Article Text

Download PDFPDF
Impact of comorbidity on health outcome after a transport-related injury
  1. Stella Samoborec1,
  2. Pamela Simpson1,
  3. Behrooz Hassani-Mahmooei2,
  4. Rasa Ruseckaite1,
  5. Melita Giummarra1,
  6. Darshini Ayton1,
  7. Sue Evans1
  1. 1 Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria, Australia
  2. 2 Strategic Intelligence and Insights Unit, Monash University, Clayton, Victoria, Australia
  1. Correspondence to Stella Samoborec, Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, VIC 3004, Australia; Stella.samoborec{at}monash.edu

Abstract

Introduction Understanding the impact of comorbidity on health outcomes is important given that comorbidities can affect survival, morbidity, service delivery costs and healthcare utilisation. However, little is known about the types of comorbidities affecting specific health outcomes after minor to moderate road trauma.

Methods This study involved 1574 participants who claimed injury compensation following transport-related injury. Cross sectional data were collected. Health outcomes were assessed using the EQ-5D-3L specific domains and summary score. Twelve self-reported pre-existing chronic conditions were assessed using a multivariate logistic regression, adjusting for demographic and injury characteristics.

Results Out of 1574 participants, only 17 (1%) participants reported no pre-existing comorbidities, 72% reported one, 13% reported two and 14% reported three or more comorbidities. Hypertension (15%), depression (14%) and anxiety (14%) were the most commonly reported comorbidities, followed by arthritis (13%), chronic pain (11%) and asthma (11%). Participants with a history of arthritis (adjusted odds ratio [AOR] 1.90, 95% CI 1.24 to 2.91); chronic back pain (AOR 1.59, 95% CI, 1.04 to 2.43); other chronic pain (AOR 2.73, 95% CI 1.42 to 4.24); depression (AOR 2.55, 95% CI 1.60 to 4.05) and anxiety (AOR 2.08, 95% CI 1.32 to 3.26) were at increased risk of poorer health outcomes, after controlling for age, gender, type of injury and time since injury.

Conclusion This study found that comorbidities such as arthritis, chronic back pain, other chronic pain, depression and anxiety significantly increase the odds of poorer health postinjury, regardless of the time since injury. Regular screening of comorbid conditions may help identify people likely to have poorer outcomes, thereby enabling the implementation of interventions to optimise health despite the presence of comorbidities.

  • injury compensation
  • outcome of injury
  • quality of life
  • comorbidities

Statistics from Altmetric.com

Introduction

Traumatic injuries, such as those arising from road trauma, are a global public health problem. According to the WHO, deaths from road transport injuries account for approximately 25% of all injury deaths, and by 2020 road trauma will become the third leading global cause of disability.1 2 Given the high burden of disability following road trauma, it is important to understand the overall impact of injury on health outcomes.3 In addition, it is crucial to understand the profile of people who are at risk of poorer health outcomes following road trauma (ie, poor health outcomes defined using the standards of the generic measure of health and WHO’s definition of health). One of the potential key factors that may affect health following road trauma are pre-existing comorbid conditions and pretrauma health state.

The Australian Burden of Disease study acknowledged the importance of coexisting conditions in estimating the overall burden of injury;4 a finding supported by observational cohort studies.5–8 Early recognition and identification of comorbidities may enable health practitioners to define strategies to prevent further negative impact on patients’ outcomes. Understanding the role of pre-existing health conditions on health outcomes is essential as it may help to identify patients at high risk for poorer outcomes and contribute to evaluation and improvement of current interventions. It may also assist public health professionals to prioritise interventions for health promotion and disability prevention.

Comorbidities, diseases or illnesses that coexist with an index injury9 affect survival10 and morbidity11 12 after major trauma and impact on service delivery costs8 and levels of healthcare utilisation.13 Pre-existing comorbidities have been shown to be a predictor of poor health outcome after non-catastrophic transport injuries.14 Several studies strongly recommend screening and assessment of preinjury health state when predicting outcomes after a road injury.15–18

Commonly used tools to obtain and evaluate information of the coexisting diseases include the Charlson-Comorbidity Index (a tool that is used to predict risk of mortality and includes conditions such as metastatic tumour, AIDS or diabetes mellitus)19 and Aggregated Diagnosis Groups.20 Self-reporting of pre-existing diagnoses is another common approach to collecting pre-existing health status.21–24 Currently, there is no gold standard for assessing comorbidity.25 Over the last two decades, there has been considerable investment in assessing the value of self-reported data in comparison to other sources, such as deriving health conditions from prescribing or pharmacy databases26 and medical records.24 A recent study found that patients with cancer reported similar comorbidity information to information extracted from the medical record, suggesting that use of patients’ reports is a reasonable approach for observational research.27 Self-reported comorbidities are therefore frequently used in outcome prediction studies.28 29

Assessing comorbidity in the early stage of recovery is important because of its potential influence on recovery trajectory.30 31 While previous research has identified that the number of comorbidities affect outcomes,29 it remains unclear as to the impact of specific comorbid disorders, such as arthritis and depression, on health outcomes after traffic accidents.

The primary objective of this study was to assess the prevalence of self-reported comorbid conditions in a cohort of compensation claimants following minor-moderate road trauma. The secondary objective was to examine the association between type and number of comorbidities and specific health outcomes.

Methods and materials

Victorian transport-injury compensation system

In the state of Victoria, Australia, people injured in land or rail-based transport accidents are eligible to claim compensation for treatment, income replacement, rehabilitation and long-term support services via the Transport Accident Commission (TAC), regardless of fault. The TAC is the statutory insurer of third-party personal liability for transport accidents.32

Data source

The TAC has conducted an annual cross-sectional Client Outcomes Survey (COS) since 2009 to assess health, vocational and claimant experience outcomes in clients without catastrophic injury, who are managed in the scheme’s Recovery Division. The survey includes standardised measures of vocational and health status prior to injury, current vocational status, current physical and mental health status, persistent pain, mobility and functional independence, access to and satisfaction with healthcare and satisfaction with the TAC service. Data were collected via computer automated telephone interview (CATI) by a third-party social research organisation, and took approximately 30 min per interview.

Study participants in the COS survey

Participants in the COS comprised randomly selected claimants with non-catastrophic injuries (ie, minor, moderate and severe injuries, but no severe brain injury or complete spinal cord injury), aged 16–89 years with claim duration of at least 5 months. During the 2015–2016 and 2016–2017 surveys, 2210 (67.3%) claimants participated in the COS. Participants were ineligible for inclusion in the COS survey if they had claims involving work-related injuries; they were current TAC staff or if they were previously identified to have behavioural ‘risks’ (ie, demonstrated anger, violence, abusive language and drug abuse identified through prior communication).

Specific inclusion criteria for this study

Participants were eligible for inclusion in this study if they were 18 years of age or older and if they had sustained minor (eg, whiplash, neck and back sprains and strains, laceration, abrasions, contusions) or moderate injuries (fractures and dislocations).

Study design

This cross-sectional study involved a retrospective analysis of deidentified COS data linked with some additional claimant demographic and injury data.

Study variables

Sociodemographic and injury details

Sociodemographic and injury data were extracted from the claimant records, and comorbidities, injury details and health outcomes were assessed with the EQ-5D-3L as part of the COS survey. Demographic details collected at the time of the survey included: age, gender, education level, language, household structure, employment status at the time of the accident, return to work status and time since injury. Injury-related characteristics included minor and moderate injuries, admission to hospital and length of hospital stay. Injury severity was defined according to the TAC’s hierarchical classification of injuries; the hierarchy is evaluated by finding the most serious injury on the claim. At the TAC, injuries are coded in the claims system using the Systematized Nomenclature of Medicine (SNOMED) classification whereas International Classification of Diseases 10th edition Australian Modification (ICD 10 AM) codes are available only for hospitalised claimants.

Pre-existing self-reported comorbidities

To record comorbid conditions methods consistent with the National Health Survey by the Australian Bureau of Statistics were used.33 That is, participants were asked:

I would also like to ask you about any long term health conditions that you may have had at the time of the accident, which had been confirmed by a doctor or nurse. That is a condition that had lasted for 6 months or longer. It may have been controlled through treatment or may not have affected you all the time.

Before your injury did you have any of the following? (eg, Asthma, arthritis, stroke, diabetes, hypertension, heart condition, cancer, chronic back pain, other chronic pain, depression, anxiety).

Individual comorbidities were dummy coded, and the total comorbidity count was calculated per participant.

Outcome measures

Health outcomes were measured using the domains of the EQ-5D-3L health-related quality of life measure. The EQ-5D-3L is a standardised health utility instrument that measures five dimensions of health-related quality of life: mobility, self-care, usual activities, pain/discomfort and anxiety/depression.34 Both the EQ-5D-3L dimensions and summary score were used as outcomes in this study. The summary index scores were calculated using age and gender population weights and the UK norms.35 It ranges from −0.594 to 1 with a score of 0 representing a health state equivalent to death and 1 representing perfect health. The EQ-5D-3L dimensions were measured in three levels: no problem, moderate problem or severe problem in descriptive analysis. The EQ-5D-3L dimensions were dichotomised into ‘no problems’ (ie, level 1) and ‘problems’ (levels 2 and 3) consistent with other studies,36 and the EQ-5D-3L summary score was dichotomised into ‘poor health’ (ie, poor and fair category 0–0.70 summary score) and ‘good health’ (ie, very good and good category 0.71–1.00 summary score) in accordance with the EQ-5D-3L user guide.34

Statistical analysis

Baseline characteristics of the study sample were presented using descriptive statistics. To assess the relationship between number of comorbidities at different time points after injury, participants were divided into three main groups (ie, one comorbidity, two comorbidities and ≥three comorbidities) and the mean health summary score was compared across the comorbidity groups separately for claimants at different time points postinjury (ie, 6–12 months, 13–24 months, 25–36 months and ≥36 months postinjury). This was done using the Mann Whitney test for continuous data because it was not normally distributed and χ² statistics were used for categorical variables, such as age group and gender.

Logistic regression was used to determine the relationship between comorbid conditions and EQ-5D-3L domains and EQ-5D-3L summary index score. The assumptions for performing binary logistic regression analysis were satisfied (eg, large sample size, no multicollinearity among the independent variables and a binary depended variable).

The relationship between each comorbidity and each health outcome was determined using univariate logistic regression; comorbidities that were not statistically significant (p>0.10) were not added to the multivariate model. Comorbidities demonstrating a statistically significant association (p<0.10) with health outcomes (ie, health summary index score and domain specific scores) were concurrently added to a multivariate logistic regression model that adjusted for common confounders such as age, gender, time since injury, and type and severity of injury. The analyses were performed using STATA V.15.0 (Stata, College Station, Texas, USA).

Results

Study sample characteristics

Detailed characteristics of the cohort are presented in table 1. This study included 1574 participants aged between 18 and 89 years, with 22% of participants surveyed less than 6 months postinjury, 32% surveyed 6–12 months postinjury, 22% surveyed 13–24 months postinjury, 12% surveyed 25–36 months postinjury, and 12% at 36–48 months postinjury. There was a higher proportion of males (56%) than females and a mean age of 44 years (SD=16.6). The biggest road user group were drivers (46%); people who had sustained injuries that were classified as minor (56%) or moderate (44%) and just over half had a tertiary education qualification (52%), including vocational training/diplomas. There were 946 people who were hospitalised (60%) with a length of hospital stay ranging from one to 182 days, with a mean length of stay in hospital of 4.1 days (SD=12.4). The mean EQ-5D-3L summary score for the sample was 0.71 (SD=0.25).

Table 1

Study sample characteristics

Prevalence of comorbidities

In this sample (n=1574), 17 (1%) participants reported no previous comorbidities, 72% reported one, 13% reported two and 14% reported three or more comorbidities. Hypertension (15%), depression (14%) and anxiety (14%) were the most commonly reported comorbidities, followed by arthritis (13%), chronic pain (11%) and asthma (11%). More than half of the participants (55%) reported having another long-term condition that was not defined in the survey (figure 1).

Figure 1

Prevalence of comorbidities (n=1574).

Females were more likely to report more than one comorbidity compared with males (p<0.001) and were more likely to report arthritis (p<0.001), chronic back pain (p=0.005), other chronic pain (p<0.001), depression (p<0.001) and anxiety (p<0.001).

Most participants who sustained a moderate injury (ie, fracture and/or dislocation) reported only one comorbidity (n=533, 34%). Participants with minor injuries (ie, soft-tissue, neck and back sprain, strain, whiplash, abrasion, laceration, protruding disc with no surgery or contusion) were more likely to report more than one comorbidity than those with moderate injuries (n= 269 vs 151, p<0.001).

Association between comorbidities and health outcome

A higher proportion of people with pre-existing arthritis (18%), hypertension (24%), chronic back pain (16%), other chronic pain (15%), depression (26%) and other long-term condition (55%) reported poor and fair health states than good and very good health states postinjury (figure 2).

Figure 2

Distribution of participants by comorbidities and the EQ-5D-3L health status category.

There was a relationship between the number of comorbidities reported at the time of the injury and health scores recorded as assessed by the EQ-5D-3L score (figure 3). For participants interviewed at 36 months postinjury, the mean EQ-5D-3L score was lower than that reported by people interviewed up to 12 months postinjury, regardless of the number of comorbidities reported at baseline (one comorbidity n=214, mean=0.63, SD=0.29; two comorbidities n=26, mean=0.64, SD=0.17; three or more comorbidities n=36, mean=0.59, SD=0.21). Participants interviewed within the first year postinjury who reported three or more comorbidities (n=57) had significantly lower mean EQ-5D-3L scores compared with those with only one comorbidity (n=281), mean=0.60, SD=0.24 vs mean=0.77, SD=0.23; p=0.011, respectively).

Figure 3

Relationship between number of comorbidities and health outcome (EQ-5D-3L mean score) at different time points since the injury.

Table 2 outlines the association between comorbidities and health outcome at different periods following the road trauma. Anxiety was significantly associated with poor health outcomes at each of the four periods. Comorbidities associated with poor health outcomes at 6–12 months post-trauma were arthritis, stroke, chronic back pain, other chronic pain, diabetes and depression. At 13–24 months post-trauma, comorbidities associated with poor health outcomes included arthritis, stroke, chronic back pain, depression and anxiety. Having a pre-existing heart condition was associated with poor longer-term outcomes only (i.e. >36 post injury).

Table 2

The association of comorbidities on health outcomes (EQ-5D-3L summary score—good vs poor health) at different time points postinjury

Table 3 describes the extent to which comorbidities were associated with specific EQ-5D-3L domains. Pre-existing depression and anxiety were associated with reporting problems with anxiety and depression on the EQ-5D-3L instrument. Pre-existing chronic pain and depression were associated with problems in each of the health domains. Comorbidities with significantly higher odds (p<0.05) of overall poorer health outcome (ie, summary index score) were chronic pain (5.35, 95% CI 2.90 to 9.89), depression (OR 4.40, 95% CI 2.96 to 6.54) and anxiety (OR 3.67, 95% CI 2.51 to 5.36).

Table 3

Univariate analysis of the relationship between self-reported comorbidities and poorer health outcomes described as mobility, ability to perform personal care, ability to perform usual activities, pain, depression and anxiety and EQ-5D-3L summary score

Table 4 provides results of the logistic multivariable analysis. In the multivariate model, older participants had higher odds of poor outcomes in the mobility domain. Participants in middle to older age groups (35–65 years of age) reported poorer ability to perform personal care and usual activities. For the overall health outcome (summary score), being middle aged (35-54) was associated with poorer outcomes. Gender and injury severity were not associated with the general health summary score in the multivariate model. Participants who were >36 months postinjury had worse health outcomes for all domains, and general health, compared with participants only 6–12 months postinjury.

Table 4

Multivariate analysis of factors and comorbidities associated with poorer health outcomes after a minor-moderate transport-related injury

Participants with a history of arthritis (AOR 1.90, 95% CI 1.24 to 2.91); chronic back pain (AOR 1.59, 95% CI 1.04 to 2.43) and other chronic pain (AOR 2.73, 95% CI 1.42 to 4.24); depression (AOR 2.55, 95% CI 1.60 to 4.05); and anxiety (AOR 2.08, 95% CI 1.32 to 3.26) were at significantly increased odds of poorer health summary score after controlling for age, gender, type of injury and time since injury.

Discussion

This study aimed to examine the prevalence of pre-existing comorbid conditions and their association with health outcomes after minor-moderate transport-related injuries. People who sustain injuries, in comparison to non-injured cohorts, are almost five times more likely to have one or more comorbidity.37 Prior studies have demonstrated that comorbidities have a negative impact on health and recovery after major trauma,7 10 yet these studies did not report which comorbidities were associated with specific health outcomes such as pain, depression/anxiety and mobility. In total, 99% of the participants in the present study reported having at least one chronic condition preinjury. We found that pre-existing arthritis, chronic back pain, depression and anxiety had an association with poorer health outcomes following minor-moderate road trauma.

Pre-existing arthritis, chronic back pain, depression and anxiety increased the odds of having a poor health outcome by more than twofold. Moreover, an increased number of comorbidities increased the odds of poorer health outcomes. However, while an increased number of comorbidities appeared to be associated with poorer health in people who were interviewed in the initial 3 years following the injury, the number of comorbid conditions had no association with health outcomes in people who were interviewed after 3 years.

People with moderate injuries were more likely to report only one comorbidity, whereas multiple comorbidities were seen in people with minor injuries. It could be that people with minor injuries and multiple comorbidities are more likely to lodge a compensation claim for a ‘minor’ injury, which could possibly explain the high rates of poor health outcomes found in minor injury cohorts across other observational studies.38–41 However, further studies using prospective population data are required in order to understand who lodges a compensation claim, and the subsequent causal association between comorbid conditions and poor health outcomes for people with minor compensable injuries.42 A study by Spearing et al explored the effect of reverse causality between compensation claim and health outcomes and demonstrated that some characteristics like previous neck pain actually influence whether someone will claim compensation (in fault-based setting), yet it is not known if this cohort differs from people who do not make a claim.42 There is lack of research examining comorbid conditions in people who do and do not claim compensation and it is likely that many other demographic and clinical factors have an impact on both the decision to lodge a claim, and subsequent health and recovery outcomes after transport-related injury.

Previous studies that have examined the association between preinjury factors and poor recovery outcomes reported that preinjury pain16 and preinjury mental health status15 were significant independent predictors of poor outcomes. Our study is consistent with the findings from previous studies and highlights that early screening for these factors will ensure that people receive appropriate treatment that will attenuate the impact of comorbid conditions on health after injury.

The impact of gender on recovery has been examined in a number of studies with conflicting findings. A systematic review published in 2001 found that female gender was a predictive factor of recovery after whiplash injury43 while another published in 2003 demonstrated no such effect.44 These differences have been attributed mostly to heterogeneity in study designs (eg, prospective vs retrospective).45

Our findings suggested that women may have worse health after injury because they have higher rates of preinjury comorbidities, particularly the comorbid conditions that had the greatest impact on health outcomes (ie, pain, depression and anxiety). A recent Victorian study reported that incidence of those treated for mental health conditions after transport-related injuries was higher in females.46 In our study, gender was not associated with poorer overall general health but it was associated with pain and depression or anxiety, which confirms that females may be in a greater risk of suffering pain and mental health issues following road trauma.

The main implication of the present findings is that screening for comorbid conditions, especially arthritis, chronic pain, depression and anxiety, should be introduced and regularly assessed to enable delivery of appropriate services in primary care, and via the compensation system. Given the high prevalence of comorbidities in the present cohort, as in other injury population studies,11 12 finding ways to support or tailor rehabilitation in the presence of comorbid conditions may facilitate better long-term health outcomes. A patient-centred care involving collaborative care approach could provide a foundation for delivery of interventions to improve outcomes, yet the extent to which this role falls to the compensation system requires further consideration. An ongoing trial in the USA may assist in informing the trauma and compensation systems nationwide on the effectiveness of the collaborative care treatment approaches for trauma survivors with mental health conditions such as PTSD and other comorbidities.47

This study has several strengths. First, the sample size was sufficiently large to enable a considerable comparison and analysis in different age, gender and injury severity groups. Second, the sample is a random selection of all patients claiming under Recovery Division at the TAC, therefore selection bias was minimalised. Third, there was no missing information detected on self-reported comorbidities and all participants provided responses on their preinjury health status. Last, although this study included cross-sectional survey data, all multivariate analyses adjusted for time since injury.

Some of the limitations of this analysis should be carefully considered.

First, the study population only included those who claimed for injury compensation in Victoria and therefore this population is not considered strictly representative of the general injured population. Consequently, the implications can only be considered in relative terms within the limitation of the sample.

Second, recall bias may affect the present findings, particularly for the 12% of respondents who were contacted more than 3 years after the injury and asked to comment on comorbidities present at the time of the injury. Thus, we were not able to gauge the extent of how accurately were the participants able to recall past exposures (ie, whether or not they had this comorbidity at the time of their injury).

Third, the comorbidity question has not been previously validated or compared with other validated comorbidity scales. However, as previously mentioned, self-reported comorbidities have been proven as reliable assessment in most studies22 27 48 and comorbidities were assessed using methods consistent with the National Health Survey standards for assessing pre-existing conditions.33 This gives some level of confidence in how the comorbidities were assessed.

Fourth, the participants were informed that their participation in the survey would not affect their claim; however, some may have not been honest about their pre-existing conditions for fear that their claim may be denied or limited. This may be due to suspicion that persistent symptoms or incapacity to work were related to previous health condition and not to the road trauma. Nonetheless, all but 1% of respondents reported at least one pre-existing condition.

The high prevalence of comorbid conditions detected in this study might have been influenced by the, as already mentioned, homogenous ‘compensation’ sample; however, we do not have data on a transport-injured population who did not claim compensation. Future studies should explore whether these findings extend to those who do not claim compensation. Given the high prevalence of other long-term conditions in this cohort (55%), it would be interesting to understand these other conditions in greater detail. However, despite their prevalence, they had no significant impact of health outcomes at any time point, except that a higher proportion of people reporting ‘other conditions’ had better health outcome (summary score).

In summary, there is now substantial evidence that health outcomes following road trauma are affected by several pre-existing comorbidities.7 8 10–12 29 Participants with comorbidities such as arthritis, chronic back pain or other chronic pain, depression and anxiety were at higher risk of poorer health postinjury, regardless of the time since injury. Early identification of these comorbidities may provide a good indication of high-risk groups requiring enhanced support and amplified interventions. Such a tailored rehabilitation management, based on specific clinical needs and using patient-centred care principles, might accelerate better outcomes for those who have sustained minor or moderate transport-related injuries.

What is already known on the subject

  • It is known that recovery after transport-related injury is multilayered and complex. Nevertheless, recovery outcome does not purely depend on the severity of injury or accident.

  • Patients with medically classified minor-moderate injuries also suffer long-term physical and mental disabilities and the reasons are multidimensional.

What this study adds

  • Recovery from a minor transport-related injury is complex, challenging and multifaceted phenomenon. It involves various biological, psychological and social elements. One of the crucial factors to consider in patients’ recovery is patients’ preinjury health status.

  • This study clearly demonstrates that a number of pre-existing conditions are associated with poor health outcomes. Screening for these comorbidities should be introduced on a regular basis, yet further studies are needed to demonstrate the causal effect of these conditions, as it is not known how these conditions affect recovery trajectory.

Acknowledgments

We would like to thank the funders and the Steering Committee members. The authors gratefully acknowledge the technical advice and support from the TAC’s representatives. Finally, we express our gratitude to the individuals who participated in the study.

References

Footnotes

  • Contributors SS and SE conceptualised the study. SS conducted the statistical analysis. PS and BHM provided assistance and advice on data analysis. SS drafted the manuscript. DA, RR, PS, SE and BHM reviewed and revised the manuscript. MG provided critical revision of the manuscript.

  • Funding SS, Monash ID 26381494, has received Capital Markets Cooperative Research Centre living allowance scholarship for conducting this study.

  • Disclaimer The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Ethics approval Monash University Human Research Ethics Committee (MUHREC Project Number Approval 2016 1044 760) approved the study.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Data are available on reasonable request.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.