Predictors of seeking financial compensation following motor vehicle trauma: inception cohort with moderate to severe musculoskeletal injuries

Study design

Participants were recruited for this inception cohort study from two trauma hospitals (Liverpool and St George) in Sydney, NSW, Australia from November 2007 to February 2011. The selected hospitals provided a representative sample of motor vehicle related orthopaedic trauma in NSW that required inpatient hospitalisation in NSW. Participants were followed up at 6, 12 and 24 months post injury.

Eligible patients identified via a hospital trauma database of orthopaedic admissions were invited to participate. Informed consent was obtained. Inclusion criteria were: admission to hospital within 2 weeks of injury; involvement in a motor vehicle crash; age 18 years or over; and an upper or lower extremity fracture (humerus, radius, ulna, pelvis, acetabulum, femur, patella, tibia, fibula, talus, calcaneus). An English speaking family member was used to interview eligible patients from Culturally and Linguistically Diverse (CALD) backgrounds. Patients were excluded if they had: dementia or a significant pre-existing cognitive impairment preventing the ability to consent; spinal cord injury; Glasgow Coma Score (GCS) less than 12 on admission; amputation of a limb; or isolated phalangeal, carpal, metacarpal, tarsal or metatarsal fractures.

There were 32 variables measured for each participant. Calculations for the sample size of 450 were based on an allowance of 10 participants per variable [25] and accommodated a 25% loss to follow up with reference to similar published research [8, 26].

At six, 12 and 24 months after injury follow up questionnaires were posted to participants. By 3 weeks, if there was no response participants were contacted up to six times by telephone. Questionnaires could be completed by telephone or by mail. Participants were removed from the study if non-contactable or they declined to participate. Additional information about the study methodology is published in two separate papers investigating injury recovery and return to work in the same cohort [27, 28].

Baseline data were collected in hospital within 2 weeks post injury by written questionnaire. Demographic data including date of birth, age, gender, and injury related information were obtained from hospital records and a trauma database. Selected study factors were based on relevant research and referred to the aims and objectives of the study [6, 7, 29, 30].

Setting

In NSW at the time of the study, the Motor Accidents Authority (MAA) was the government insurance regulator of the CTP personal injury scheme, and is a privately underwritten modified common law scheme. WorkCover was the government insurance regulator of the WC scheme, and is a publically underwritten statutory benefit scheme where private insurers manage claims on behalf of WorkCover [31, 32]. In 2015, the scheme regulators amalgamated and formed the State Insurance Regulatory Authority (SIRA).

All motor vehicles travelling on public roads must be registered. To make a CTP claim a motor vehicle must also be registered and the claim is made against the driver at fault. From April 2010, anyone injured in a motor vehicle crash (regardless of fault) can access limited entitlements, that is: medical expenses and lost wages up to AUD (Australian Dollar) $5000. Before 2012, to make a WC claim a motor vehicle crash must have occurred during travel between place of employment and home, and/or any work-related place, and a person injured (regardless of fault) [31, 32].

For both schemes, a claim must be lodged within six months of injury and insurers have 3 months to determine final liability (accept/deny the claim). Provisional liability facilitates access to treatment by earlier payment of medical expenses and for WC weekly wage benefits. In WC, within 48 h insurers must be informed of an injury [32]. Entitlements include past and future losses: for example, medical expenses, loss of income, and pain and suffering/impairment. In both schemes, payments for medical expenses are made as incurred. In CTP, loss of income (including past and future), future medical expenses, and permanent impairment are paid as a lump sum at claim settlement. In WC, loss of income is paid weekly and can be lifetime depending on the level of impairment. Future medical expenses can be lifetime, and permanent impairment is paid as a lump sum. In addition, legal representation can also be sought at any time for either scheme. [31, 32]

Injury related study factors

The Abbreviated Injury Scale (AIS) (1990 Revision, Update 98) was used to code all injuries [33]. The Injury Severity Score (ISS) and New Injury Severity Score (NISS) were calculated as measures of injury severity; these are considered indicators of potential mortality [34] and are the sum of the squares of the three highest AIS scores from different body regions (ISS) regardless of body region (NISS). The AIS ranks injuries to particular body regions on a scale from one to six (six is not survivable). Injuries were classified as minor–moderate (1-8), serious (9-15) or severe–critical (16-75) based on ISS/NISS scores [35].

Socio-demographic study factors

Socio-demographic factors included age, gender, marital status, occupation, and education. Occupation was measured using the Australian Standard Classification of Occupations (ASCO), classifications were divided by skill level (e.g. managers/professionals, tradespersons, intermediate clerical and elementary related) [28, 36]. Current work status (yes/no) was asked with additional variables for full/modified duties (e.g. lifting restrictions, reduced hours) and full-time (usually working at least 35 h per week) or part-time (usually working one hour to 35 h per week) [37].

Household income was measured exclusive and inclusive of household structure to cater for any differences in income distribution. To calculate an adjusted income (inclusive of household structure), household income was divided by the sum of points, 1 for the first person aged ≥15 years, 0.5 for each additional person aged ≥15 years, and 0.3 for each person aged <15 years [38, 39].

The Index of Relative Socioeconomic Disadvantage (IRSD) summarises economic and social conditions within a particular area/postcode such as employment, fluency in English and household size [38]. The lowest score is indicative of greatest or most socioeconomic disadvantage and the highest score indicates least disadvantage. It can be used as a continuous variable or divided into quintiles.

Health related study factors

Different health conditions measured as indicators of general health status at baseline were chronic illnesses – asthma, cancer, heart and circulatory conditions, diabetes, arthritis, osteoporosis, mental and behavioural problems, and neck and back problems/disorder/pain. The National Health Priority Areas initiative list these conditions as imposing high social and financial costs on Australian society) [40]. The classification was based on the Australian Bureau of Statistics (ABS) Health Survey which defines a long term condition as one which the patient currently has, and which has lasted or they expect to last for 6 months or more [39–41]. Body Mass Index (BMI) was calculated from the participant’s self-reported weight and height.

Other factors measured in accordance with the ABS Health Survey included: recent injuries (other than the motor vehicle crash) in the last 4 weeks that required medical intervention or were associated with a decrease in usual activities; medication use in the last 2 weeks for asthma, arthritis, osteoporosis, heart or circulatory conditions, diabetes, high sugar levels, mental wellbeing; and smoker status [39].

In previous research, associations were found between poor recovery and poor expectations for return to usual activities and work [29, 42–44]. There were few validated measures for self-efficacy and other similar constructs, therefore, two applicable measures from a large Canadian study of soft tissue injuries were used [43]. These were ‘do you expect to return to work (yes/no)’ and ‘when do you expect to return to usual activities’ (number of days).

A validated scale measured alcohol consumption with the first three questions of the Alcohol Use Disorders Identification Test: Self-Report Version (AUDIT-C) [45, 46]. The word ‘standard’ and ‘in the past year’ were added. Alcohol quantity was based on an Australian standard drink [47, 48]. The AUDIT-C questions measure number and frequency of drinking. The National Health and Medical Research Council (NHMRC) levels were used to assess risk of long term harm (alcohol related disease) and/or short term harm (alcohol related injury) due to alcohol consumption [47]. The risk levels for long term harm were low risk, risky, and high risk based on the number of standard drinks/week consumed over the past year. The risk level for short term harm was if ≥6 standard drinks were consumed on one occasion, on one day over the past year.

Because these measures did not match the AUDIT-C categories, to compare results an algorithm was calculated from the Bettering the Evaluation of Care and Health (BEACH) Survey, (Professor K Conigrave, personal communication March 19, 2007). Categories for other study factors are explained in the tables and our other published research from the same cohort that investigated predictors of injury recovery and return to work. [27, 28]

Compensation related measures

Most compensation related factors were not recorded at baseline because the questions were unanswerable (within 2 weeks of injury). At 6 months post injury the questions asked were: claim made (yes/no), claim type (Compulsory Third Party [CTP]/Workers Compensation [WC]/other), claim accepted (yes/no/don’t know), and legal representation obtained (yes/no). Claim made ‘yes’ was defined as making a personal injury claim of any type (CTP, WC or other) to access entitlements, which included a CTP Accident Notification Form (ANF) for expenses < AUD$5000 within 28 days of injury. Self-reported fault of the driver was measured at baseline (i.e. whether the driver considered themselves to have caused the crash). Pedestrians and passengers were considered not at fault (at baseline) because road rules dictate that vehicles must give way to pedestrians. However, both have a responsibility for their own safety and where they fail to take care, rules of apportionment under contributory negligence apply, that is: the insurer believes that the person contributed to the crash and/or their injuries (e.g. not wearing a seatbelt or crossing the road at a red traffic light can result in reduced financial entitlements at settlement [49].

Data analysis

Descriptive statistics were used to summarise baseline characteristics of the participants by claim status (i.e. made a claim Yes/No) at 6 months. The differences in the baseline characteristics between those that claimed compensation and those that did not were compared using Analysis of Variance (ANOVA) tests for continuous variables and chi-squared tests for categorical variables. The variables met the assumptions of independence, homoscedasticity and normality. Chi-squared tests were also undertaken to determine relationships between claim type and legal representation as well as claim type and claim acceptance. Logistic regression models were employed to determine predictors of claiming compensation and legal representation at six months. All potential predictor variables were considered for the final predictive models. We included variables with a p-value of <0.20 in univariate analyses and retained only variables with a p-value of ≤0.10 in the final models. Variables for the final model were selected using a backward elimination technique based on changes in likelihood ratios. The Spearman’s Rank-Order Correlation Coefficient and the Variance Inflation Factor (VIF) tests were performed to check for multicollinearity between variables in the model. The C-statistic (equivalent to the area under the Receiver Operating Characteristic curve) was used as an indication of the predictive accuracy of the final models. To assess the impact of eligibility factors, sensitivity analysis was conducted using only those participants that were eligible (n = 180). All data analysis was performed using SPSS statistical software version 21 (SPSS Inc, USA).