Comparison of Four Quality of Life Inventories for Patients with Traumatic Brain Injuries and Orthopedic Injuries

Abstract

The value of assessing health-related quality of life (HRQoL) in traumatic brain injury (TBI) patients has been increasingly recognized in recent years. Yet, research examining generic and TBI-specific quality of life (QOL) methodologies within this population has been limited, rendering decisions to use one alternative over another difficult and based largely on conceptual grounds. The current study compared widely used generic QoL/HRQOL measures (Satisfaction With Life Scale, 36-item Short Form Survey) and newer population-specific HRQoL measures (Quality of Life after Brain Injury [QOLIBRI], Trauma-Quality of Life [TQoL]) among 77 TBI and 23 orthopedically injured trauma control patients. The QOLIBRI Cognition and Physical Problems subscales were the only HRQoL scores across the four instruments administered that differentiated between patient groups: participants with TBI reported being significantly less satisfied with their cognitive abilities and more bothered by physical problems. Analyses of the unique population-specific QOLIBRI content revealed that 12.2-31.5% of TBI patients endorsed dissatisfaction and 28.8-51.4% endorsed being bothered by items unique to the QOLIBRI. Endorsement rates for unique TQoL items ranged from 1.4-75.7%. Overall, the QOLIBRI and TQoL appear to capture important information pertinent to patients with TBI and trauma. Inclusion of these disease-specific HRQoL measures is recommended over the use of only generic measures among TBI populations.

Introduction

Traumatic brain injury (TBI) is a leading cause of long-term disability and frequently results in important physical, social, cognitive, and psychological changes following injury.¹ Researchers have traditionally emphasized relatively objective outcomes among TBI survivors, such as physical or cognitive recovery, disability status, or degree of functional independence.² However, it is increasingly recognized that in addition to documenting how patients are functioning, it is also important to understand patients' subjective experiences of the impact of their health conditions on their functioning and overall well-being (i.e., to attend to health-related quality of life [HRQoL]).^3,4 Correlations between functional outcome and HRQoL measures are low to moderate, suggesting that these constructs are relatively independent of one another.^5,6

HRQoL measurement in patients with TBI can be assessed through the use of generic or disease-specific instruments. Generic HRQoL measures are developed for broad populations and can therefore be used across individuals with a variety of diseases. While such measures allow for population comparisons, they may fail to assess factors that are particularly salient for certain health conditions.⁷ This may mask important HRQoL issues in the TBI literature, where generic QoL and HRQoL measures have predominated. In particular, the Satisfaction With Life Scale (SWLS),⁸ a generic measure of life satisfaction, and the 36-item Short Form Survey (SF-36),⁹ which emphasizes general health and functioning, have been the most widely recommended and used QoL/HRQoL measures in TBI populations.^10,11 However, these measures may insufficiently capture TBI-pertinent information, such as cognitive and social changes.^12-14

Disease-specific measures are designed with consideration of a particular health condition and, therefore, may be more sensitive to the problems within that population. Relevant to patients with TBI, QoL measures have been developed to target the TBI population specifically and the trauma population more broadly. The Quality of Life after Brain Injury (QOLIBRI) instrument is one such novel HRQoL measure designed to cover domains affected by TBI.¹⁵ It has been validated in large international TBI populations and demonstrates good psychometric properties.^5,6,16–20 Additionally, the QOLIBRI is appropriate to use with cognitively impaired populations,^15,21 demonstrates sensitivity to changes in Glasgow Outcome Scale-Extended (GOSE) scores over time,^18,21 and is believed to be as easy to understand as other HRQoL measures, an important feature given that TBI can adversely impact cognitive functioning.^22,23 The QOLIBRI shows promise for adding unique value over the SF-36 given that it produces fewer floor and ceiling effects as compared with the SF-36.^18,24,25 The QOLIBRI can also differentiate between TBI and non-TBI trauma populations,²¹ although its discriminatory power has not been directly compared with other QoL measures.

Given that TBI often occurs in the context of broader emotional trauma and other physical injuries, HRQoL measures intended for the general trauma population may be valuable to consider for patients with TBI. In particular, the recently developed Trauma-Quality of Life (TQoL) scale has demonstrated promising validity and reliability²⁶ and could provide important information about a broad array of injury-related concerns relevant to trauma patients. Although trauma-specific HRQoL measures have not been widely used, there has been a call for further evaluation of these instruments to examine whether utilization would enhance outcome research.^27,28

Overall, research directly comparing the QOLIBRI and TQoL to other generic QOL/HRQoL methodologies within this population has been limited, rendering decisions to use one alternative over another difficult and based largely on conceptual grounds.¹⁰ Additionally, there are still major gaps in our knowledge regarding the impact of TBI on HRQoL.^10,29 Thus, the aim of the current study was to directly compare widely-used generic QoL/HRQOL measures (SWLS, SF-36) and newer population-specific HRQoL measures (QOLIBRI, TQoL) to establish their degree of overlap and sensitivity to TBI-specific problems. We hypothesized that: 1) as a TBI-specific HRQoL instrument, patients with TBI would report more HRQoL issues on subscales of the QOLIBRI (e.g., cognitive functioning) than orthopedically-injured trauma controls (TC), and 2) content unique to the TQoL and QOLIBRI (vs. SWLS/SF-36) would be commonly endorsed in TBI and TC groups, indicating relevant content necessitating population-specific HRQoL scales. We also performed exploratory analyses to characterize the degree of association between HRQoL measures (which we expected to be moderate-to-strong overall) as well as the degree of association between HRQoL ratings and other clinical outcomes (e.g., emotional, TBI-related, and physical symptoms; acquiescence response bias; and degree of injury-related functional limitations).

Methods

Participants

Patients admitted to the emergency department (ED) or inpatient trauma service at a Level 1 trauma center in Milwaukee, WI between May 2017 and August 2018 were identified through the Epic medical record system. Screening and recruitment flow charts are available in the online Supplementary Material S1. Patients who had been involved in a potential TBI-related event based on cause of injury (i.e., motor vehicle-traffic crash, fall, etc.) were screened for study inclusion in the TBI group. Participants in the TBI group also were required to meet the study definition of TBI (see below) and to be enrolled within 2 weeks of injury. Of the 677 patients who appeared eligible based on record review, 162 remained eligible following additional screening and consented to participate. A smaller group of TCs also were enrolled between July 2017 and April 2018. Prospective TC participants were contacted before 3 months post-injury and were selected to broadly match the TBI group in demographics and level of care. Of the 130 TC patients who appeared eligible, 28 patients remained eligible and chose to enroll in the study. Inclusion criteria for all participants were 18+ years old, English speaking, and able to give informed consent. Prisoners and participants unable to consent (e.g., activated power of attorney) were excluded. This study was approved by the institutional review board at the principal investigator's institution.

Of the consenting patients, 101 returned for follow-up at 3 months post-injury. One subject did not complete any HRQoL measures and was thus excluded, resulting in a final sample of 100 participants (77 TBI and 23 TC participants). A comparison of enrolled participants who did versus those who did not return for 3-month follow-up revealed no significant difference in gender, race, initial Glasgow Coma Scale (GCS) score, or the proportion of participants who experienced a work-related injury or who were involved in litigation related to their injury. There was a small but statistically significant difference in age (t(188) = 2.11, p = 0.036), such that participants who returned were older than those who did not (mean [M] difference = 4.92 years). Of note, 99 of the 100 participants in the current study sample were included in another recent study focused on different outcome measures.³⁰

Definition of TBI

The definition of TBI was adapted from the American Congress of Rehabilitation Medicine's definition of mild TBI (mTBI)³¹—a traumatically induced physiological disruption of brain function, as manifested by at least one of the following: 1) any period of loss of loss of consciousness (LOC); 2) any loss of memory for events immediately before or after the injury; 3) any alteration in mental state at the time of the accident (e.g., feeling dazed, disoriented, or confused); or 4) focal neurologic deficits that may or may not be transient. Participants could have sustained TBIs of any severity (i.e., we did not apply exclusions based on GCS scores or duration of LOC/post-traumatic amnesia). However, given that the majority of enrollment occurred in the ED and given the typical characteristics of the TBI population, a majority of the sample would be classified as having mTBI given historical TBI severity grading systems. Specifically, of the 77 TBI participants, 56 met criteria for an uncomplicated mild TBI, 18 had a complicated mild TBI, and three had a severe TBI based on GCS scores and computed tomography findings. The TBI group had a prevalence of witnessed loss of consciousness of 27.3% (88.9% ≤ 30 min duration, 11.1% 60–70 min), and 62.3% endorsed posttraumatic amnesia (82.1 ≤ 24 h duration, 15.4% 1–7 days, 2.5% ≥ 8 days).

Study protocol

Participants completed an in-person clinical outcome assessment (roughly 90-120 min in duration) at about 3 months post-injury (M = 92 days, standard deviation [SD] = 5, range = 85-111). The assessment measures administered were selected to cover the major domains of patient-reported outcomes: symptoms, day-to-day functioning, and quality of life. Relevant measures from the assessment battery that are included in analyses for this study are described below.

Quality of life measures

To obtain information within the domain of quality of life, we administered the SWLS,⁸ SF-36 (version 1),⁹ TQoL,²⁶ and QOLIBRI.¹⁵ The SWLS is a 5-item measure which provides a single general life satisfaction score (range 5-35). The SF-36, a measure of generic health status, provides subscale scores for eight domains: Physical Functioning, Role-Physical, Role-Emotional, Pain, General Health, Vitality, Social Functioning, and Mental Health (range 0-100). Subscale scores are also combined to create two aggregate T scores, the Physical Component Summary (PCS) and Mental Component Summary (MCS) scores (M = 50, SD = 10). These component summary scores were computed using a standardized scoring procedure in which domain scores are first transformed into linear z-scores based on US population norms prior to calculating PCS and MCS T scores.³² The TQoL is a 43-item measure assessing trauma-specific HRQoL and provides summary scores in five domains: Emotional Well-Being, Functional Engagement, Recovery/Resilience, Peri-Traumatic Experience, and Physical Well-Being. The QOLIBRI contains 37-items and consists of four subscales assessing satisfaction with one's Cognition, Self, Daily Life and Autonomy, and Social Relationships, and two subscales assessing how bothered one is by Emotions and Physical Problems (range 0-100). Responses on the Emotions and Physical Problems items were reverse coded so all QOLIBRI scales are keyed in the same direction (i.e., higher scores reflect better QoL). Finally, TC participants were verbally instructed to consider their “injury” for any QOLIBRI items asking about “brain injury.”

Table 1 provides an overview of the content evaluated through the four HRQoL instruments used in the current study. The SWLS provides a single generic score of overall life satisfaction. The SF-36, TQoL, and QOLIBRI all assess psychiatric and physical functioning, as well as level of or satisfaction with autonomy/independence. The TQoL is unique in assessing a patient's peri-traumatic experience and level of resiliency, whereas cognitive factors and the patient's self-perception are assessed by only the QOLIBRI.

Table 1.

Summary of Quality of Life Scales Evaluated and their Content

Content domain represented	SWLS	SF-36	TQOL	QOLIBRI
Life satisfaction	+
Psychiatric		+	+	+
Physical		+	+	+
Autonomy/independence		+	+	+
Social		+		+
Peri-traumatic experience			+
Recovery/resilience			+
Cognitive				+
Self-perception/satisfaction				+

SWLS, Satisfaction With Life Scale; SF-36, 36-item Short Form Survey; TQOL, Trauma Quality of Life Scale; QOLIBRI, Quality of Life After Brain Injury

Post-injury symptoms

The Rivermead Post Concussion Symptoms Questionnaire (RPQ) ³³ was used to assess the severity of post-injury symptoms. Participants rate the degree to which they have experienced, in the last 24 h, 16 symptoms more so than at pre-injury baseline. Ratings of 0 (“not experienced at all”), 2 (“mild problem”), 3 (“moderate problem”), and 4 (“severe problem”) are summed to achieve a total symptom severity score (range 0–64)

Day-to-day functioning

To assess the domain of overall functioning, we administered the GOSE, ³⁴ an interview-based index of the impact of one's injury on day-to-day functioning (e.g., work, home independence, travel/shopping, social/interpersonal functioning; range 1-8 where 1 = Death, 8 = Upper Good Recovery)

Other measures

The Minnesota Multiphasic Personality Inventory-2 Restructured Form (MMPI-2-RF) ³⁵ was administered to collect a measure of acquiescence response bias (TRIN-r scale). We used the following measures to assess for general distress/psychiatric symptoms: Brief Symptom Inventory-18 (BSI-18), ³⁶ Post-traumatic Stress Disorder (PTSD) Checklist for DSM-5 (PCL-5),³⁷ Personality Inventory for DSM-5 Brief Form (PID-5-BF),³⁸ Depression, Anxiety, and Stress Scale (DASS-21), ³⁹ and the MMPI-2-RF Somatic Complaints (RC1) scale. A word reading test (WRAT-4)⁴⁰ was administered at the beginning of each visit to establish that participants had sufficient reading ability to independently complete the questionnaires. If performance on the WRAT-4 suggested less than a 6th grade reading level, or if requested, the questionnaires were read to the participant by research staff

Statistical analysis

Descriptive statistics including skewness and kurtosis of continuous variables and cell counts of categorical variables were first evaluated to develop the analysis plan. Independent samples t-tests and Fisher's exact tests were used to compare the TBI and TC groups on demographic, history, and injury variables. Among the primary QoL-related variables and criterion clinical assessment variables only the PCL-5, and TQoL Functional Engagement fell outside the recommended bounds for skewness and kurtosis (± 2) to employ parametric statistics.⁴¹ Because of our desire to leverage well-known effect size measures (Cohen's d, r) and because parametric statistics are reasonably robust to violations of distributional assumptions ⁴² we present the primary analyses of the QoL measures as parametric statistics but also performed sensitivity analyses using nonparametric alternatives for non-normal measures to ensure that this decision did not bias the results. In particular, descriptive statistics for the four HRQoL measures were computed as means and standard deviations. Independent samples t-tests were used to examine group differences on the four HRQoL instruments (SWLS, SF-36, TQoL, and QOLIBRI). Cohen's d effect sizes were computed to facilitate interpretation of group differences. Finally, Pearson's r correlations were calculated to examine the linear relationships between the scales assessed by the HRQoL instruments, as well as between QOL scales and other outcome measures of interest. Although effect sizes (r, d) were of primary interest rather than statistical significance, we adopted α = 0.01 to accommodate the large number of statistical comparisons.

To identify item content on the QOLIBRI and TQoL that was not present on the SF-36, items were coded by two independent raters for their degree of uniqueness or overlap with the SF-36. Discrepancies were examined by a third rater in order to reach a consensus. TQoL items assessing positive sequela were reverse coded per scoring instructions such that all items were scaled in the same direction (with endorsement reflecting more problems). QOLIBRI “satisfied” items were dichotomized such that responses of “not at all” and “slightly” satisfied were considered indicative of dissatisfaction, whereas “moderately,” “quite,” and “very” satisfied were considered indicative of satisfaction. Likewise, responses of “moderately,” “quite,” and “very” bothered were grouped for “bothered” items, whereas “not at all” and “slightly” bothered were labeled “not bothered.” For the TQoL, responses were dichotomized with “strongly disagree” and “moderately disagree” classified as disagreement and “strongly agree” and “moderately agree” classified as agreement. Prevalence of endorsement of unique items from the QOLIBRI and TQoL were computed as percentages (95% confidence intervals) in the TBI group. Data analysis was carried out using IBM SPSS Statistics Version 20.0 (IBM Corp., Armonk, NY).

Results

The socio-demographic, injury, and clinical characteristics of the TBI and TC groups are reported in Table 2. The TBI and TC groups were well matched in demographic (age, gender, race), history (education, estimated pre-morbid IQ), and injury variables (cause of injury, highest level of care, work-related injury, litigation). The groups also were well-matched in 3-month general emotional distress (BSI-18 Global Severity Index [GSI]), somatic complaints, personality scores, and PTSD symptoms. As expected, the TBI group reported more post-injury symptoms (RPQ total score) at 3 months than the TC group (t[72.08] = 3.1, p = .003; d = 0.62). Abbreviated Injury Scale (AIS)⁴³ Head/Neck Injury scores (available for admitted patients) were significantly higher for the TBI group (t[33.00] = 6.08, p < 0.001). However, peripheral injury severity estimated from a Peripheral Injury Severity Score ⁴⁴ (ISS; i.e., ISS calculated without the AIS Head/Neck or Face regions) was no different between groups (t[45] = 0.73, p = 0.469).

Table 2.

Sample Characteristics

	TBI (n = 77)	Trauma control (n = 23)
	n (%) or M (SD)	n (%) or M (SD)	p
Demographics and history
Age (years)	45.0 (15.8)	45.3 (18.5)	0.924
Male gender	44 (57.1%)	13 (56.5%)	0.572
Race			0.868
White	40 (51.9%)	15 (65.2%)
Black	31 (40.3%)	7 (30.4%)
Unknown	4 (5.2%)	1 (4.3%)
Not reported	2 ( 2.6%)	0 (0%)
Years of education	13.0 (2.1)	13.9 (2.8)	0.165
WRAT-4 Word Reading SS	94.7 (15.6)	97.7 (13.2)	0.392
Injury characteristics
Highest level of care			0.121
Emergency department	43 (55.8%)	9 (39.1%)
Inpatient unit	34 (44.2%)	14 (60.9%)
Cause of injury			0.361
Motor vehicle crash	49 (63.6%)	14 (60.9%)
Fall	18 (23.4%)	8 (34.8%)
Assault	5 (6.5%)	0 (0%)
Struck by/against object	4 (5.2%)	0 (0%)
Other	1 (1.3%)	1 (4.3%)
TBI severity
Uncomplicated mild (GCS 13-15, CT-)	56 (72.7%)	-
Complicated mild (GCS 13-15, CT+)	18 (23.4%)	-
Severe (GCS 3-8)	3 (3.9%)	-
AIS Head/Neck Score (TBI n = 61; TC n = 14)	1.8 (1.7)	0.0 (0.0)	< 0.001
ISS Peripheral Injury Score (TBI n = 61; TC n = 14)	11.5 (7.8)	9.8 (4.9)	0.469
Litigation related to injury	23 (29.9%)	7 (30.4%)	0.575
Work-related injury	6 (7.8%)	3 (13.0%)	0.341
Clinical assessment variables (3 months)
RPQ total	14.9 (16.3)	6.9 (8.5)	0.003
PCL-5 total	17.2 (19.0)	10.5 (14.8)	0.144
BSI-18 GSI standard score	105.1 (18.0)	104.3 (17.7)	0.846
MMPI-2-RF RC1	63.4 (14.8)	58.4 (12.4)	0.145
PID-5-BF total	0.51 (0.56)	0.48 (0.51)	0.821

TBI, traumatic brain injury; M, mean; SD, standard deviation; WRAT-4, Wide Range Achievement Test 4; SS, standard score; GCS, Glasgow Coma Scale; AIS, Abbreviated Injury Scale; CT, computed tomography; ISS, Injury Severity Score; TC, trauma control; RPQ, Rivermead Post Concussion Symptom Questionnaire; PCL-5, Post-Traumatic Stress Disorder Checklist for Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; BSI-18, Brief Symptom Inventory-18; GSI, Global Severity Index; MMPI-2-RF, Minnesota Multidimensional Personality Inventory-2 Restructured Form; RC1, Somatic Complaints scale; PID-5-BF, Personality Inventory for DSM-5 Brief Form.

The subscale and total scores for the four HRQoL instruments for both TBI and TC groups are presented in Table 3. The two groups were not statistically significantly different in any of the SWLS, SF-36, or TQoL scores. In contrast, participants with TBI reported being less satisfied with cognitive performance (QOLIBRI Cognition subscale; d = 0.48) and more bothered by physical problems (QOLIBRI Physical Problems scale; d = 0.58). However, no significant differences were found for the QOLIBRI total score or remaining subscales. Interestingly, both groups obtained PCS mean scores roughly a full standard deviation below the mean, indicating higher degrees of physical disability.

Table 3.

TBI versus Trauma Control Group Differences

	TBI	Trauma control
	M (SD)	M (SD)	d	p
SWLS1	21.3 (7.6)	19.8 (8.8)	0.20	0.408
SF-362
PCS	39.2 (12.0)	40.2 (11.4)	-0.09	0.728
MCS	46.8 (11.3)	47.1 (12.2)	-0.03	0.906
General Health	63.6 (20.4)	66.6 (19.1)	-0.15	0.537
Mental Health	70.3 (20.0)	68.0 (22.2)	0.11	0.646
Pain	55.1 (29.6)	57.2 (31.7)	-0.07	0.769
Physical Functioning	58.8 (30.7)	58.2 (30.2)	0.02	0.930
Role—Physical	43.5 (42.9)	45.5 (45.4)	-0.05	0.853
Role—Emotional	60.7 (43.5)	62.1 (40.2)	-0.03	0.894
Social Functioning	60.1 (28.8)	63.1 (30.2)	-0.10	0.676
Vitality	49.4 (24.3)	54.1 (21.7)	-0.20	0.417
TQOL3
Emotional Well-Being	47.7 (11.8)	52.3 (10.1)	-0.37	0.101
Functional Engagement	29.0 (4.7)	28.8 (6.0)	0.04	0.885
Peri-Traumatic Experience	14.1 (3.0)	14.8 (2.1)	-0.25	0.291
Physical Well-Being	20.3 (7.0)	20.8 (6.3)	-0.08	0.753
Recovery/Resilience	15.7 (3.5)	16.4 (3.4)	-0.21	0.377
QOLIBRI4
Total	64.5 (24.0)	69.1 (18.8)	0.20	0.400
Cognition	65.3 (27.4)	77.5 (19.0)	0.48	0.020
Self	60.6 (26.8)	57.9 (26.8)	0.10	0.666
Daily Life and Autonomy	65.4 (28.0)	64.1 (28.5)	0.04	0.852
Social Relationships	67.6 (28.1)	69.4 (23.6)	0.07	0.780
Emotions	65.1 (29.5)	69.6 (27.2)	0.16	0.518
Physical Problems	63.2 (29.4)	79.2 (20.6)	0.58	0.005

Bolded where p < .05.

¹Higher scores indicate greater overall life satisfaction.

²Higher scores indicate greater health-related quality of life (HRQoL) and lower levels of disability.

³Higher score indicate greater HRQoL and fewer problems in these domains.

⁴Higher total score indicates greater HRQoL; higher Cognition, Daily Life and Autonomy, Self, and Social Relationships scores indicate greater satisfaction in these domains; higher Emotions and Physical Problems scores indicate less issues and/or frustration in these domains.

TBI, traumatic brain injury; M, mean; SD, standard deviation; SWLS, Satisfaction With Life Scale; SF-36, 36-item Short Form Survey; PCS, Physical Component Summary; MCS, Mental Component Summary; TQOL, Trauma Quality of Life Scale; QOLIBRI, Quality of Life After Brain Injury.

Associations between the four HRQoL scales are shown in Table 4. Consistent with expectations, the strongest correlations were generally found between subscales with similar content. Regarding physical functioning, the SF-36 PCS, TQoL Physical Well-Being, and QOLIBRI Physical Problems subscales demonstrated strong correlations. The SF-36 MCS demonstrated high correlations with the TQoL Emotional Well-Being and the QOLIBRI Self and Emotions subscales. Similarly, the TQoL Functional Engagement and QOLIBRI Daily Life and Autonomy (DLA) subscales, which both assess independence/functioning, showed a moderate relationship. Regarding unique associations, the TQoL Functional Engagement subscale demonstrated weak associations with the SWLS, SF-36 MCS, TQoL Peri-Traumatic Experiences, and QOLIBRI Self scale. Likewise, the TQoL Peri-Traumatic Experiences scale showed weak associations with the SF-36 PCS, and TQoL Resilience and Physical Well-Being scales. Finally, the SF-36 PCS and MCS were not significantly correlated. Of the TQoL subscales, Emotional Well-Being and Physical Well-Being demonstrated the highest correlations with subscales from other measures, with ranges between 0.51 to 0.82 and 0.42 to 0.81. Aside from the weak relationships mentioned above, correlations for the QOLIBRI with other HRQoL measure subscales ranged from 0.20 to 0.81.

Table 4.

Associations (Correlations) between HRQoL Measures

	a	b	c	d	e	f	g	h	i	j	k	l	m	n
a. SWLS
b. SF-36 PCS	0.34
c. SF-36 MCS	0.50	0.16
d. TQOL EW	0.51	0.61	0.58
e. TQOL FE	0.13	0.54	0.23	0.52
f. TQOL PE	0.35	.22	0.30	0.40	0.21
g. TQOL PW	0.48	0.81	0.42	0.75	0.48	0.29
h. TQOL R	0.38	0.45	0.39	0.55	0.41	0.15	0.58
i. QOLIBRI TOT	0.68	0.59	0.69	0.82	0.44	0.40	0.71	0.53
j. QOLIBRI COG	0.52	0.45	0.49	0.65	0.30	0.34	0.54	0.41	0.83
k. QOLIBRI Self	0.72	0.41	0.68	0.63	0.20	0.34	0.58	0.46	0.88	0.67
l. QOLIBRI DLA	0.62	0.60	0.61	0.74	0.49	0.30	0.69	0.56	0.90	0.70	0.80
m. QOLIBRI SR	0.57	0.40	0.56	0.69	0.38	0.34	0.51	0.37	0.86	0.70	0.74	0.73
n. QOLIBRI EM	0.46	0.46	0.62	0.66	0.39	0.37	0.54	0.34	0.74	0.44	0.57	0.55	0.56
o. QOLIBRI PP	0.48	0.65	0.49	0.75	0.46	0.33	0.70	0.46	0.77	0.56	0.52	0.64	0.53	0.67

Bolded where p < 0.01.

HRQOL, health-related quality of life; SWLS, Satisfaction With Life Scale; SF-36 PCS, SF-36, 36-item Short Form Survey; PCS, Physical Component Summary; MCS: Mental Component Summary TQOL, Trauma Quality of Life Scale; EW, Emotional Well-Being; FE, Functional Engagement; PE, Peri-Traumatic Experience; PW, Physical Well-Being; R, Recovery/Resilience; QOLIBRI, Quality of Life After Brain Injury; TOT, Total; COG, Cognition; DLA, Daily Life and Autonomy; SR, Social Relationships; EM, Emotions; PP, Physical Problems.

Table 5 shows the correlations between the four HRQoL instruments and the PID-5-BF, BSI-18 Global Severity Index (GSI), DASS-21 Total Score, PCL-5, RPQ, MMPI-2-RF RC1 and TRIN-r scales, and GOSE overall score. Reporting higher levels of maladaptive personality (PID-5-BF) was associated with lower QoL on most QoL scales, with the exception of TQoL Resiliency, Physical Well-being, and Functional Engagement, as well as the SF-36 PCS (which were nonsignificantly related to maladaptive personality). Elevated scores on all HRQoL subscales indicate high quality of life and showed significant negative correlations with measures of general emotional distress, PTSD and post-injury symptoms, and somatic complaints (r = -0.28 to -0.82). Higher levels of acquiescence (MMPI-2-RF TRIN-r raw scores) was associated with greater physical difficulties (SF-36 PCS, TQoL Physical Well-Being, and QOLIBRI Physical Problems) and less satisfaction in one's daily autonomy (QOLIBRI DLA). With the exception of the TQoL Peri-Traumatic Experience scale, better functional outcome (GOSE scores) was associated with significantly greater QoL/HRQoL on all other administered scales, with the strongest relationships for HRQoL subscales associated with physical functioning.

Table 5.

Correlations between Quality of Life Measures and Other Outcomes

	Maladaptive personality	General emotional distress		PTSD symptoms	mTBI symptoms	Somatic complaints	Acquiescence response bias	Functional Independence
	PID-5-BF	BSI-18 GSI	DASS total	PCL-5	RPQ	MMPI-2-RF RC1	MMPI-2-RF TRIN-r Raw	GOSE
SWLS	-0.41	-0.52	-0.50	-0.39	-0.41	-0.34	-0.18	0.40
SF-36
PCS	-0.14	-0.58	-0.34	-0.31	-0.46	-0.51	-0.31	0.71
MCS	-0.41	-0.68	-0.61	-0.57	-0.58	-0.53	-0.10	0.40
TQOL
EW	-0.49	-0.72	-0.67	-0.73	-0.75	-0.64	-0.26	0.62
FE	0.03	-0.40	-0.28	-0.36	-0.39	-0.35	-0.16	0.55
PE	-0.38	-0.35	-0.39	-0.39	-0.44	-0.40	-0.25	0.11
PW	-0.23	-0.67	-0.49	-0.50	-0.61	-0.64	-0.32	0.66
R	-0.09	-0.44	-0.34	-0.31	-0.51	-0.43	-0.22	0.44
QOLIBRI
Total	-0.49	-0.82	-0.76	-0.65	-0.73	-0.65	-0.31	0.61
COG	-0.38	-0.58	-0.56	-0.48	-0.65	-0.60	-0.26	0.43
Self	-0.43	-0.71	-0.70	-0.47	-0.58	-0.48	-0.22	0.44
DLA	-0.29	-0.69	-0.60	-0.52	-0.57	-0.52	-0.29	0.62
SR	-0.48	-0.66	-0.65	-0.48	-0.54	-0.49	-0.25	0.49
EM	-0.53	-0.77	-0.70	-0.71	-0.63	-0.56	-0.25	0.56
PP	-0.39	-0.73	-0.61	-0.68	-0.74	-0.61	-0.28	0.67

Bolded where p < 0.01.

Pearson correlations for all measures except GOSE (which are Spearman correlations).

PTSD, post-traumatic stress disorder; mTBI, mild traumatic brain injury; PID-5-BF, Personality Inventory for DSM-5 Brief Form; BSI-18, Brief Symptom Inventory-18; GSI, Global Severity Index; DASS, Depression, Anxiety, and Stress Scale; PCL-5, PTSD Checklist for DSM-V; RPQ, Rivermead Post Concussion Symptom Questionnaire; MMPI-2-RF, Minnesota Multidimensional Personality Inventory-2 Restructured Form; RC1, Somatic Complaints scale; TRIN-r, True Response Inconsistency scale; GOSE, Glasgow Outcome Scale-Extended; SWLS, Satisfaction With Life Scale; SF-36, 36-item Short Form Survey; PCS, Physical Component Summary; MCS, Mental Component Summary; TQOL, Trauma Quality of Life Scale; EW, Emotional Well-Being; FE, Functional Engagement; PE, Peri-Traumatic Experience; PW, Physical Well-Being; R, Recovery/Resilience; QOLIBRI, Quality of Life After Brain Injury; COG, Cognition; DLA, Daily Life and Autonomy; SR, Social Relationships; EM, Emotions; PP, Physical Problems.

Endorsement rates for unique population-specific QOLIBRI content ranged from 12.2-31.5% dissatisfaction on the Cognition, DLA, Self, and Social Relationships items and 28.8-51.4% reported being bothered on Emotions and Physical Problems items (Table 6). Of these items, participants were most likely to endorse emotional, physical, and sexual difficulties. Endorsement rates for unique TQoL items ranged from 1.4-75.7% (Table 7). Participants reported the most difficulties with sleep disruption, increased negativity about the future, feeling as though their life is worse than before their injury, and feeling as though their healing has not improved as expected. A high percentage of participants also reported experiencing fear during the injury and current difficulties emotionally coping with the injury. Finally, of the administered QoL measures, the QOLIBRI is the only one to assess cognitive factors and patients' self-perception. On these scales, the degree of patients with TBI reporting dissatisfaction ranged from 12.2-29.7% across all Cognition items and from 17.8-33.8% across the Self items. The TQoL is unique in assessing a patient's peri-traumatic experience and level of resiliency. On these scales, 12.2-75.7% of patients with TBI reported problems on Resiliency items and 8.1-66.2% reported problems on Peri-Traumatic Experience items.

Table 6.

Frequency of QOLIBRI Unique Item Endorsement in TBI Sample

Dissatisfied with:	% (95% CI)
A1. Ability to concentrate	27.0 (17.0-37.0)
A2. Ability to express self and comprehend conversations	17.6 (9.0-26.2)
A3. Memory for everyday things	29.7 (19.4-40.0)
A4. Ability to plan and problem-solve every day problems	21.6 (12.3-30.9)
A5. Ability to make decisions	17.6 (9.0-26.2)
A6. Ability to navigate around	12.2 (4.8-19.7)
A7. Cognitive processing speed	29.7 (19.4-40.0)
B2. Motivation levels	29.7 (19.4-40.0)
B3. Self-esteem	25.7 (15.9-35.5)
B4. Physical appearance	23.3 (13.7-32.9)
B6. Self-perception	17.8 (9.1-26.5)
B7. Perception of the future	18.9 (10.1-27.8)
C1. Level of independence	16.4 (8.0-24.8)
C2. Ability to leave one's home and get around in public	21.6 (12.3-30.9)
C3. Ability to complete household activities	23.0 (13.5-32.5)
C4. Ability to manage finances	14.9 (6.9-23.0)
C7. Control over one's own life	18.9 (10.1-27.8)
D1. Ability to feel affection	16.2 (7.9-24.5)
D2. Relationships with family	14.9 (6.9-23.0)
D3. Relationships with friends	16.2 (7.9-24.5)
D4. Relationship with a significant other or lack of relationship	24.3 (14.6-34.0)
D5. Sex life	31.5 (21.0-42.1)
D6. Others' opinions towards you	21.9 (12.5-31.3)
Bothered by:
E1. Loneliness	35.6 (24.8-46.5)
E2. Boredom	46.6 (35.3-57.9)
E5. Anger or aggression	34.2 (23.5-45.0)
F1. Physical slowness and/or clumsiness	43.8 (32.6-55.1)
F2. Other injuries sustained during the TBI	51.4 (40.0-62.8)
F4. Visual or hearing difficulties	28.8 (18.5-39.1)
F5. The impact of the TBI	43.8 (32.6-55.1)

QOLIBRI, Quality of Life After Brain Injury; CI, confidence interval; TBI, traumatic brain injury.

Table 7.

Frequency of TQoL Unique Item Endorsement in TBI Sample

	% severe or worse (95% CI)
A. degree of injury severity.	49.3 (38.0-60.7)

	% endorsing a problem (95% CI)
B2. Physical recovery has occurred as expected.	17.6 (9.0-26.2)
B3. Needing assistance with driving.	28.4 (18.3-38.6)
B8. Needing assistance with eating.	2.7 (0-6.5)
B9. Needing assistance with using the bathroom.	1.4 (0-4.3)
B10. Needing assistance with cooking/meal preparation.	13.5 (5.8-21.3)
B14. Daily use of pain medications.	37.8 (26.9-48.8)
C16. Sleep disruption	58.1 (47.0-69.2)
C19. Being bothered by injury scars.	25.7 (15.9-35.5)
C20. Appetite changes	29.7 (19.4-40.0)
C21. Changes in eating habits due to subsequent physical problems.	20.3 (11.2-29.4)
C22. Feelings of fear when thinking about the injury.	40.5 (29.5-51.6)
C24. Increased irritation following the injury.	40.5 (29.5-51.6)
C25. Anger about the injury.	47.3 (36.1-58.5)
D26. Feeling that the medical care received was good.	8.1 (1.9-14.4)
D27. Feeling that the care from physicians and nurses was good.	8.1 (1.9-14.4)
D28. Healing has occurred without problems.	27.4 (17.3-37.5)
D29. Being able to make changes to compensate for existing limitations.	12.2 (4.8-19.7)
E30. Being able to remember the injury.	29.7 (19.4-40.0)
E31. Feeling fear during the injury.	66.2 (55.6-76.9)
E32. Feeling strong emotions during the injury.	62.2 (51.3-73.2)
E33. Currently feeling safe on a daily basis.	21.6 (12.3-30.9)
F34. Difficulty dealing with the injury emotionally.	58.1 (47.0-69.2)
F35. Having increased positivity about the future than before the injury.	59.5 (48.5-70.1)
F36. Feeling that life is currently better than it was before the injury.	75.7 (66.0-85.4)
F37. Being able to handle physical limitations stemming from the injury.	28.4 (18.3-38.6)
F38. Being able to handle mood changes following the injury.	32.4 (21.9-43.0)
F39. Having increased negativity about the future than before the injury.	27.0 (17.0-37.0)
F40. Feeling that one's recovery has occurred faster than expected.	56.2 (45.0-67.5)
F41. Having to rely on other individuals and programs due to financial limitations.	39.2 (28.3-50.2)
F43. Feeling that one's injuries have negatively impacted others emotionally.	33.8 (23.2-44.5)

CI, confidence interval.

Discussion

The current study investigated the use of newer population-specific HRQoL measures (QOLIBRI, TQoL) and commonly used generic measures (SF-36, SWLS) within TBI and orthopedic trauma populations. It is unique in its inclusion of multiple HRQoL measures within a single study design and adds to the literature by examining the use of the QOLIBRI within a trauma cohort without head injury. Additionally, a limited number of studies have undertaken QOLIBRI comparative analyses with mTBI populations. Therefore, the current study also holds value through the examination of HRQoL within a predominantly mTBI cohort.

First, we hypothesized that QOLIBRI scores would be lower for TBI versus TC patients. Findings revealed the QOLIBRI Cognition and Physical Problems subscales were the only HRQoL scores across the four instruments administered that differentiated between patient groups, with participants with TBI being significantly less satisfied with their cognitive abilities and more bothered by physical problems. This finding supports the notion that the QOLIBRI evaluates content particularly relevant to patients with TBI. This is consistent with prior investigations demonstrating that the Cognition subscale differentiates between patients with TBI and those with soft-tissue injuries,²¹ as well as between patients with major TBI and those with no or mild TBI.⁴⁵ This is noteworthy because other HRQoL measures (e.g., SF-36 and SWLS) and TBI outcome measures (e.g., GOSE) fail to assess cognition.^46,47 Additionally, that the two groups were well matched in peripheral injury severity implies that the more substantial physical problems reported by patients with TBI may have been driven by brain injury. Thus, the current findings add to a growing body of literature suggesting that the SF-36 may not capture many TBI-relevant symptoms, such as cognitive deficits, emotional problems, and significant physical changes.^12-14,45

Previous studies have suggested that QOLIBRI scores of <60 may indicate impaired HRQOL.^22,48 However, studies examining the interpretation of QOLIBRI scores have been limited, and a more complete understanding of the meaning of scores is still needed.⁴⁸ Interestingly, investigations of the QOLIBRI in TBI populations at 1-year follow-up have found average total scores of around 65,^5,15,22 whereas a 4-year follow-up study demonstrated an average of 70.⁴⁹ The average QOLIBRI total score within the current TBI sample was 64.5 at 3-month follow-up, suggesting that while HRQoL may be negatively impacted following a TBI, QOLIBRI scores may improve modestly in the chronic recovery period. Because QOLIBRI items primary assess satisfaction with and distress about one's level of functioning, this could reflect the gradual recruitment of adaptive coping strategies after TBI. Time since injury has been shown to be related to generic HRQoL.⁵⁰ Thus, while the QOLIBRI should be investigated with longitudinal research, these data provide preliminary evidence that TBI-specific HRQoL may conform to the same pattern over time.

Correlational analyses between the SF-36 and QOLIBRI displayed the expected pattern of relationships and were consistent with other study findings.^5,17,18,29 Consistent with prior results, the QOLIBRI total score also demonstrated a lower association with the SF-36 PCS scale than the MCS.^5,17,29,51 Regarding measure overlap, the QOLIBRI total score was not redundant with the SWLS or SF-36 scales based upon conventional thresholds (r > 0.7). This diverges from another study that found a substantial correlation between the QOLIBRI and SWLS total scores.²⁹ Within the current sample, the QOLIBRI Self was the only subscale to demonstrate potential overlap with the SWLS. Findings regarding this association have previously been mixed, suggesting that further research may help to elucidate this relationship.^16,29 Although the SWLS has been recommended as a central measure of HRQoL after TBI,⁵² our findings support prior suggestions that the two inventories yield distinct information, which may be intuited by comparing their distinct item content (Table 1).²⁹

Associations between the TQoL and other HRQoL measures also displayed the same anticipated pattern of relationships, which support the construct validity of this measure. While the PCS and TQoL Physical Well-Being scales demonstrate potential redundancy, none of the other TQoL scales show the same degree of overlap. Moreover, the magnitude of correlations, which were generally 0.30-0.60, suggests that while the TQoL shares something in common with these measures, it also offers something conceptually and psychometrically distinct.

Analyses between HRQoL and other outcome measures revealed a number of interesting relationships. As expected, the HRQoL scales demonstrated negative relationships with measures of maladaptive personality traits, emotional distress, PTSD symptoms, concussion symptoms, and somatic complaints. HRQoL were generally correlated more strongly with current emotional distress (e.g., M r with BSI-18 GSI = -0.63) and functional independence (M r = 0.52) than with maladaptive personality (M r = -0.34), suggesting that HRQoL does not predominantly reflect premorbid personality. Our findings of robust associations (r ∼ 0.5) between HRQoL and both post-injury symptoms and PTSD symptoms was consistent with prior studies finding significant relationships between QOLIBRI scores and measures of these constructs (RPQ and PCL-C) in other TBI samples.^45,53 Estimates of acquiescence response bias were mostly nonsignificant, with small and significant associations with measures of physical and functional abilities. With the exception of QOLIBRI total score, TQoL Physical Well-Being, and SF-36 PCS, the isolated association between acquiescence and QOL could be inflated by the fact that items were keyed in the same direction on other scales.

Finally, the strong association between functional outcome and HRQoL, with the strongest associations seen for physical functioning subscales, confirms previously reported patterns.^16,27,51 Interestingly, the TQoL Peri-Traumatic Experiences scale demonstrates a unique lack of association with GOSE scores, suggesting that degree of disability is not related to the patient's subjective experience during injury. Taken together, these findings suggest that HRQoL is robustly associated with general emotional distress, PTSD symptoms, TBI symptoms, and functional limitations but is related to a more limited degree to maladaptive personality and acquiescence.

In line with hypotheses, high endorsement rates of the unique QOLIBRI and TQoL items suggests that these measures may capture important information pertinent to TBI and trauma patients. For example, 75.4% of current participants reported that they feel their life is worse than it was before their injury and 58.1% reported experiencing sleep disruption on the TQoL. Such information is not captured on the other administered HRQoL measures, although it may have important clinical implications. Likewise, a very high percentage of patients reported that their physical recovery has not occurred as expected, which raises the question of whether patients should be provided with a higher degree of psycho-education to help them better understand what to expect. Overall, the TQoL provides unique information regarding psychosocial domains and patient recovery. An additional strength of the TQoL is that questions are more specifically anchored to the patient's injury thus facilitating interpretation of how one's injury has impacted HRQoL. The QOLIBRI, in contrast, primarily solicits ratings of satisfaction with functioning overall, rendering it challenging to ascertain the relationship between injury and HRQoL.

Given that the QOLIBRI and TQoL appear to offer a more fine-grained assessment of TBI-pertinent symptoms, collection of these disease-specific HRQoL measures is recommended in concert with, or instead of, generic measures. Likewise, the TQoL and QOLIBRI both possess subscales covering content not assessed by these other measures. High endorsement rates on these subscales further highlight the unique value of these measures and suggest that the TQoL and QOLIBRI do not serve as substitutes for one another.

Despite several strengths of this study, including prospective study enrollment, the use of multiple HRQoL measures, and the well characterized TBI sample, certain limitations should be noted. First, the current study possessed a relatively small sample size, and therefore comparisons across these instruments in a larger sample that enables use of modern psychometric methods should be made. Additionally, a high proportion of consenting participants were lost to follow-up. However, that the group of participants who returned for 3-month follow-up were similar in demographic and injury characteristics to those who did not return is reassuring. The present study also used the SF-36 version 1, but it is worth noting that there is a more recent version available. Thus, it is unclear to what degree these findings would generalize to the SF-36 version 2. Finally, although the present sample is notable for its diversity in TBI injury characteristics, the majority were participants with a GCS score of 13-15 and negative acute clinical neuroimaging findings (i.e., uncomplicated mild TBIs), which limits generalizability to more severe brain injury populations.

Conclusions

The current study suggests that the QOLIBRI assesses clinically relevant TBI-specific information not captured by other generic HRQoL measures. The TQoL also appears to contain unique content relevant for all trauma populations. Overall, findings support the application of these disease-specific measures within TBI populations.

Funding Information

The study was funded by the Medical College of Wisconsin Center for Advancing Population Science (CAPS) and NIH grant R01NS110856. The REDCap database used for the study was supported by the MCW Clinical and Translational Science Institute (National Institutes of Health grant UL1TR001436).

Author Disclosure Statement

No competing financial interests exist.

Supplementary Material

Supplementary Material S1