 39 years versus 45 years and 47
Data and statistical analysis Analysis of variance (ANOVA), Chi-square statistics (dichotomous variables), and Student’s t tests (continuous variables) were used for between-group comparisons on socio-demographic and injury-related variables, and the influence of AISH (
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Health-related quality of life after mild, moderate and severe traumatic brain injury: Patterns and predictors of suboptimal functioning during the first year after injury A.C. Scholten a,*, J.A. Haagsma a, T.M.J.C. Andriessen b, P.E. Vos c, E.W. Steyerberg a, E.F. van Beeck a, S. Polinder a a
Department of Public Health, Erasmus MC, PO Box 2040, 3000 CA Rotterdam, The Netherlands Department of Neurology, Radboud University Medical Center, PO Box 9101, 6500 HB Nijmegen, The Netherlands c Department of Neurology, Slingeland Hospital, PO Box 169, 7000 AD Doetinchem, The Netherlands b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 10 June 2014 Received in revised form 15 October 2014 Accepted 26 October 2014
Background: The Glasgow Outcome Scale Extended (GOSE) is the established functional outcome scale to assess disability following traumatic brain injury (TBI), however does not capture the patient’s subjective perspective. Health-related quality of life (HRQL) does capture the individual’s perception of disability after TBI, and has therefore been recognized as an important outcome in TBI. In contrast to GOSE, HRQL enables comparison of health outcome across various disease states and with healthy individuals. We aimed to assess functional outcome, HRQL, recovery, and predictors of 6 and 12-month outcome in a comprehensive sample of patients with mild, moderate or severe TBI, and to examine the relationship between functional impairment (GOSE) and HRQL. Methods: A prospective cohort study was conducted among a sample of 2066 adult TBI patients who attended the emergency department (ED). GOSE was determined through questionnaires or structured interviews. Questionnaires 6 and 12 months after ED treatment included socio-demographic information and HRQL measured with Short-Form Health Survey (SF-36; reflecting physical, mental and social functioning) and Perceived Quality of Life Scale (PQoL; measuring degree of satisfaction with functioning). Results: 996 TBI survivors with mild, moderate or severe TBI completed the 6-month questionnaire. Functional outcome and HRQL after moderate or severe TBI was significantly lower than after mild TBI. Patients with moderate TBI showed greatest improvement. After one year, the mild TBI group reached outcomes comparable to population norms. TBI of all severities highly affected SF-36 domains physical and social functioning, and physical and emotional role functioning. GOSE scores were highly related to all SF-36 domains and PQoL scores. Female gender, older age, co-morbidity and high ISS were strongest independent predictors of decreased HRQL at 6 and 12 months after TBI. Conclusions: HRQL and recovery patterns differ for mild, moderate and severe TBI. This study indicates that GOSE, although clinically relevant, fails to capture the subjective perspective of TBI patients, which endorses the use of HRQL as valuable addition to established instruments in assessing disability following TBI. Influence of TBI severity on recovery, together with female gender, older age, co-morbidity and high ISS should be considered in long-term follow-up and intervention programs. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Traumatic brain injury Follow-up studies Glasgow outcome scale-extended Health-related quality of life SF-36
Background Traumatic brain injury (TBI) is a leading cause of death and long-term disability, particularly in young adults. TBI can cause
* Corresponding author. Tel.: +31 10 7038465. E-mail address: [email protected] (A.C. Scholten). http://dx.doi.org/10.1016/j.injury.2014.10.064 0020–1383/ß 2014 Elsevier Ltd. All rights reserved.
assorted impairments and disabilities in functional, physical, emotional, cognitive, and social domains which drastically reduce health-related quality of life (HRQL) [1,2]. Because of major improvements in trauma care, the number of survivors of severe TBI has rapidly grown [3]. However, the disability due to TBI has not appreciably reduced [4]. This has resulted in a shift in attention from mortality towards disability of TBI patients.
A.C. Scholten et al. / Injury, Int. J. Care Injured 46 (2015) 616–624
Disability following TBI is often assessed by functional measurement scales that have been designed for TBI specifically, e.g. the Glasgow Outcome Scale (GOS) and the GOS Extended (GOSE) [4,5]. The GOS is a descriptive outcome scale with 5 categories. Five categories are believed to be too few to represent the wide range of mental and physical disability a patient can suffer following TBI [6,7]. Therefore, the GOS was extended to 8 categories, by dividing 3 categories into a lower and upper one. The GOSE is more sensitive to change than the GOS [6,8], is quick to administer, can be applied to all cases, and has clinically relevant categories. These practical advantages have led to its widespread adoption in early management studies and clinical trials. However, one criticism of scales such as the GOSE is that they fail to capture the subjective perspective (e.g. HRQL) of TBI patients [8]. HRQL reflects an individual’s perception of how an illness and its treatment affect physical, mental and social aspects of his/her life [9]. HRQL has been recognized as an important outcome in TBI, because it provides well-standardized information on recovery patterns and frequency, nature, and predictors of disabilities [10]. In contrast to the GOSE, HRQL measures enable comparison of health outcome after TBI with other diseases and the general population, and their outcome in terms of an health status on a scale from 0 (death) to 1 (perfect health) scale can be used in economic evaluations. Research has shown that even years after injury, many TBI patients still report significantly lower HRQL than the general population [1,2,11–14]. Most studies however, focus on recovery after mild [15–17] or moderate and severe [1,12,13,18,19] TBI. HRQL and recovery pattern differences between mild, moderate and severe TBI are not often studied. Large variation exists in the use of HRQL instruments to quantify the impact of TBI on population health over time. The most widely used instrument to estimate HRQL after TBI is the 36Item Short-Form (SF-36) Health Survey [20]; a multidimensional questionnaire, reflecting features of health including physical, mental, and social functioning. Another HRQL instrument that has previously been used in TBI [21,22], is the Perceived Quality of Life Scale (PQoL); a measure of the degree to which the individual is satisfied with his/her functioning, or global life satisfaction [23]. Findings from earlier studies suggest similar SF-36 and PQoL patterns after TBI [7,21,24]. Due to the heterogeneity of TBI patients and their wide array of short- and long-term recovery patterns, accurate measurement of HRQL and the impact of all severities of TBI over time is needed. Furthermore, more insight is needed in the assessment of HRQL following TBI as a potential addition to established instruments, such as the GOSE. Therefore, the current study focused on HRQL after mild, moderate and severe TBI, and on the relationship between functional outcome measured with GOSE and HRQL measured with the SF-36 (including all domains) and PQoL. The objectives of the present prospective cohort study were to (1) assess the functional outcome (GOSE), HRQL (SF-36 and PQoL), and recovery patterns at 6 and 12 months after mild, moderate and severe TBI, (2) assess the relationship and discrepancies between GOSE and HRQL for all TBI severity levels, and (3) test sociodemographic and injury-related characteristics as predictors for suboptimal functioning after TBI.
617
prospective observational cohort study on the association between demographic and clinical variables, posttraumatic complaints, and functional outcome of patients with brain injury. This study encompassed multiple outcome measures (GOSE, SF-36 and PQoL) of patients 6 and 12 months after mild, moderate and severe TBI. Between 1998 and 2010, patients admitted to the emergency department (ED) of the Radboud University Nijmegen Medical Centre (RUNMC), a level I trauma centre, with a diagnosis of mild, moderate or severe TBI were included in the RUBICS database. TBI was defined as an acute insult to the brain caused by an external physical force [28]. Mild and moderate TBI were defined by an ED Glasgow Coma Scale (GCS) score of, respectively, 13–15 [29] and 9–12 [30] after initial resuscitation at the ED or an admission GCS of, respectively, 13–15 and 9–12 followed by sedation and intubation during resuscitation for a non-neurological cause. Severe TBI was characterized by an ED GCS 8 [31] after resuscitation. Clinical data registered by a neurologist and/or neurosurgeon in the ED were collected by a research nurse and entered into the RUBICS database. The RUBICS database comprised demographic data, trauma mechanism, hospitalization, clinical injury variables, and comorbidities. Co-morbidity was defined as the presence of any co-existing medical diseases or disease processes additional to the injury that the injury patients sustained. The following diseases were assessed as co-morbid disease: asthma, chronic bronchitis, chronic non-specific lung disease (not questioned), heart disease, diabetes, back hernia or chronic backache, osteoarthritis, rheumatoid arthritis, and cancer. Further, Abbreviated Injury Scale of the Head (AISH) revised 1990 (AIS-90) [32], Injury Severity Score (ISS), and GOSE were recorded. Study participants In the current study, all patients, aged 16 years and older, with mild, moderate and severe TBI, admitted to the ED of RUNMC, between June 2003 and June 2010, who completed the 6 month questionnaire, were selected from the RUBICS database. Exclusion criteria were no informed consent, alcohol or drug abuse or dementia, unknown address, and inability to speak or write Dutch. Furthermore, patients who died within 6 months were excluded. Written informed consent was obtained from all participating patients. Functional outcome measure The GOSE scores functional outcome with eight questions covering consciousness, independence at home, major social roles (work, social and leisure activities, family and friendships), and return to normal life [33]. It results in an 8-point scale classifying functional outcome from 1 (dead) to 8 (complete recovery). GOSE scores were determined using a structured interview during regular visits to the outpatient clinic or during consultation by telephone [34]. Patients not visiting the outpatient clinic were sent a GOSE questionnaire by regular mail, and when not returned a reminder was sent [35]. Finally, we attempted to reach all nonresponding patients by telephone to acquire an outcome score. Assessment often took place at 6 (70%) and 12 (66%) months postinjury. Outcomes obtained within a 2 months range were also accepted if no outcome at exactly 6 or 12 months was available. Patients with a GOSE score of 1 (dead) were excluded from this study.
Methods Health-related quality of life measures Study design Data for the present study were obtained from the Radboud University Brain Injury Cohort Study (RUBICS) [25–27]. RUBICS is a
HRQL was determined using the SF-36 (Version 1) and PQoL. Patients were asked to fill in a questionnaire, which included the HRQL measurements at 6 and 12 months post-injury.
618
A.C. Scholten et al. / Injury, Int. J. Care Injured 46 (2015) 616–624
The SF-36 is the most frequently used HRQL instrument in TBI and showed positive results for internal consistency and validity in a TBI population [36,37]. It is a 36-item questionnaire that covers eight domains of health status: physical functioning (PF), role limitations related to physical health problems (RP), bodily pain (BP), general health perception (GH), vitality (VT), social functioning (SF), role limitations related to emotional problems (RE), and mental health (MH) [38]. For each domain, a summation of item responses is linearly transformed into a score ranging from 0 to 100. Physical (PCS) and mental summary scores (MCS) are calculated by standardizing patients’ scores, by subtracting Dutch subscale means from each individual’s subscale scores and dividing the result by Dutch standard deviation to generate Z-scores [39]. In order to facilitate international comparison [40], Z-scores are multiplied by United State (US) subscale factor coefficients for PCS and MCS and summed over all eight subscales into PCS and MCS sums. Both sums were re-scaled into T-scores, with a mean of 50 and standard deviation of 10 for the US norm population [40]. Missing values at 6 and 12 months of 10.8% and 10.7% of the respondents were replaced by the mean value of the respondents’ completed items in the same scale, provided that at least 50% of the items within that scale had been completed [38]. The PQoL was initially developed as a cognitive appraisal of life satisfaction for patients after intensive medical care [41]. It has been used for adults with chronic neurologic disability [21,42] and showed good internal reliability in a TBI population [22]. PQoL measures the degree to which the individual is satisfied with his/ her functioning on an 11-point scale ranging from 0 (extremely dissatisfied) to 10 (extremely satisfied). It consists of 19 items in 3 domains (physical, cognitive and social), assessing 10 areas of functioning including physical health, thinking and remembering, family relationships, community participation and leisure, work and income, and meaning and purpose of life. PQoL scores may be considered a measure of global life satisfaction, with PQoL < 7.5 ‘‘Dissatisfied’’ and PQoL > 7.5 ‘‘Satisfied’’ [23]. A previous study showed that the PQoL scores in adults without chronic conditions, range between 8.3 and 8.5 [42]. We used the mean score (range of 0 to 10) in our analyses. Because PQoL scores can only be computed in case of complete information on all items, missing values of 11.4% respondents at both 6 and 12 months were estimated by hot deck imputation per domain if at least 50% of the items within that domain had been completed, using the reported values of respondents with similar scores on the items that were reported in that domain [43].
Socio-demographic and injury-related characteristics were tested as predictors of HRQL measured with the SF-36 domains and PQoL score 6 and 12 months after TBI in a simple linear regression analysis. We included the socio-demographic variables gender (male/female), age (continuous), a dummy-coded variable for primary/secondary (reference), higher and academic education, and co-morbidity (continuous). The injury-related variables were a dummy-coded variable for GCS 13–15 (mild TBI, reference), GCS 9– 12 (moderate TBI), and GCS 3–8 (severe TBI), ISS (continuous), and AISH (continuous). Variables associated with outcome (p < 0.20 in the univariable analysis) were included in stepwise multivariable linear regression analyses [44,45]. Analysis of the variance inflation factor (VIF) showed low VIF-values (all VIFs < 1.9), indicating that higher order collinearity was not problematic in this study. None of the variables showed high correlations between each other (all variables r < 0.7). All statistical analyses were carried out using the statistical package SPSS for Windows, version 21 (IBM SPSS Statistics, SPSS Inc, Chicago, IL). Results Patient characteristics Between June 2003 and June 2010, 4576 patients with TBI of 16 years old and older were admitted to the ED of the Radboud University Nijmegen Medical Centre (Fig. 1). Of these patients, 688 were excluded from the database due to inability to speak Dutch, or death. Mortality for mild, moderate and severe TBI was, respectively, 4.1%, 17.2%, and 35.7%. Of the 3888 eligible patients, 1602 did not receive a first questionnaire due to various reasons (e.g. dementia, unknown address). Of the 2286 TBI patients that received a 6 month questionnaire, 996 (44%) completed the 6 month questionnaire, of whom 386 (39%) filled in the 12 month questionnaire. Of the 996 respondents, 797 had mild (80%), 50 moderate (5%) and 149 severe (15%) TBI (Table 1). The median age of the respondents was 44 years and 63% were male. Respondents with severe TBI were significantly younger than respondents with mild or moderate TBI (median age 39 years versus 45 years and 47
Data and statistical analysis Analysis of variance (ANOVA), Chi-square statistics (dichotomous variables), and Student’s t tests (continuous variables) were used for between-group comparisons on socio-demographic and injury-related variables, and the influence of AISH (
