Health status and psychological outcomes after trauma: A prospective multicenter cohort study.

INTRODUCTION: Survival after trauma has considerably improved. This warrants research on non-fatal outcome. We aimed to identify characteristics associated with both short and long-term health status (HS) after trauma and to describe the recovery patterns of HS and psychological outcomes during 24 months of follow-up.
METHODS: Hospitalized patients with all types of injuries were included. Data were collected at 1 week 1, 3, 6, 12, and 24 months post-trauma. HS was assessed with the EuroQol-5D-3L (EQ-5D-3L) and the Health Utilities Index Mark 2 and 3 (HUI2/3). For the screening of symptoms of post-traumatic stress, anxiety and depression, the Impact of Event Scale (IES) and the Hospital Anxiety and Depression Scale (HADS) subscale anxiety (HADSA) and subscale depression (HADSD) were used. Recovery patterns of HS and psychological outcomes were examined with linear mixed model analyses.
RESULTS: A total of 4,883 patients participated (median age 68 (Interquartile range 53-80); 50% response rate). The mean (Standard Deviation (SD)) pre-injury EQ-5D-3L score was 0.85 (0.23). One week post-trauma, mean (SD) EQ-5D-3L, HUI2 and HUI3 scores were 0.49 (0.32), 0.61 (0.22) and 0.38 (0.31), respectively. These scores significantly improved to 0.77 (0.26), 0.77 (0.21) and 0.62 (0.35), respectively, at 24 months. Most recovery occurred up until 3 months. At long-term follow-up, patients of higher age, with comorbidities, longer hospital stay, lower extremity fracture and spine injury showed lower HS. The mean (SD) scores of the IES, HADSA and HADSD were respectively 14.80 (15.80), 4.92 (3.98) and 5.00 (4.28), respectively, at 1 week post-trauma and slightly improved over 24 months post-trauma to 10.35 (14.72), 4.31 (3.76) and 3.62 (3.87), respectively. DISCUSSION: HS and psychological symptoms improved over time and most improvements occurred within 3 months post-trauma. The effects of severity and type of injury faded out over time. Patients frequently reported symptoms of post-traumatic stress. TRIAL REGISTRATION: ClinicalTrials.gov identifier: NCT02508675.

Introduction

Trauma poses a large burden on public health [1]. Reduction of trauma-related mortality in high-income countries [2] has resulted in increased numbers of trauma survivors with long-term injury impact [3], including reduced health status (HS) [4]. An improved understanding of the quality of survival of patients is critically important for improving health care quality and in evaluating trauma care. Furthermore, it is important to understand short and long-term recovery patterns of HS in terms of injured patient characteristics and to identify predictors of outcome of seriously injured patients [5, 6].

Establishing recovery patterns in the short and long-term requires longitudinal data [7]. Non-fatal outcomes after trauma can be assessed with an overall measure of HS. HS includes patients’ physical functioning, state of mind and social activities [8]. In general, trauma has a large impact on HS [4, 9–12], but large variations between patients have been observed [11, 12].

In addition, several relevant non-fatal outcomes after trauma may be assessed more specifically. These include psychological outcomes, such as anxiety and depression. Psychological problems are often reported among trauma patients [13-18] and are associated with worse HS [9, 16].

The fact that trauma has an impact on diverse aspects of patient health, illustrates that a multidimensional approach is necessary for a comprehensive understanding of non-fatal outcomes after trauma. This also allows studying the mutual relations between non-fatal outcomes. Using a multidimensional approach to measure outcomes including HS and symptoms of depression, anxiety and post-traumatic stress will result in a comprehensive understanding of non-fatal outcomes after trauma. In addition, to assess prognostic factors for a poor outcome it is important to cover the entire spectrum of the trauma population without exclusion of particular patient groups (e.g. elderly). The number of longitudinal cohort studies that examine multiple non-fatal outcomes in a large sample with a broad inclusion of type and severity of injury is limited [12, 19–23]. Most studies start measuring outcomes at least 3 months after trauma, resulting in little knowledge about the real short-term consequences [11, 24–27].

The overall aim of the Brabant Injury Outcome Surveillance (BIOS), a population based longitudinal study, is to provide more insight into recovery patterns and determinants of non-fatal outcomes after trauma. The aims of this population-based study are 1. to identify characteristics associated with short, mid and long-term HS and 2. to describe the 2 year recovery patterns of HS and psychological outcome for different categories of trauma patients. This information is important for understanding the short and long-term recovery patterns and for best informing provision of trauma care to injured patients with long-term disability.

Methods

Study design and participants

Data were obtained from the BIOS. The BIOS-study is a prospective observational cohort study in which HS and psychological outcomes are assessed in injured patients in the first 24 months after trauma. The methods of the BIOS have been described in detail in a published research protocol (doi: 10.1136/injuryprev-2016-042032) [28].

Recruitment occurred in all ten hospitals of the Noord-Brabant region (the Netherlands) from August 2015 until November 2016. Adults (≥18 years) who visited an emergency department ≤48 hours after trauma and who were admitted to the hospital, were invited to participate. All types of injuries were included regardless of the intent or severity of the injury. Patients who died between hospital discharge and the first week post-trauma, non-Dutch speaking patients, patients with no permanent address or patients with a pathological fracture were excluded. A proxy informant (caregiver or family member) was asked to complete the self-administered questionnaires if the patient was incapable of participating in the BIOS study him- or herself. Proxy informants were invited to enroll in the study 1 month post-trauma. Informal caregivers (e.g. family members) and paid caregivers (e.g. nurses) were allowed to function as proxy informants.

The Brabant Trauma Registry (BTR) compiles pre-hospital and hospital data of all trauma patients admitted after presentation to the ED in the Noord-Brabant region. Quality of the data of the BTR and BIOS was checked on outliers and completeness by a trauma coordinator and researcher respectively. Data of the BTR was checked as part of routine practice while the data of the BIOS was checked randomly by the researchers. Furthermore, data from a sample of the trauma registry was checked manually by a trauma surgeon.

The study was approved by the Medical Ethics Committee Brabant, the Netherlands (project numbers NL50258.028.14 and NW2016-09). Prior to participation, participants signed an informed consent form.

Data collection

Questionnaires were sent at 1 week and 1, 3, 6, 12 and 24 months after trauma. Based on the participants’ preference, follow-up questionnaires were either completed by paper and pencil or digitally. The questionnaires collected data on general patient characteristics (date of birth, gender), self-reported comorbidities (by using a modified version of the Cumulative Illness Rating Scale [29]), self-reported HS (i.e. EuroQol-5D-3L (EQ-5D-3L) [30], the Health Utilities Index (HUI) Mark 2 and Mark 3 [31]) and self-reported psychological functioning (i.e., Hospital Anxiety and Depression Scale (HADS) [32] and the Impact of Event Scale (IES) [33]). Proxy informants did not complete questionnaires regarding psychological outcome.

To increase the response rate, patients who did not complete a questionnaire up until 3 or 6 months post-trauma were asked to complete a short version of the BIOS-questionnaire. Patients who completed the shortened questionnaire included those who could not be reached by phone and did not return a BIOS questionnaire. In this short questionnaire, educational level, comorbidities, the EQ-5D-3L and the IES were included.This short questionnaire did not include proxy assessments. In the shortened questionnaire, pre-injury HS was not collected.

If participants did not complete the questionnaire, they were not excluded from the study but they were still invited at the subsequent time points.

Outcome measures

The EQ-5D and HUI are used in various studies measuring HS after trauma [9, 11, 24, 26, 34–39]. The EQ-5D provides valid results for trauma patients when it is completed by a proxy informant [40]. Moreover, a combination of the EQ-5D and the HUI is recommended for use in trauma patients since the combination of these measures covers all relevant dimensions of health [37, 41].

The EQ-5D consists of the EQ-5D descriptive system and the EQ-visual analogue scale (EQ-VAS). The EQ-5D comprises the following five dimensions: ‘mobility’, ‘self-care’, ‘usual activities’, ‘pain/discomfort’ and ‘anxiety/depression’. Each dimension can be scored as ‘no problems’, ‘moderate problems’ or ‘severe problems’ [30]. A scoring algorithm is available by which each HS description can be expressed as a summary score. This summary score ranges from 0 for death and 1 for full health and can be interpreted as a judgment on the relative desirability of an HS compared with perfect health. A summary score of these five dimensions (EQ-5D utility) can be calculated by using the Dutch tariffs [42]. The EQ-VAS is a vertical visual analogue scale with 0 indicating the worst imaginable health state and 100 indicating the best imaginable health state. The EQ-5D and EQ-VAS were also measured pre-injury, by asking participants 1 week or 1 month and proxy informants 1 month after the trauma for the patients’ HS before sustaining the injury. The EQ-VAS was not included in the short questionnaire.

The HUI is a self-administered HS questionnaire that covers the main health domains that are affected by injury, with a particular focus on functional capacities. The HUI consists of 15 questions, classifying respondents into either the HUI2 or HUI3 health states [31]. Single-attribute and overall HS utility scores are calculated using the respective HUI2 and HUI3 utility functions. The results of the HUI questionnaires are converted by an algorithm into the levels of the complementary HUI2 and HUI3 classification system to form seven-element and eight-element health state vectors. From these vectors, single-attribute and overall health state utility scores are calculated [31].

For both the EQ-5D and the HUI, a scoring algorithm is used in which a score of 1 represents full health, 0 represents death and negative values indicate a HS of worse than death [30, 31].

The Hospital Anxiety and Depression Scale (HADS) was used to assess symptoms of anxiety and depression [32]. The HADS consists of 14 questions, 7 for symptoms of anxiety (HADSA) and 7 for depressive symptoms (HADSD). All questions have a 4-point response scale (0–3) and the scores for both subscales ranged from 0 to 21. A higher subscale score indicates greater severity of symptoms for anxiety and depression with a subscale value of ≥11 indicating a probable case (i.e., clinical symptoms) [32].

The IES was used to assess self-reported symptoms of post-traumatic stress [33]. The IES consists of 15 items of which the patient could use a 4-point scale (0 = not at all, 1 = rarely, 3 = sometimes and 5 = often) to determine whether the statement is present during the last seven days. The IES measures intrusive re-experience of the trauma and avoidance of trauma-related stimuli. A total sum score for the IES could be calculated ranging from 0 to 75. A score of ≥35 is considered as having symptoms of post-traumatic stress [33].

Prognostic factors

Hospital length of stay (LOS), admission to an Intensive Care Unit (ICU) and type and severity of injury were collected from the Brabant Trauma Registry and merged with the BIOS-data. The Abbreviated Injury Scale (AIS) codes (AIS-90, update 2008) [43] were used to create 14 injury group classifications (e.g. hip fracture, severe abdominal injury) representing the most common types of injuries (see S1 Table). Patients who suffer multiple injuries could be classified into one or more injury group classifications.

Trauma severity was based on the Injury Severity Score (ISS). The ISS is based on the square of the highest Abbreviated Injury Scales (AIS) scores of the three most severely injured body regions with a range of 1 to 75. An ISS of ≥16 is considered severely injured [43].

To determine socio-economic status (SES), educational level was used. Educational level was categorized into three levels; low (primary education, preparatory secondary vocational education or without diploma), middle (university preparatory education, senior general secondary education or senior secondary vocational education and training), and high (academic degree or university of applied science).

Statistical analyses

Patients were included in the analyses if they completed a questionnaire for at least one of the predetermined time points. For the non-responders of the BIOS, we could not obtain educational level. Therefore, in the non-responders analysis, status scores from 2014 were used as a proxy to indicate SES. Status scores were based on the mean income, % of people with a low income, % of people with low educational level and % of unemployed people in the neighborhood. In 2014, the mean status score in the Netherlands was 0.28 [44]. A lower status score indicates a lower SES whereas a higher status score indicates a higher SES. Responders and non-responders were compared on age, gender, status score, ISS, type of trauma, LOS and admission to an ICU using Mann-Whitney U tests and Chi-square tests (χ2).

Means and standard deviations (SDs) of the EQ-5D-3L, HUI2, HUI3, HADSD, HADSA and IES summary scores were calculated and reported for the total BIOS population and for the different subcategories.

Multiple imputation was conducted with the Multivariate Imputation by Chained Equations procedure [45] to handle missing baseline characteristics and sum scores of the questionnaires due to missing item scores (see S1 File). The dataset was imputed 15 times with 5 iterations. Sensitivity analysis was performed in which only complete cases were included to compare results with the imputed datasets. S2 Table shows the differences between the original and imputed data.

Score options from each dimension of the EQ-5D were dichotomized into 0 = ‘no problems’ and 1 = ‘moderate problems’/‘severe problems’.

Four linear mixed models [46] with a random intercepts were performed to assess longitudinal association between prognostic factors and HS over the 24 months after trauma, which were divided into short-term (1 week and 1 month), mid-term (3 and 6 months) and long-term (12 and 24 months) associations. HS was measured with the EQ-5D-3L summary score.

The results were considered statistically significant at a level of p<0.05. All analyses were conducted in SPSS V.24 (Statistical Package for Social Sciences, Chicago, Illinois, USA), except of the multiple imputation which was performed in R version 3.4.0 (The R Project for Statistical Computing).

Results

BIOS cohort

During the inclusion period of the BIOS, a total of 10,227 patients were hospitalized because of trauma in one of the participating study centers. Patients were excluded if they did not speak the Dutch language (n = 194), had no permanent address (n = 32), died during their hospital stay within the first week after trauma (n = 219) or had other reasons (n = 8) (e.g., living abroad). Thus, 9,774 patients were eligible for participation in the BIOS, of whom 4,883 patients provided informed consent and were included (50% response rate). Of these 4,883 participants, 1,099 filled out the shortened questionnaires (see Fig 1).

Fig 1

Number of participants troughout the Brabant Injury Outcome Surveillance (n = 4,883).

At 1 week and 1, 3, 6, 12 and 24 months following the trauma, we collected data from of 1,776, 2,971, 3,109, 3,418, 3,105 and 2,734 participants (36.4%, 60.8%, 63.7%, 69.9%, 63.6% and 56.0%, respectively, of the study population) (see Fig 1). A total of 1,105 participants (22.6% of the study population) completed all BIOS questionnaires at each time point. In addition, data on pre-injury HS were obtained from 3,366 participants (69% of the study population). After the first week assessment, missing questionnaires were the result of non-response (i.e., patients who had provided no data at any of the previous time points) and loss to follow-up (i.e., patients who had provided data for at least one of the previous time points). The main reason for participants to be lost to follow-up during the study period was that completing the questionnaires was too time consuming. Elderly, participants with low educational levels, longer hospital LOS, moderate injury (ISS 9–15), a hip fracture, severe traumatic brain injury (TBI) and those with severe abdominal trauma showed lower response rates to the 1 week questionnaire but provided data thereafter. In the BIOS, patients aged 18–24 and those who recovered completely were most likely to be lost to follow-up.

Study population

The median age of the study population was 68 years (IQR 53–80) (Table 1). Responders had a median ISS of 5 (IQR [4-9]) and a large part of the population reported comorbidities. Mild TBI (27.1%) and hip fracture (25.9%) were the most common types of trauma among the participants included in the BIOS. The majority of the participants (n = 2,562, 52.5%) had a low educational level. A total of 407 participants (8% of the study population) were represented by a proxy informant.

Table 1

Characteristics of responders and non-responders of the Brabant Injury Outcome Surveillance.

Characteristics	Responders (n = 4,883) %	Non-responders (n = 4,891) %
Gender (male)	2,329 (47.7%)	2,407 (49.0%)
Median age (yrs)	68 (IQR 53–80)	70 (IQR 46–84)
18–24	217 (4.4%)	400 (8.2%)
25–44	516 (10.6%)	767 (15.7%)
45–64	1,364 (27.9%)	1,006 (20.6%)
65–74	963 (19.7%)	563 (11.5%)
75–84	1,102 (22.6%)	1,030 (21.1%)
≥85	721 (14.8%)	1,125 (23.0%)
Missing	0 (0.0%)	0 (0.0%)
Median SES status score	0.33 (IQR -0.24–0.84)	0.13 (IQR -0.36–0.73)
Missing	60 (1.2%)	68 (1.4%)
Median days hospital LOS	4 (IQR 2–8)	4 (IQR 2–8)
≤2	1,325 (27.1%)	1,528 (31.2%)
3–7	1,944 (39.8%)	1,642 (33.6%)
8–14	937 (19.2%)	911 (18.6%)
≥15	346 (7.1%)	421 (8.6%)
Missing	331 (6.8%)	389 (8.0%)
Type of injury
Pelvic injury	293 (6.0%)	151 (3.1%)
Hip fracture	1,266 (25.9%)	1,099 (22.5%)
Tibia, complex foot or femur fracture	569 (11.7%)	505 (10.3%)
Shoulder and upper arm injury	473 (9.7%)	417 (8.5%)
Radius, ulna or hand fracture	308 (6.3%)	283 (5.8%)
Mild TBI	1,324 (27.1%)	1,443 (29.5%)
Serious TBI	126 (2.6%)	130 (2.7)
Severe TBI	77 (1.6%)	77 (1.6)
Facial fracture	249 (5.1%)	303 (6.2%)
Thoracic injury	198 (4.1%)	162 (3.3%)
Rib fracture	541 (11.1%)	398 (8.1%)
Mild abdominal injury	87 (1.8%)	89 (1.8%)
Severe abdominal injury	36 (0.7%)	30 (0.6%)
Spinal cord injury	27 (0.6%)	10 (0.2%)
Stable vertebral fracture or disc injury	301 (6.2%)	249 (5.1%)
Injury severity	5 (IQR 4–9)	5 (IQR 2–9)
ISS 1–3	1,145 (23.4%)	1,360 (27.8%)
ISS 4–8	1,597 (32.7%)	1,320 (27.0%)
ISS 9–15	1,857 (38.0%)	1,627 (33.3%)
ISS ≥16	239 (4.9%)	194 (4.0%)
Missing	45 (0.9%)	390 (8.0%)
ICU-admission (yes)	358 (7.3%)	292 (6.0%)
Missing	0 (0.0%)	0 (0.0%)

Abbreviations:SES, social-economic status; ICU, intensive care unit; ISS, Injury Severity Score; IQR, Interquartile range; LOS, length of stay; TBI, traumatic brain injury; yrs, years.

Compared to the non-responders, participants were more severely injured and had a higher probability of being admitted to the ICU. In addition, responders had a higher median status score (based on the mean income, % of people with a low income, % of people with low educational level and % of unemployed people in the neighborhood) compared to the general Dutch population (mean 0.28) and compared to the median status score of the non-responders (median 0.33, min. score -3.03, max. score 2,58). Patients aged 18–44 and ≥85 years showed relatively low response rates (35%-40% and 39%, respectively). Patients with minor injuries (ISS 1–3) revealed a low response rate (46%), as well as patients with a hospital LOS of ≤2 or ≥15 days (46% and 45%, respectively).

Health status

The mean EQ-5D-3L summary score increased from 0.49 (SD 0.32) at 1 week post-trauma to 0.77 (SD 0.26) at 24 months post-trauma. The mean pre-injury EQ-5D score was 0.85 (SD 0.23). In addition, the mean (SD) HUI2 and HUI3 scores increased from 0.61 (0.22) and 0.38 (0.31) at 1 week post-trauma to 0.77 (0.21) and 0.62 (0.35) at 24 months post-trauma, respectively (see Table 2). With regard to the individual domains of the EQ-5D, trauma patients reported various problems on the ‘mobility, ‘usual activities’ and ‘pain/discomfort’ dimensions during the 24 months of follow-up (see Fig 2). In addition, during the 24 months, the prevalence of problems on all dimensions of the EQ-5D decreased, but remained higher at 24 months compared to pre-injury (46% and 32%, respectively for mobility, 23% and 16%, respectively for self-care, 44% and 26%, respectively for usual activities, 52% and 32%, respectively for pain/discomfort and 22% and 16%, respectively for anxiety/depression).

Table 2

Mean (SD) summary scores of self-reported health status and psychological outcomes up to 2 years post-trauma.

Time post-trauma	EQ-5D-3L*		HUI2**		HUI3**		HADSA***		HADSD***		IES****
	N	Mean (SD)	N	Mean (SD)	N	Mean (SD)	N	Mean (SD)	N	Mean (SD)	N	Mean (SD)
1 week	1,776	0.49 (0.32)	1,776	0.61 (0.22)	1,776	0.38 (0.31)	1,776	4.92 (3.98)	1,776	5.00 (4.28)	1,776	14.80 (15.80)
1 month	2,971	0.56 (0.30)	2,971	0.67 (0.22)	2,971	0.45 (0.35)	2,644	4.81 (3.95)	2,644	4.77 (4.17)	2,644	14.44 (15.73)
3 months	3,131	0.69 (0.27)	2,735	0.72 (0.22)	2,735	0.53 (0.35)	2,423	4.57 (3.80)	2,423	4.24 (4.02)	2,819	12.75 (15.47)
6 months	3,433	0.74 (0.25)	2,442	0.75 (0.22)	2,442	0.58 (0.35)	2,191	4.21 (3.79)	2,191	3.91 (4.01)	3,182	11.42 (15.28)
12 months	3,151	0.76 (0.25)	2,391	0.76 (0.22)	2,391	0.60 (0.36)	2,165	4.32 (3.78)	2,165	3.74 (3.97)	2,925	10.98 (14.98)
24 months	2,734	0.77 (0.26)	2,137	0.77 (0.21)	2,137	0.62 (0.35)	1,990	4.31 (3.76)	1,990	3.62 (3.87)	2,587	10.35 (14.72)

*Completed by total study population.

**Not administered to patients who completed only the short questionnaire.

***Not administered to proxy participants or to patients who completed only the short questionnaire.

****Not administered to patients aged ≥65 with a hip fracture who completed only the short questionnaire.

Range of EQ-5D-3L: 0–1. Mean EQ-5D-3L of the general Dutch population: 0.87 [47].

Range of HUI 2/3: 0–1

Range of HADSA and HADSD: 0–21

Range of IES: 0–75

Abbreviations: SD, standard deviation; EQ-5D, EuroQol-5D-3L; HUI2/3, Health Utilities Index Mark 2/3; HADS, Hospital Anxiety and Depression Scale; HADSA, Hospital Anxiety and Depression scale, subscale anxiety; HADSD, Hospital Anxiety and Depression scale, subscale depression; IES, Impact of Event Scale.

Fig 2

Prevalence of moderate or severe problems (%) on each EuroQol-5D-3L dimension up until 2 years of follow-up.

Psychological outcomes

The mean summary scores (SD) of the HADSA (clinical symptoms of anxiety), HADSD (clinical symptoms of depression) and IES (symptoms of post-traumatic stress) slightly decreased from 4.92 (3.98), 5.00 (4.28) and 14.80 (15.80), respectively, at 1 week post-trauma to 4.31 (3.76), 3.62 (3.87) and 10.35 (14.72), respectively, at 24 months post-trauma (see Table 2).

The prevalence rates of clinical symptoms of anxiety (HADSA≥11), depression (HADSD≥11) and post-traumatic stress (IES≥35) reported at 1 week post-trauma were 10.2%, 12.3% and 13.5%, respectively, and showed a small decrease over time to 7.8%, 6.8% and 11.0% at 24 months post-trauma, respectively.

The results revealed that patients with symptoms of post-traumatic stress (IES≥35) showed worse outcomes on the EQ-5D-3L than patients with no symptoms of post-traumatic stress (Fig 3).

Fig 3

Mean summary scores of the EuroQol-5D-3L questionnaire and patients reporting symptoms on the Impast of Event Scale.

Prognostic factors of health status

Overall, HS measured as with the EQ-5D-3L increased over at least up until 6 months for all groups of patients and stabilized between 6 and 12 months post-trauma for most groups (see Table 3). Female patients had a lower HS compared to males at every time point.

Table 3

Mean (SD) of the self-reported health status for patient and injury characteristics as measured with the EQ-5D-3L.

		Mean (SD) EQ-5D-3L summary score
Characteristics		Pre-injury	1 week	1 month	3 months	6 months	12 months	24 months
Gender	Male	0.90 (0.19)	0.54 (0.32)	0.62 (0.28)	0.74 (0.25)	0.79 (0.23)	0.83 (0.21)	0.84 (0.21)
	Female	0.80 (0.25)	0.43 (0.31)	0.50 (0.30)	0.64 (0.28)	0.69 (0.26)	0.70 (0.27)	0.72 (0.28)
Age (yrs)	18–24	0.95 (0,13)	0.50 (0.32)	0.63 (0.24)	0.79 (0.24)	0.85 (0.21)	0.86 (0.21)	0.86 (0.22)
	25–44	0.95 (0.12)	0.45 (0.30)	0.59 (0.29)	0.74 (0.24)	0.79 (0.24)	0.84 (0.22)	0.87 (0.19)
	45–64	0.91 (0.17)	0.49 (0.31)	0.60 (0.27)	0.73 (0.24)	0.79 (0.21)	0.83 (0.21)	0.83 (0.22)
	65–74	0.88 (0.20)	0.51 (0.31)	0.62 (0.28)	0.73 (0.24)	0.79 (0.22)	0.80 (0.22)	0.81 (0.23)
	75–84	0.78 (0.26)	0.52 (0.33)	0.53 (0.31)	0.66 (0.27)	0.70 (0.25)	0.70 (0.26)	0.70 (0.28)
	≥85	0.63 (0.28)	0.40 (0.33)	0.39 (0.32)	0.50 (0.29)	0.55 (0.29)	0.57 (0.29)	0.56 (0.30)
Educational level	Low	0.78 (0.27)	0.49 (0.33)	0.52 (0.31)	0.65 (0.28)	0.70 (0.27)	0.71 (0.27)	0.72 (0.28)
	Middle	0.90 (0.17)	0.50 (0.31)	0.59 (0.29)	0.71 (0.26)	0.77 (0.24)	0.80 (0.23)	0.81 (0.23)
	High	0.93 (0.13)	0.48 (0.30)	0.62 (0.26)	0.75 (0.23)	0.80 (0.21)	0.84 (0.20)	0.86 (0.20)
Comorbidity	0	0.96 (0.12)	0.53 (0.30)	0.65 (0.26)	0.77 (0.23)	0.83 (0.19)	0.86 (0.18)	0.87 (0.18)
	1	0.85 (0.21)	0.49 (0.32)	0.56 (0.29)	0.69 (0.25)	0.75 (0.24)	0.77 (0.24)	0.78 (0.25)
	2	0.75 (0.24)	0.45 (0.30)	0.47 (0.31)	0.61 (0.29)	0.66 (0.26)	0.67 (0.27)	0.68 (0.28)
	≥3	0.64 (0.29)	0.40 (0.33)	0.42 (0.32)	0.56 (0.29)	0.58 (0.29)	0.59 (0.29)	0.59 (0.29)
ISS	1–3	0.86 (0.22)	0.63 (0.29)	0.69 (0.27)	0.78 (0.23)	0.79 (0.24)	0.79 (0.25)	0.81 (0.25)
	4–8	0.89 (0.20)	0.46 (0.31)	0.56 (0.29)	0.71 (0.25)	0.77 (0.23)	0.80 (0.22)	0.81 (0.23)
	9–15	0.80 (0.26)	0.43 (0.30)	0.50 (0.30)	0.63 (0.28)	0.68 (0.27)	0.70 (0.28)	0.72 (0.28)
	≥16	0.90 (0.18)	0.37 (0.35)	0.50 (0.31)	0.65 (0.30)	0.74 (0.25)	0.77 (0.24)	0.77 (0.25)
LOS (days)	≤2	0.91 (0.18)	0.61 (0.29)	0.70 (0.25)	0.81 (0.21)	0.83 (0.20)	0.84 (0.22)	0.86 (0.21)
	3–7	0.85 (0.23)	0.47 (0.30)	0.55 (0.28)	0.69 (0.25)	0.74 (0.24)	0.77 (0.24)	0.79 (0.24)
	8–14	0.78 (0.27)	0.31 (0.31)	0.46 (0.29)	0.59 (0.29)	0.65 (0.27)	0.68 (0.28)	0.67 (0.29)
	≥15	0.74 (0.28)	0.28 (0.34)	0.32 (0.31)	0.50 (0.29)	0.58 (0.28)	0.60 (0.28)	0.62 (0.28)

Range of EQ-5D-3L: 0–1. Mean EQ-5D-3L of the general Dutch population: 0.87.

Abbreviations: SES, socio-economic status; SD, standard deviation; EQ-5D, EuroQol-5D-3L; ISS, Injury Severity Score; LOS, length of hospital stay; yrs, years

At all time points, patients aged 85 and older had the lowest HS compared to the other age categories. At 3 and 6 months, all patient groups between 25 and 74 years reported the same HS whereas patients aged between 18 and 24 reported a higher EQ-5D summary score. HS stabilized at 6 or 12 months for every age group, except for patients between 25 and 44 years for whom HS increased further.

Except for 1 week, patients with a high educational level had the highest HS.

With an increasing number of comorbidities, HS decreased. Patients with moderate injuries (ISS 9–15) showed on almost each time point the lowest HS. At 1 week, severely injured patients (ISS ≥16) showed the lowest mean HS (0.28, SD 0.35).

Patients with the longest hospital LOS (≥15 days) had the lowest mean HS at all time points, ranging from 0.28 (SD 0.34) at 1 week up to 0.62 (SD 0.28) at 24 months after trauma.

After adjustment for confounding factors, short-term (1 week and 1 month) prognostic factors for a significant lower EQ-5D summary score were female gender, a higher number of comorbidities, a longer LOS, a higher ISS, pelvic injury, tibia/complex foot or femur fracture, radius/ulna/hand fracture, shoulder/upper arm injury, rib fracture, spinal cord injury and stable vertebral fracture/disc injury (see Table 4). Mid-term (3 and 6 months) prognostic factors were a higher number of comorbidities, an ISS between 4 and 15, a longer LOS, radius/ulna/hand fracture, tibia/complex foot or femur fracture, severe TBI, spinal cord injury and stable vertebral fracture/disc injury. Long-term (12 and 24 months) prognostic factors for a lower HS were: age 75 and above, 2 or more comorbidities, a longer LOS, tibia/complex foot or femur fracture, spinal cord injury and stable vertebral fracture/disc injury were prognostic factors. A high educational level was associated with higher HS in the long-term analysis.

Table 4

Multivariable longitudinal analysis of short, mid and long-term prognostic factors for decreased health status as measured with the EQ-5D-3L.

			Total (1 week-24 months)	Short-term (1 week-1 month)	Mid-term (3–6 months)	Long-term (12–24 months)
		Number of patients	* Beta (95% C.I.)	* Beta (95% C.I.)	* Beta (95% C.I.)	* Beta (95% C.I.)
Characteristics		4,809	n = 4,809	n = 3,314	n = 4,048	n = 3,422
Gender	Male	2,291	ref	ref	ref	ref
	Female	2,518	-0.05 (-0.06; -0.04)	-0.06 (-0.08; -0.05)	-0.05 (-0.06;-0.03)	-0.06 (-0.08; -0.05)
Age (yrs)	18–44	721	ref	ref	ref	ref
	45–64	1,345	0.04 (0.02; 0.05)	0.06 (0.04; 0.09)	0.02 (-0.00; 0.04)	0.01 (-0.01; 0.03)
	65–74	956	0.09 (0.08; 0.11)	0.14 (0.11; 0.17)	0.07 (0.05; 0.09)	0.04 (0.02; 0.06)
	≥75	1,787	0.02 (0.01; 0.04)	0.09 (0.06; 0.12)	-0.01 (-0.03; 0.01)	-0.05 (-0.07; -0.02)
Educational level	Low	2,637	ref	ref	ref	ref
	Middle	1,280	0.02 (0.00; 0.03)	0.01 (-0.01; 0.03)	0.01 (-0.00; 0.02)	0.03 (0.01; 0.04)
	High	893	0.03 (0.02; 0.05)	0.02 (-0,01; 0,04)	0.04 (0.02; 0.05)	0.05 (0.04; 0.07)
Number of comor-bidities	None	1,801	ref	ref	ref	ref
	1	1,426	-0.06 (-0.07; -0.05)	-0.06 (-0.08; -0.04)	-0.06 (-0.07; -0.04)	-0.05 (-0.07; -0.04)
	≥2	1,582	-0.15 (-0.16; -0.14)	-0.14 (-0.17; -0.12)	-0.14 (-0.16;-0.13)	-0.16 (-0.18; -0.15)
ISS	1–3	1,139	ref	ref	ref	ref
	4–8	1,596	0.00 (-0.01; 0.02)	-0.03 (-0.06; -0.01)	0.01 (-0.05; 0.03)	0.03 (0.01; 0.05)
	9–15	1,838	-0.02 (-0.04; -0.00)	-0.06 (-0.09; -0.02)	-0.02 (-0.04; -0.01)	0.00 (-0.02; 0.03)
	≥16	235	0.01 (-0.01; 0.04)	-0.02 (-0.08; 0.04)	0.01 (-0.04; 0.05)	0.04 (-0.01; 0.08)
LOS	≤2	1,425	ref	ref	ref	ref
	3–7	2,053	-0.05 (-0.06; -0.04)	-0.08 (-0.10; -0.06)	-0.06 (-0.07; -0.04)	-0.03 (-0.04; -0.01)
	8–14	975	-0.11 (-0.12; -0.09)	-0.15 (-0.18; -0.12)	-0.11 (-0.13; -0.09)	-0.08 (-0.11; -0.06)
	≥15	357	-0.20 (-0.22; -0.18)	-0.24 (-0.28; -0.20)	-0.20 (-0.23; -0.17)	-0.17 (-0.20; -0.14)
Injury	Pelvic injury	286	-0.05 (-0.07; -0.03)	-0.13 (-0.16; -0.10)	-0,01 (-0.04; 0.01)	0.00 (-0.02; 0.03)
	Hip fracture	1,242	-0.02 (-0.04; -0.01)	-0.02 (-0.05; 0.01)	-0.03 (-0.05; -0.00)	-0.02 (-0.05; 0.00)
	Tibia, complex foot or femur fracture	561	-0.05 (-0.06; -0.04)	-0.11 (-0.14; -0.08)	-0.05 (-0.07; -0.02)	-0.02 (-0.04; 0.00)
	Shoulder and upper arm injury	469	-0.03 (-0.04; -0.01)	-0.07 (-0.10; -0.05)	-0.02 (-0.04; 0.00)	-0.00 (-0.02; 0.02)
	Radius, ulna or hand fracture	304	0.00 (-0.02; 0.02)	-0.03 (-0.06; 0.00)	0.00 (-0.02; 0.03)	0.02 (-0.00; 0.05)
	Mild TBI (AIS 1–2)	1,302	0.03 (0.02; 0.04)	0.06 (0.04; 0.08)	0.03 (0.02; 0.05)	0.02 (-0.00; 0.03)
	Serious TBI (AIS 3)	125	0.04 (0.01; 0.07)	0.07 (0.02; 0.13)	0.05 (0.01; 0.10)	0.01 (-0.03; 0.05)
	Severe TBI (≥ 4)	76	-0.03 (-0.06; 0.02)	-0.02 (-0.10; 0.07)	-0.06 (-0.12; 0.00)	-0.01 (-0.07; 0.05)
	Facial fracture	243	0.02 (0.01; 0.04)	0.05 (0.01; 0.09)	0.02 (-0.01; 0.04)	0.01 (-0.01; 0.04)
	Thoracic injury	198	0.04 (0.02; 0.06)	0.05 (0.01; 0.10)	0.04 (0.02; 0.07)	0.03 (-0.01; 0.06)
	Rib fracture	533	0.01 (-0.00; 0.03)	-0.01 (-0.04; 0.01)	0.03 (0.01; 0.05)	0.01 (-0.01; 0.04)
	Mild abdominal injury	85	0.02 (-0.02; 0.05)	0.04 (-0.02; 0.10)	0.01 (-0.04; 0.05)	0.02 (-0.03; 0.07)
	Severe abdominal injury	36	0.03 (-0.02; 0.08)	-0.05 (-0.15; 0.05)	0.06 (-0.04; 0.13)	0.05 (-0.02; 0.13)
	Spinal cord injury	27	-0.11 (-0.17; -0.05)	-0.10 (-0.22; 0.03)	-0.12 (-0.20; -0.03)	-0.15 (-0.23; -0.07)
	Stable vertebral fracture or disc injury	296	-0.05 (-0.07; -0.03)	-0.08 (-0.12; -0.05)	-0.05 (-0.08; -0.03)	-0.03 (-0.06; -0.01)

* Mixed models, adjusted for all other variables

Beta: mean increase in EQ-5D-3L score (improvement of Health Status) compared to the reference category.

Range of EQ-5D-3L: 0–1. Mean EQ-5D-3L of the general Dutch population: 0.87 [47].

Abbreviations: ISS, Injury Severity Score; AIS, Abbreviated Injury Scale; LOS, Hospital lenght of stay in days; TBI, traumatic brain injury; yrs = years

Discussion

This study describes HS and psychological outcomes over 24 months after trauma. HS markedly improved during the 24 months after trauma, most of which occurred within the first 3 months. Compared to pre-injury HS, a large decrease in HS was found 1 week post-trauma. The prevalence rates of symptoms of anxiety and depression were relatively low. In contrast, symptoms of post-traumatic stress were highly prevalent and were present five times as often as compared with the Dutch general population [48]. Additionally, in line with the literature, having symptoms of post-traumatic stress were associated with a lower HS [27]. The addition of an assessment at 1 week post-trauma in the present study adds detailed insight into (baseline) functioning shortly after trauma. Therefore, it provides a more valid assessment of the magnitude of recovery thereafter.

Between 1 week and 3 months, the percentage of patients who reported problems on the ‘pain/discomfort’ and ‘self-care’ dimensions of the EQ-5D decreased steeply. For the ‘mobility’ and ‘usual activities’ dimensions, this decrease started 1 month after trauma. The percentage of patients who reported problems on the ‘anxiety and depression’ dimension was the highest at 1 month after trauma. Within 6 months post-trauma, patients showed the most recovery. From 6 months post-trauma onwards, little improvement in overall HS was found. The percentage of patients who reported improvements on the different EQ-5D domains increased through the 24 months of follow-up. However, the vast majority of trauma patients did not recover to their pre-injury HS. The mean EQ-5D-3L summary score for the total Dutch adult population is considered 0.87 (SD 0.18) (for males 0.89 (SD 0.16) and for females 0.85 (SD 0.19)) [47].

At short-term (up to 1 month post-trauma) female gender, lower extremity, spine, shoulder and upper arm injuries, injury severity, comorbidities and a longer hospital stay were associated with lower HS. At mid-term (3 and 6 months), almost the same prognostic factors were significant and relevant, however, only injury severity seemed to be less important. In the long-term a greater age, two or more comorbidities, a longer hospital stay and only a few injuries (i.e., lower extremity fracture and spine injury) showed a significantly lower HS. The effect of injury severity seemed to fade over time. Spinal cord injury patients had the highest risk (long-term Beta -0,18, CI -0,27;-0,08) of a lower HS during both the short (not significant), mid and long-term post-trauma. Middle and high educational levels were associated with a higher HS in the long term compared to those with low educational levels.

Most recovery in HS occurs up to 3 months post-trauma, which is in agreement with previous studies on this topic [9, 10, 19]. In this regard, the addition of an assessment at 1 week post-trauma in the present study adds detailed insight into (baseline) functioning shortly after trauma. Prior work also confirms the finding that a large proportion of patients have a considerably lower HS 1 year post-trauma compared to pre-injury HS [9, 11, 16] or compared to the HS of the general population [49].

In addition to the physical injury itself, other characteristics largely affect HS after trauma. This was particularly in the long-term. This finding extends those of previous studies [15, 35], confirming that patients who were the most severely injured, were not necessarily those with the lowest HS.

In this study, the prevalence of symptoms of anxiety and depression was slightly higher compared to the prevalence of an anxiety disorder or depression in the general Dutch population (both disorders are estimated to be present in 10.0% of the Dutch population) [50, 51]. As a result, the prevalence of symptoms of anxiety and depression slightly decreased over time. Recent and comparable studies have documented a slightly higher prevalence of symptoms of anxiety and depression [9, 16].

The prevalence of symptoms of post-traumatic stress was high as compared with the Dutch population (11.0% at 2 years post-trauma versus 2.6–3.3%) [48]. Our prevalence rate of long-term symptoms of post-traumatic stress is in line with previous research that also uses the ≥35 cut-off point for the IES [13]. However, the systematic review by Haagsma et al. [52] revealed prevalence of post-traumatic stress in hospitalized trauma patients that ranged from 30% (90% CI 27%-33%) within 3 months post-trauma to 6% (90% CI 4%-10%) at 1 year. Compared to those results, we found a lower prevalence of symptoms of post-traumatic stress early post-trauma whereas a higher prevalence was found at the 1 year follow-up. This discrepancy may be due to studies using various instruments and different cut-off points to indicate post-traumatic stress.

This study was conducted according to the recommended guidelines for measuring non-fatal outcomes after trauma [41]. The BIOS included a broad study population, measured both short-term and long-term outcomes, measured functioning prior to the trauma, included a large number of patient and injury-related characteristics and requested information from proxy informants for patients who were incapable of completing the set of questionnaires themselves. Recruitment for the BIOS occurred in all hospitals of the Dutch Noord-Brabant region, covering both urban and rural populations.

The inclusion of elderly individuals with a hip fracture in the current study improves the generalizability of the study findings. In this study, one out of four participants was aged 65 or older and had a hip fracture. We are aware that recovery patterns (due to comorbidities and functional decline) in the elderly population are different from recovery patterns in younger patients.

This study also has several limitations. First, there was selection bias since younger patients, elderly patients, patients with very minor injuries (ISS 1–3) and those with a low status score (used as a proxy to indicate SES; a low status score indicate a lower SES-level) were less likely to participate. It is challenging to include these specific groups. Previous studies also reported lower response rates from younger and elderly patients [7, 10, 24, 35], from patients with minor injuries [10] and from those with low educational level [53, 54]. Second, only a selected group of patients (18% of the eligible population) completed the 1 week assessment. Since most recovery occurs within the first 3 months post-trauma, it is vital to examine very early recovery patterns. This study provides unique data since it incorporates the use of a standardized 1 week assessment in a comprehensive group of trauma patients. Nevertheless in most cases, non-responders at this time point felt too disabled to complete a long questionnaire. Therefore, this most likely led to an underestimations of the reported HS and psychological outcomes 1 week post-trauma. Third, there were many missing data, especially early post-trauma. However, we were able to impute missing sum scores in several cases. Fourth, we did not correct for pre-injury HS in the longitudinal analyses since we were not able to collect pre-injury HS in all participants. Fifth, we did not include a pre-injury assessment of the symptoms of anxiety or depression. Sixth, we cannot exclude change findings due to multiple testing. Change findings might have occurred for the results of the longitudinal analyses as presented in Table 4.

Given the acute nature of trauma, it is difficult to include patients soon after they experience trauma in order to examine very early recovery patterns. To increase the response rate and to reduce loss to follow-up, future research should minimize the large number of questionnaires that patients have to complete at each time point, especially early post-trauma. A promising technique includes computerized adaptive testing. In this already proven valid technique, tailored-made short and precise domain-specific data can be collected [55-58].

We are aware that recall bias and response shift most likely led to an overestimation of the pre-injury HS as measured in this study. However, to produce valid estimates of the health impact and the decrease in functioning after trauma, information on patients functioning prior to the trauma is crucial [59-62]. Future research should focus on the effects of recall bias and response shift on retrospectively collected data.

Implications for health-care

According to the literature, early recognition, treatment and monitoring of psychological problems improve non-fatal outcomes after trauma [16, 63–66]. Therefore, early screening and interventions to reduce symptoms of post-traumatic stress should be part of standard care. Furthermore in the long long-term, patients aged ≥75 years, patients with a longer length of hospital stay and patients with ≥2 comorbidities are more likely to have a poor HS. For these patients, standard aftercare should be extended to screen for remaining problems that the patients should address after their trauma. For example by a follow-up appointment with a case manager.

Symptoms of post-traumatic stress were frequently reported in this study. Patients need to be better informed about the psychological problems they may experience after trauma. Health-care providers should not solely focus on the physical consequences but should also be aware of all other possible consequences that may occur after trauma. To achieve this, a more holistic approach towards the treatment of trauma patients should be aimed in which the patients’ own perspectives on their recovery should play a crucial role.

Conclusion

Hospitalized trauma patients experience substantial reductions in HS and frequently report symptoms of post-traumatic stress. The results of this study should be interpreted with caution since younger patients, patients with very minor injuries and those with a lower status score (used as a proxy to indicate SES, a lower status score indicates a lower SES-level) were less likely to participate. Most improvements in HS and psychological symptoms occurred within the first 3 months post-trauma. After two years post trauma, the vast majority of trauma patients did not achieve their pre-injury HS. Recovery trajectories varied widely in which female gender, age ≥75 years, spinal cord injury, having more comorbidities, low educational level, a longer hospital stay and symptoms of post-traumatic stress were associated with a higher risk of decreased HS in the long-term after trauma. In the short-term, also several lower extremity injuries are prognostic factors for decreased HS.

Injury group classification of the most common types of injury, based on the Abbreviated Injury Score [67].

(DOCX)

Click here for additional data file.

Missing sum scores of the original data and imputed data of the self-reported health status and psychological measures of the participants of the Brabant Injury Outcome Surveillance (n = 4,883).

(DOCX)

Click here for additional data file.

Methods of imputed data of the Brabant Injury Outcome Surveillance.

(DOCX)

Click here for additional data file.

8 Nov 2019

PONE-D-19-21191

Health status and psychological outcome after trauma; a prospective multicenter cohort study

PLOS ONE

Dear MSc. Kruithof,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We would appreciate receiving your revised manuscript by Dec 23 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Melita J. Giummarra

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for including your ethics statement: Ethics committee: Medical Ethics Committee Brabant (project number NL50258.028.14 and NW2016-09).

Prior to participation, participants signed an informed consent form.

Please amend your current ethics statement to confirm that your named institutional review board or ethics committee specifically approved this study.

Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer to http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research.

3. Please carefully proofread your manuscript for typographical errors. For example, on page 7 “a total of 10,227 patients was hospitalized” should be “a total of 10,227 patients were hospitalized”

4. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

5. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

6. Your ethics statement must appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please move it to the Methods section and delete it from any other section. Please also ensure that your ethics statement is included in your manuscript, as the ethics section of your online submission will not be published alongside your manuscript.

7. Please amend your authorship list in your manuscript file to include author Nena Kruithof.

8. One of the noted authors is a group or consortium: BIOS-group. In addition to naming the author group, please list the individual authors and affiliations within this group in the acknowledgments section of your manuscript. Please also indicate clearly a lead author for this group along with a contact email address.

Additional Editor Comments:

In addition to the very helpful comments raised by the reviewers I would like to highlight the following points that require your attention as you revise your manuscript for resubmission.

Methods:

How were comorbidities recorded (e.g., ICD diagnoses?)

For the injury groups, it might be better to have a single variable that codes for the broad types of isolated and combined injuries.

Please provide a STROBE chart in the body of the paper that includes the exclusions, and the data collection/loss to follow-up at each wave of the study.

Results

In addition to the responder analysis, please include analysis of loss to follow-up. I expect the same factors that differentiate between responders/non-responders will be associated with LTFU.

Only summarise data the data that is reported in tables in the text. It is not necessary to fully report the statistics again.

Correct the wording in some phrases to reduce ambiguity. For instance in line 212 change “more often” to “higher proportion” rather than  given that the ICU admission was not a frequency WITHIN patients, but between patients. On line 229 change the wording of “most problems” as the data show the proportion of the sample with problems, not what they had the "most problems" with. On line 251 replace “symptoms” with “clinically significant symptoms" given that this is focused on those with symptoms above 11, not any symptoms.

Ensure that the terms confounding and covariates are used appropriately.

In Table 4 provide columns showing the cell sizes for each level of each factor.

Discussion

There is only very limited discussion of the implications of the research. What new insights does this study add that can be logically and easily included in patient care to improve injury outcomes/reduce injury burden?

Note the level of missing data that were imputed was as high as 63.6% for some variables in the limitations

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you very much for the opportunity to review this manuscript. The authors present a prospective multi-center cohort study that examined patterns of health status and psychological outcomes (post-traumatic stress, depression, and anxiety) in the 24 months after trauma. Short and long-term prognostic factors for lower health status over time were also investigated. The study adhered to guidelines regarding best practice for the conduct of longitudinal studies examining injury-related disability. Furthermore, the study design ensured that data collection occurred more frequently than many other studies involving trauma populations to date. Investigation of psychological outcomes is a significant strength of the study given the absence of research in this area with respect to general injury.

I believe this manuscript makes an important and novel contribution to the literature. However, there are several issues that could be addressed prior to acceptance for publication.

1) There is significant editing that is required to improve the readability of the manuscript. Perhaps someone independent from the research team could complete a thorough proof-read? There are a number of sentences that need edits in order to make sense. For example, in the abstract, the sentence: “At long-term, higher age, comorbidities, longer hospital stay, lower extremity fracture and spine injury showed lower HS” needs to be modified to read: “At long-term follow-up, participants of higher age, with comorbidities, longer hospital stay, lower extremity fracture and spine injury showed lower HS”. Sentences that begin with ‘besides’ and ‘apparently’ should be removed from the manuscript. The EQ-5D is comprised of five dimensions (as opposed to compressed). There are instances of tense changes in the manuscript also.

2) If possible, the authors should examine the relationship between post-traumatic stress (which was found to be high in the study population) and health status over time. If post-traumatic stress is associated with increased risk of low health status, an important implication is that post-traumatic stress can be addressed early after injury to improve long-term outcome.

3) It would also be good if the authors could justify why they have focused on examining predictors of health status at 24 months but not prognostic factors for anxiety, depression, or post-traumatic stress. It would be interesting to know whether there are different factors that are predictive of these outcomes when compared to health status. This information could be used to identify individuals who are at increased risk of poor psychological outcomes after trauma and to inform intervention development.

3) The authors should provide more of a rationale for including the larger sample of participants who completed only health status assessments (and not the psychological outcome questionnaires) in the analyses for this manuscript, particularly because the focus of the study is on examining psychological outcome after trauma.

4) The reason for some patients (those who did not complete a questionnaire up until 3 months post-trauma) completing a short version of the BIOS-questionnaire is not made explicit. Did this short questionnaire include a pre-injury health status assessment?

5) Please describe how exactly comorbidities were assessed in the questionnaires. Furthermore, what demographic characteristics were collected? Collection of education level is described but other variables examined (such as sex/gender) are not clearly identified.

6) Although the authors describe their multiple imputation process, it is unclear how much missing data there was and how many participant scores needed to be imputed.

7) Tables 2 and 3 could include a pre-injury health status column to aid interpretation.

8) A limitation of the study is the lack of pre-injury assessments of post-traumatic stress, depression, and anxiety.

9) Should pre-injury health status also be examined as a predictor in linear mixed models?

Reviewer #2: I congratulate this research team on collecting data on the long-term impacts of injury. This is not an easy task and provides very valuable information. I have several comments.

1. What does this study add to previous research on the subject? This could be made clearer in the Introduction (justify the need for this study) and the Discussion (what this study adds).

2. Acronyms need to be spelt out in the abstract. Please do not use acronyms that are not well-known to readers. It makes reading the article very arduous.

3. ‘We aimed to describe recovery patterns of health status (HS) and psychological outcomes during 24 months of follow-up and to identify subgroups at risk of both short and long-term health problems after trauma’. Authors could better align the presentation and interpretation of results (abstract and manuscript) to these objectives.

4. ‘Reduction of trauma-related mortality in high-income countries [2] resulted in increased numbers of trauma survivors with long-term injury impact, including reduced health status (HS)’. Can authors provide a reference for the last part of this statement?

5. Document needs extensive grammar/spelling correction

6. Study inclusion criteria are not quite clear to me. Did authors include all presentations to the ED with injury (as stated in the Methods) or just hospital admissions (as stated in the Abstract)? If the latter, did they only include patients with a principal diagnosis of injury? Please clarify.

7. What proportion of the study population were elderly patients with an isolated simple fracture after a fall from their own height? These patients constitute a different population. Please comment.

8. Results – no need to repeat information given in tables. You can just highlight salient results.

9. ‘Compared to the non-responders, participants were more severely injured, were more often admitted to the ICU and had a lower SES’. However, responders had a higher SES score. Please explain how SES status scores should be interpreted.

10. Please provide definitions of mild, serious and severe TBI. These are usually referred to as mild, moderate and severe. Furthermore, as I understand it ‘type of injury’ refers to any injury rather than the most serious or isolated? For example a patient with three injuries including AIS severity score of 5 to the head, another injury of 2 to the head and a tibial fracture would be classed in ‘tibia, complex foot or femur fracture’, ‘mild TBI’ and ‘severe TBI’?

11. In Tables 2 and 3 please add Ns and a note in the legend on how to interpret the scales presented. Means in a healthy population would be useful addition. Why not add pre-injury data here? Also presenting mean differences (inter-patient) with regard to baseline would be interesting. I wonder whether presenting as a Figure would facilitate interpretation (as in Figure 1).

12. It would be useful to have a flow diagram describing numbers included in each phase of the study.

13. Table 4 – Are colours intended to represent statistically significant results? If so, it is applied inconsistently. Furthermore, I would avoid bringing attention to results based on statistical significance only. In addition, given the problem of multiple comparisons, statistical significance of variables with more than two categories should be verified with a global test. Please indicate to readers (in the table legend) how to interpret the betas.

14. Were any sensitivity analyses conducted other than patients with no missing data?

15. Please comment on the external validity of the results given less than 10% of the study population had major trauma.

16. Can authors give any information on the validity of data in the trauma registry?

17. It would be interesting to see results stratified for broad groups of injury: TBI, thoraco-abdo, spinal cord, orthopaedic, multi-system.

18. The summary of key results in line with study objectives (first paragraph of the discussion) may be more effective if it were more succinct.

19. Authors should take account of potential biases (e.g. underestimation of the frequency of health problems due to selection bias) when interpreting results in the Conclusion sections (abstract and manuscript).

Reviewer #3: This manuscript concerns a relevant question on recovery after trauma.

The authors chose a novel approach, intending to include all victims of trauma from a specific region, irrespective of the severity.

The methodology seems to be sound, and the manuscript is fairly easy to read.

I have a number of suggestions to increase reader friendliness, and some comments that I think the authors should address and correct.

First: The inclusion criteria are confusing. You say that “Adults (≥18 years) who visited an emergency department ≤48 hours after trauma were invited to participate” – but all patients included seem to have been admitted, “During the inclusion period of the BIOS, a total of 10,227 patients was hospitalized because of a trauma in one of the participating study centers”. Which group of patients was actually included?

Two: You are using a number of scales/tools (EQ-5D-3L, HUI2/3, HADS-A/D, IES). It would be helpful to readers if you explained a little more detailed how the scores were calculated (EQ-5D-3L – “A summary score of these five dimensions (EQ-5D utility) can be calculated by using the Dutch tariffs [40].”). In addition, I suggest that you inform readers of the “normal values” in the Dutch population of the different scales, as done in the discussion line 336-337 for post-traumatic stress.

Socioeconomic status is not explained. How was the SES assessed, and how did respondents compare to the general population?

The challenge with retrospective assessment of pre-injury health is of course that respondents may see their previous health in light of the present status. This is addressed appropriately in the limitations section.

The tables are numerous and difficult to read when not deeply engaged in the study. I suggest that you elaborate on the legends (e.g. table 4: no explanation of the meaning of the colors). You could also exchange tables for figures. The reader when seeing a figure better perceives a longitudinal change in groups.

In short, an interesting study with need for clarification, and presently looking as you have been working hard and long time with data, and forgotten that potential readers does not have the same confidence with the material.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Lynne Moore

Reviewer #3: Yes: Torben Wisborg

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

24 Jan 2020

Rebuttal letter

Response to comments

General comments on the manuscript:

-We placed the DOI-number of the protocol of the Brabant Injury Outcome Surveillance (BIOS) in the manuscript

-Affiliations of the individual members of the BIOS-group are now provided in the acknowledgements. Furthermore, a contact person (including contact details) of the BIOS-group is added

-In this study, there are ethical restrictions on sharing the data set. In the manuscript, we explained in detail were the data can be requested.

Editor comment:

How were comorbidities recorded (e.g., ICD diagnoses?)

A: In the method section, we now described how self-reported comorbidities were reported (i.e. by using a modified version of the Cumulative Illness Rating Scale). We added this information in the subsection ‘data collection’.

For the injury groups, it might be better to have a single variable that codes for the broad types of isolated and combined injuries.

A; There is an infinite number of combined injuries. Therefore, it is not possible to have a single variable for all the codes. In the analyses, we made corrections for the most common types of injuries and we also made corrections for the Injury Severity Scale.

Please provide a STROBE chart in the body of the paper that includes the exclusions, and the data collection/loss to follow-up at each wave of the study.

A: In Fig. 1, we now provide a more detailed description of the patient flow of the study. Since patients could flow into the study at 5 different time points, it was not possible to make a STROBE chart.

In addition to the responder analysis, please include analysis of loss to follow-up. I expect the same factors that differentiate between responders/non-responders will be associated with LTFU.

A: In the second paragraph of the result section, we now provide information about non-responders, responders and those who did not complete the questionnaire at each time point. We described in the second paragraph of the result section that younger patients and those who report to be fully recovered, were most likely to be LTFU. In addition, in the method section, we added the following information: ‘In the BIOS, participants who did not complete a questionnaire were not excluded from the study but they were still invited at the subsequent time points’.

Only summarise data the data that is reported in tables in the text. It is not necessary to fully report the statistics again.

A: In the result section, we deleted the description of the statistics that were used and we more clearly summarised the most important study findings.

Correct the wording in some phrases to reduce ambiguity. For instance in line 212 change “more often” to “higher proportion” rather than given that the ICU admission was not a frequency WITHIN patients, but between patients. On line 229 change the wording of “most problems” as the data show the proportion of the sample with problems, not what they had the "most problems" with. On line 251 replace “symptoms” with “clinically significant symptoms" given that this is focused on those with symptoms above 11, not any symptoms.

A: We thank the editor for this specific feedback. We changed the words as suggested above.

Ensure that the terms confounding and covariates are used appropriately.

A: We have checked this in the manuscript and we did not find inappropriate use of the terms confounding and covariates.

In Table 4 provide columns showing the cell sizes for each level of each factor.

A: We provided the cell sizes for each level of each factor in Table 4.

There is only very limited discussion of the implications of the research. What new insights does this study add that can be logically and easily included in patient care to improve injury outcomes/reduce injury burden?

A: We thank the editor for this valuable comment. We extended the discussion with implications for health-care givers.

Note the level of missing data that were imputed was as high as 63.6% for some variables in the limitations.

A: Not all missing values were imputed; only missing sumscores due to missing items were imputed. This resulted in maximal: 13.9% imputed values per variable. We made a mistake by reporting that 63.6% of the data was imputed for some variables. In the discussion, we mentioned the high number of missing data. This was especially the case early post-trauma.

Reviewer 1:

There is significant editing that is required to improve the readability of the manuscript. Perhaps someone independent from the research team could complete a thorough proof-read? There are a number of sentences that need edits in order to make sense. For example, in the abstract, the sentence: “At long-term, higher age, comorbidities, longer hospital stay, lower extremity fracture and spine injury showed lower HS” needs to be modified to read: “At long-term follow-up, participants of higher age, with comorbidities, longer hospital stay, lower extremity fracture and spine injury showed lower HS”. Sentences that begin with ‘besides’ and ‘apparently’ should be removed from the manuscript. The EQ-5D is comprised of five dimensions (as opposed to compressed). There are instances of tense changes in the manuscript also.

A: We changed the specific grammatical errors in the sentences as supposed by reviewer. To further improve the readability of the manuscript, correction of grammatical errors and English improvement were made by a native English-speaking person.

If possible, the authors should examine the relationship between post-traumatic stress (which was found to be high in the study population) and health status over time. If post-traumatic stress is associated with increased risk of low health status, an important implication is that post-traumatic stress can be addressed early after injury to improve long-term outcome.

A: We thank the reviewer for this comment. In the result section, we now described that patients with symptoms of post-traumatic stress reported worse HS compared to patients who did not report symptoms of post-traumatic stress (see also S5 figure for more detailed information). In the conclusion section, we described that symptoms of post-traumatic stress is a determinant of worse HS post-trauma.

It would also be good if the authors could justify why they have focused on examining predictors of health status at 24 months but not prognostic factors for anxiety, depression, or post-traumatic stress. It would be interesting to know whether there are different factors that are predictive of these outcomes when compared to health status. This information could be used to identify individuals who are at increased risk of poor psychological outcomes after trauma and to inform intervention development.

A:. In the result section, we now described that patients with symptoms of post-traumatic stress reported worse HS compared to patients who did not report symptoms of post-traumatic stress (see also Figure 3 for more detailed information). We also would like to refer the reviewer to a recent BIOS-study publication concerning psychological outcome post-trauma:

-de Munter L, Polinder S, Haagsma J.A, Kruithof N, van de Ree C.L.P, Steyerberg E.W, de Jongh M.A.C. Prevalence and prognostic factors for psychological distress after trauma. Archives of Physical Medicine and Rehabilitation. In Press (23 December 2019).

The authors should provide more of a rationale for including the larger sample of participants who completed only health status assessments (and not the psychological outcome questionnaires) in the analyses for this manuscript, particularly because the focus of the study is on examining psychological outcome after trauma.

A: Patients who completed the short form questionnaire also completed the Impact of Event Scale. Indeed, those patients did not complete the Hospital Anxiety and Depression Scale. Therefore, we created Figure 3 which represents the course of HS of patients with symptoms of post-traumatic stress and the course of HS of patients without symptoms of post-traumatic stress.

The reason for some patients (those who did not complete a questionnaire up until 3 months post-trauma) completing a short version of the BIOS-questionnaire is not made explicit. Did this short questionnaire include a pre-injury health status assessment?

A: We thank the reviewer for this valuable comment. In the methods section, we now described the reason why patients were invited for the shortened questionnaire; ’Patients who completed the shortened questionnaire included those who stated that the BIOS-questionnaire was to comprehensive to complete or included those who were not traceable by phone and did not returned a BIOS-questionnaire’..We now also described in the methods section that the shortened questionnaire did not include a pre-injury assessment of HS.

Please describe how exactly comorbidities were assessed in the questionnaires. Furthermore, what demographic characteristics were collected? Collection of education level is described but other variables examined (such as sex/gender) are not clearly identified.

A: We thank the reviewer for this valuable comment. We now described, in the method section, how information regarding comorbidities was collected. Besides, we described in the method section that gender and date of birth were extracted from the self-reported questionnaires.

Although the authors describe their multiple imputation process, it is unclear how much missing data there was and how many participant scores needed to be imputed.

A: We added an extra appendix (S3 Table) to provide more insights into the missing data and the number and percentage of data that were imputed in this study.

Tables 2 and 3 could include a pre-injury health status column to aid interpretation.

A: In Table 2, we did not provide the pre-injury HS since we only collected the pre-injury HS as measured with the EQ-5D-3L. In Table 3, we now described the pre-injury outcomes.

A limitation of the study is the lack of pre-injury assessments of post-traumatic stress, depression, and anxiety.

A: We are aware of the added value of a pre-injury assessment of symptoms of post-traumatic stress, depression and anxiety. In the Brabant Injury Outcome Surveillance Study, participants had to complete a comprehensive set of questionnaires (almost 30 pages). Besides, it is not possible to include a pre-injury assessment of post-traumatic stress since we cannot refer to a certain incident that has not taken place yet. In the limitation section we now described: ‘We did not include a pre-injury assessment of symptoms of anxiety or depression.’

Should pre-injury health status also be examined as a predictor in linear mixed models?

A: Pre-injury HS was not examined as a predictor in the linear mixed models. We excluded pre-injury HS in this analysis since we had information regarding pre-injury HS in only 69% of the study population. In the exploratory analyses, we added pre-injury HS as a predictor in the linear mixed models and the variables which explains pre-injury status (such as age and comorbidity) had a small larger effect in the analyses without pre-injury HS.

Reviewer 2:

What does this study add to previous research on the subject? This could be made clearer in the Introduction (justify the need for this study) and the Discussion (what this study adds).

A: We thank the reviewer for this valuable comment. In the introduction, we described in more detail the specific need for this study. In the discussion, we added some implications for health-care givers, (see the ‘Implications for health care’ paragraph in the discussion section).

Acronyms need to be spelt out in the abstract. Please do not use acronyms that are not well-known to readers. It makes reading the article very arduous.

A: We thank the reviewer for this important and valuable comment. All the acronyms in the abstract are now spelt out.

‘We aimed to describe recovery patterns of health status (HS) and psychological outcomes during 24 months of follow-up and to identify subgroups at risk of both short and long-term health problems after trauma’. Authors could better align the presentation and interpretation of results (abstract and manuscript) to these objectives.

A: We thank the reviewer for this valuable comment. By providing more information about patients’ psychological functioning post-trauma and its relation to HS (see Fig. 3), we hope that the results better align to the study objective.

Reduction of trauma-related mortality in high-income countries [2] resulted in increased numbers of trauma survivors with long-term injury impact, including reduced health status (HS)’. Can authors provide a reference for the last part of this statement?

A: In the manuscript, we added two references behind this part of our statement.

Document needs extensive grammar/spelling correction

A: Correction of grammatical errors and English improvement were made by a native English-speaking person.

Study inclusion criteria are not quite clear to me. Did authors include all presentations to the ED with injury (as stated in the Methods) or just hospital admissions (as stated in the Abstract)? If the latter, did they only include patients with a principal diagnosis of injury? Please clarify.

A: We thank the reviewer for this valuable comment. In the methods, it was unclear that we only included hospitalized patients. We changed this in the method section.

As stated in the method section: all types of injuries were included, regardless of the intent or severity of the injury.

What proportion of the study population were elderly patients with an isolated simple fracture after a fall from their own height? These patients constitute a different population. Please comment.

A: One out of four of the included trauma patients in our study was aged 65 or older and had a hip fracture. Indeed, this is a large proportion of the study sample. However, to examine the total burden of trauma patients it is vital to not exclude certain subgroups.

In the discussion, we added the following text: ‘The inclusion of elderly with a hip fracture in the current study improves the generalizability of the study findings. We are aware that recovery patterns (due to comorbidities and functional decline) in elderly are different from recovery patterns in younger patients.’

Results – no need to repeat information given in tables. You can just highlight salient results.

A: In the result section, we deleted a lot of information that was already presented in the tables and figures.

‘Compared to the non-responders, participants were more severely injured, were more often admitted to the ICU and had a lower SES’. However, responders had a higher SES score. Please explain how SES status scores should be interpreted.

A: We thank the reviewer for this important and valuable comment. Indeed, we made a mistake by reporting that responders had a lower median SES/status score. In the result section, we now changed ‘lower SES’ into ‘higher status score’. We added the following sentence in the result section: ‘non-responders had higher status score compared to the general Dutch population (0.28) and compared to the non-responders’. As we already did in the method section, we now also described in the result section on how to interpret the status score (i.e. Status scores were based on the mean income, % of people with a low income, % of people with low educational level and % of unemployed people in the neighborhood).

Please provide definitions of mild, serious and severe TBI. These are usually referred to as mild, moderate and severe. Furthermore, as I understand it ‘type of injury’ refers to any injury rather than the most serious or isolated? For example a patient with three injuries including AIS severity score of 5 to the head, another injury of 2 to the head and a tibial fracture would be classed in ‘tibia, complex foot or femur fracture’, ‘mild TBI’ and ‘severe TBI’?

A: In the S1 Table, we provided more detail about diagnosis that were considered as mild TBI (AIS severity .1 and .2) or as serious/severe TBI (AIS severity. .3 or higher). Indeed as described in the method section (section prognostic factors), a patient with multiple injuries was classified in one or more injury of the 14 group classifications.

Dit helderder opschrijven in de methods

In Tables 2 and 3 please add Ns and a note in the legend on how to interpret the scales presented. Means in a healthy population would be useful addition. Why not add pre-injury data here? Also presenting mean differences (inter-patient) with regard to baseline would be interesting. I wonder whether presenting as a Figure would facilitate interpretation (as in Figure 1).

A: In Table 2, we did not provide the pre-injury HS since we only collected the pre-injury HS as measured with the EQ-5D-3L, and only for 69% of the study population. In Table 3, we provided the pre-injury scores. In Table 2 and 3, we added the numbers and a note in the legends on how to interpret the scales.

It would be useful to have a flow diagram describing numbers included in each phase of the study.

A: We included a more precise flow diagram of the study participants in the manuscript (Fig. 1).

Table 4 – Are colours intended to represent statistically significant results? If so, it is applied inconsistently. Furthermore, I would avoid bringing attention to results based on statistical significance only. In addition, given the problem of multiple comparisons, statistical significance of variables with more than two categories should be verified with a global test. Please indicate to readers (in the table legend) how to interpret the betas.

A: We deleted the colours in Table 4. Indeed due to multiple testing, change findings might have occurred for the results in Table 4. We described the risk of change findings in the limitation section of the discussion. Under Table 4, we shortly described how the Beta should be interpreted (i.e. measures how strong each predictor variable influences the dependent variable.).

Were any sensitivity analyses conducted other than patients with no missing data?

A: We did not perform any sensitivity analyses.

Please comment on the external validity of the results given less than 10% of the study population had major trauma.

A: Since we included both rural and urban areas and we included level I, II and III trauma centers, results of this study are at least generalizable to European countries. Nevertheless, we are aware that results may not be generalizable to Australian and USA trauma populations.

Can authors give any information on the validity of data in the trauma registry?

A; Accurate and consistent registration in the trauma registry is essential. In our study, we used data of the Brabant Trauma Registry, which is part of the Dutch National Trauma Registry (NTR). The registration collaborators of the NTR are well trained before they start to register. Besides, the researchers randomly controlled the data of the trauma registry. In the third paragraph of the method section, we now provide information concerning the validity of the data of the trauma registry.

It would be interesting to see results stratified for broad groups of injury: TBI, thoraco-abdo, spinal cord, orthopaedic, multi-system.

A: We thank the reviewer for this valuable comment. However, this manuscript contains already a lot of information. It is a good suggestion to stratify broad injury groups. In the future, our research group can examine the effect of stratification on non-fatal outcome after trauma.

The summary of key results in line with study objectives (first paragraph of the discussion) may be more effective if it were more succinct.

A: We made several changes in the first paragraph of the discussion section in order to provide a short overview of the most important results of our study.

Authors should take account of potential biases (e.g. underestimation of the frequency of health problems due to selection bias) when interpreting results in the Conclusion sections (abstract and manuscript).

A: We thank the reviewer for this comment. In the conclusion section, we described the high risk of selection bias in our study sample. Due to the restricted word count, we were not able to describe the high risk of selection bias in the abstract.

Reviewer 3:

The inclusion criteria are confusing. You say that “Adults (≥18 years) who visited an emergency department ≤48 hours after trauma were invited to participate” – but all patients included seem to have been admitted, “During the inclusion period of the BIOS, a total of 10,227 patients was hospitalized because of a trauma in one of the participating study centers”. Which group of patients was actually included?

A: See previous comment of reviewer 2 about the inclusion criteria of participants of the Brabant Injury Outcome Surveillance: In the methods, it was unclear that we only included hospitalized patients. We changed this in the method section.

You are using a number of scales/tools (EQ-5D-3L, HUI2/3, HADS-A/D, IES). It would be helpful to readers if you explained a little more detailed how the scores were calculated (EQ-5D-3L – “A summary score of these five dimensions (EQ-5D utility) can be calculated by using the Dutch tariffs [40].”). In addition, I suggest that you inform readers of the “normal values” in the Dutch population of the different scales, as done in the discussion line 336-337 for post-traumatic stress.

A: To improve the readability our manuscript, we added extra information about the used questionnaires in the method section. In the discussion, we added information about the Dutch norm scores of the EQ-5D-3L.

Socioeconomic status is not explained. How was the SES assessed, and how did respondents compare to the general population?

A: In the method section, we added the following sentence: ‘To determine socio-economic status, we used educational level’. Since we had could not collect educational level of the non-responders, we used status scores (based on mean income, % of people with a low income, % of people with low educational level and % of unemployed people in the neighbourhood, see also the first paragraph of the statistical analyses) to indicate SES. In the result section, we added the following sentence: Besides, non-responders had higher statuscores? compared to the general Dutch population (0.28) and compared to the non-responders.

The challenge with retrospective assessment of pre-injury health is of course that respondents may see their previous health in light of the present status. This is addressed appropriately in the limitations section.

A: We thank the reviewer for this compliment.

The tables are numerous and difficult to read when not deeply engaged in the study. I suggest that you elaborate on the legends (e.g. table 4: no explanation of the meaning of the colors). You could also exchange tables for figures. The reader when seeing a figure better perceives a longitudinal change in groups.

A: We removed the colours in Table 4. Furthermore, we tried to improve the readability of the tables and figures in our manuscript by adding extra information (e.g. by explaining what a Beta means and by providing the ranges of the questionnaires that were used in the Tables). We did not exchange tables for figures since we think that that may reduce the readability of the findings.

In short, an interesting study with need for clarification, and presently looking as you have been working hard and long time with data, and forgotten that potential readers does not have the same confidence with the material.

A: Hopefully, we improved the readability of the manuscript by the feedback from the editor and the three reviewers.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

10 Feb 2020

PONE-D-19-21191R1

Health status and psychological outcomes after trauma; a prospective multicenter cohort study

PLOS ONE

Dear Dr. Kruithof,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

We would appreciate receiving your revised manuscript by Mar 26 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Melita J. Giummarra

Academic Editor

PLOS ONE

Additional Editor Comments (if provided):

I have a few minor comments that should be addressed, in addition to those raised by the reviewers below, in your revision:

The additional explanation of SES needs to describe whether higher or lower scores indicate greater disadvantage to avoid potential mis-interpretation.

The axes on Figure 5 are not clear. Which data do the titles on the left and right-hand x axes correspond to? Moreover, are these truly demonstrating "differences" or mean values? Ideally it is a good idea to depict variability (e.g., sd, se or 95%CI around the mean).

Supplementary Figure 5 does not appear to have been uploaded.

The explanation of how to interpret a beta score is not sufficient. I suggest you follow the recommendation of Reviewer 2 in providing a clearer explanation of the beta value.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I would like to thank the authors for their efforts to improve the manuscript. The limitations associated with the study have now been identified and the implications of the findings are covered in more detail. There is still some language editing required which can be done at the proofing stage if the manuscript is accepted by the journal. One sentence that is important to address in the abstract is: ‘We aimed to describe the recovery patterns of health status (HS) and psychological outcomes during 24 months of follow-up and to identify subgroups at risk of both short and long-term HS after trauma’. The authors should make it clear that the focus is on identification of sub-groups at risk of poor health status.

Reviewer #2: I thank the authors for their thorough responses that address most of my comments. Remaining points are described below.

• The manuscript still requires significant editing for grammatical and spelling errors. For example, in line 71, no comma is required; line 106 should read ‘compiles’ and not ‘complies’; line 140 should read ‘comprise’ and not ‘comproise’. Errors are too extensive to list them all. Spelling/gramma correction software should help (not just the Word tool).

• Line 108. Please explain ‘randomly controlled the data’ in the data verification process more detail. All patient files or a sample? Did you check completeness, consistency, coherence and/or chronology?

• Line 240. Spell out FU. Again, using too many acronyms that are not widely recognized makes reading very arduous

• Table 4 ‘Beta: measures how strong each predictor variable influences the dependent variable’ could be replaced by ‘mean increase in EQ-5D-3L score (improvement in quality of life) compared to the reference category.

Reviewer #3: Thanks for comprehensive responses to the points raised by the reviewers. In general, my concerns have been addressed.

I am still a little confused by the status-scores. Some of the confidence intervals contains negative figures. What are the extremes of the scale? As this seems to be a Dutch scale, you could explain this better to international readers.

In the study design and participants section it is stated: “The Brabant Trauma Registry (BTR) complies pre-hospital and hospital data of all trauma patients admitted after presentation to the ED in the Noord-Brabant region.” You probably intended to say compiles? Same paragraph: “Before the data of the BTR and data of the BIOS-study were merged, the researchers randomly controlled the data of the trauma registry.” This is commendable, but what was “randomly”? 1 of 100, 1 of 1000, 20%? Please elaborate, otherwise this statement is not very descriptive.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Amy Richardson

Reviewer #2: Yes: Lynne Moore

Reviewer #3: Yes: Torben Wisborg

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

20 Mar 2020

Response to comments

Additional Editor Comments

-The additional explanation of SES needs to describe whether higher or lower scores indicate greater disadvantage to avoid potential mis-interpretation.

A: In the ‘statistical analyses’ paragraph, we described the following sentence: a lower status score indicates a lower SES whereas a higher status score indicates a higher SES. Furthermore, we described in the discussion and conclusion paragraphs that a lower status scores is an indicator of a lower SES-level (lines 425 and 473).

-The axes on Figure 5 are not clear. Which data do the titles on the left and right-hand x axes correspond to? Moreover, are these truly demonstrating "differences" or mean values? Ideally it is a good idea to depict variability (e.g., sd, se or 95%CI around the mean).

A: We changed Figure 3 as supposed by the editor. Furthermore, we made the description of the y-axis more clear.

-Supplementary Figure 5 does not appear to have been uploaded.

A: In this manuscript, there is no supplementary Figure 5. In the text, we did not refer to Supplementary Figure 5.

-The explanation of how to interpret a beta score is not sufficient. I suggest you follow the recommendation of Reviewer 2 in providing a clearer explanation of the beta value.

A: See the comment of reviewer 2: we have changed the sentence as supposed by the reviewer.

Comments reviewer 1

-There is still some language editing required which can be done at the proofing stage if the manuscript is accepted by the journal.

A: We did made several improvements in language editing throughout the manuscript.

-One sentence that is important to address in the abstract is: ‘We aimed to describe the recovery patterns of health status (HS) and psychological outcomes during 24 months of follow-up and to identify subgroups at risk of both short and long-term HS after trauma’. The authors should make it clear that the focus is on identification of sub-groups at risk of poor health status.

A: In the abstract and manuscript, we now described: ‘We aimed to 1) identify subgroups at risk of both short and long-term health status (HS) after trauma and 2) to describe the recovery patterns of HS and psychological outcomes during 24 months of follow-up.’

Comments reviewer 2

-The manuscript still requires significant editing for grammatical and spelling errors. For example, in line 71, no comma is required; line 106 should read ‘compiles’ and not ‘complies’; line 140 should read ‘comprise’ and not ‘comproise’. Errors are too extensive to list them all. Spelling/gramma correction software should help (not just the Word tool).

A: We have deleted the comma in line 71, we have changed the word ‘complies’ into ‘compiles’ and changed the word ‘comproise’ into ‘comprise’. Furthermore, we made several improvements in language editing throughout the manuscript.

-Line 108. Please explain ‘randomly controlled the data’ in the data verification process more detail. All patient files or a sample? Did you check completeness, consistency, coherence and/or chronology?

A: We now described the following sentence: ‘Quality of the data of the BTR and BIOS was checked on outliers and completeness by a trauma coordinator and researcher respectively. Furthermore, data from a sample of the trauma registry was checked manually by a trauma surgeon.’

-Line 240. Spell out FU. Again, using too many acronyms that are not widely recognized makes reading very arduous

A: We have spelled out the FU as ‘follow-up’ as suggested by the reviewer.

-Table 4 ‘Beta: measures how strong each predictor variable influences the dependent variable’ could be replaced by ‘mean increase in EQ-5D-3L score (improvement in quality of life) compared to the reference category.

A: We thank the reviewer for this suggestion. We changed the sentence as supposed by the reviewer.

Comments reviewer 3

-I am still a little confused by the status-scores.

A: In the ‘statistical analyses’ paragraph, we described the following sentence: a lower status score indicates a lower SES whereas a higher status score indicates a higher SES. Hopefully, this will lead to a better interpretation of the status score. Furthermore, we described in the discussion and conclusion paragraphs that a lower status scores is an indicator of a lower SES-level (lines 425 and 473). To interpret the results, we added the minimum and maximum scores.

-Some of the confidence intervals contains negative figures. What are the extremes of the scale? As this seems to be a Dutch scale, you could explain this better to international readers.

A: We added the minimum and maximum score of the EQ-5D-3L under Table 4.

-In the study design and participants section it is stated: “The Brabant Trauma Registry (BTR) complies pre-hospital and hospital data of all trauma patients admitted after presentation to the ED in the Noord-Brabant region.” You probably intended to say compiles?

A: Indeed, we intented to say ‘compile’ instead of ‘complies’. We changed the word ‘complies’ into ‘compiles’.

Same paragraph: “Before the data of the BTR and data of the BIOS-study were merged, the researchers randomly controlled the data of the trauma registry.” This is commendable, but what was “randomly”? 1 of 100, 1 of 1000, 20%? Please elaborate, otherwise this statement is not very descriptive.

A: We deleted the following text: ‘Before the data of the BTR and data of the BIOS-study were merged, the researchers randomly controlled the data of the trauma registry’. We replaced this text with the following sentence: ‘Quality of the data of the BTR and BIOS was checked on outliers and completeness by a trauma coordinator and researcher respectively. Furthermore, data from a sample of the trauma registry was checked manually by a trauma surgeon.’

Submitted filename: Response to Reviewers_V2_08032020(2).docx

Click here for additional data file.

26 Mar 2020

PONE-D-19-21191R2

Health status and psychological outcomes after trauma; a prospective multicenter cohort study

PLOS ONE

Dear Dr. Kruithof,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

I have uploaded a marked copy of your most recent revision highlighting the relatively minor points that still require further clarification or correction before I can make a decision about the suitability of your submission for publication.

We would appreciate receiving your revised manuscript by May 10 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Melita J. Giummarra

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: PONE-D-19-21191_R2_Editor comments.pdf

Click here for additional data file.

28 Mar 2020

Response to comments

-In Fig. 3, we changed the word ‘’sumscore’’ into ‘’summary score’’, as suggested by the editor.

-In the abstract and in the introduction, we changed the description of the aim of the study, as suggested by the editor.

-In line 111, we added the following information on how data was checked: ‘Data of the BTR (i.e. Brabant Trauma Registry) was checked as part of routine practice while the data of the BIOS was checked randomly by the researchers.’

-In line 140, we removed the question mark as supposed by the editor.

-In one of the previous versions of our submitted manuscript, we referred to S5. This was not correct, we should have referred to Fig 3 instead of S5. In the current manuscript, we now correctly refer to Fig 3.

-In line 398, we removed the ‘’s’’ as supposed by the editor.

-In line 457, we removed the ‘’n’’ as supposed by the editor.

Submitted filename: Rebuttal letter_BIOS_V3_27032020.docx

Click here for additional data file.

30 Mar 2020

Health status and psychological outcomes after trauma; a prospective multicenter cohort study

PONE-D-19-21191R3

Dear Dr. Kruithof,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Melita J. Giummarra

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

2 Apr 2020

PONE-D-19-21191R3

Health status and psychological outcomes after trauma; a prospective multicenter cohort study

Dear Dr. Kruithof:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Melita J. Giummarra

Academic Editor

PLOS ONE