Health State Preference Weights for the Glasgow Outcome Scale Following Traumatic Brain Injury: A Systematic Review and Mapping Study

Abstract

BackgroundValid and relevant estimates of health state preference weights (HSPWs) for Glasgow Outcome Scale (GOS) categories are a key input of economic models evaluating treatments for traumatic brain injury (TBI).ObjectivesTo characterize existing HSPW estimates, and model the EuroQol five-dimensional questionnaire (EQ-5D) from the GOS, to inform parameterization of future economic models.MethodsA systematic review of HSPWs for GOS categories following TBI was conducted using a highly sensitive search strategy implemented in an extensive range of information sources between 1975 and 2016. A cross-sectional mapping study of GOS health states onto the three-level EQ-5D UK tariff index values was also performed in patients with significant TBI (head region Abbreviated Injury Scale score ≥3) from the Victoria State Trauma Registry. A limited dependent variable mixture model was used to estimate the 12-month EQ-5D UK value set as a function of GOS category, age, and other explanatory variables.ResultsSix unique HSPWs from five eligible studies were identified. All studies were at high risk of bias with limited applicability. The magnitude of HSPWs differed significantly between studies. Three class mixture models demonstrated excellent goodness of fit to the observed Victoria State Trauma Registry data. GOS category, age at injury, sex, comorbidity, and major extracranial injury all had significant independent effects on mean EQ-5D utility values.ConclusionsThe few available HSPWs for GOS categories are challenged by potential biases and restricted generalizability. Mixture models are presented to provide HSPWs for GOS categories consistent with the National Institute for Health and Care Excellence reference case.

Development of Methods for the Mapping of Utilities Using Mixture Models: Mapping the AQLQ-S to the EQ-5D-5L and the HUI3 in Patients with Asthma

“…The first is an ALDVMM implemented using the publicly available Stata command “aldvmm” [18] . This model has previously been applied for mapping across a range of clinical areas, including rheumatoid arthritis [18] , [19] , osteoarthritis [27] , ankylosing spondylitis [28] , and traumatic brain injury [29] . It has been shown to outperform other methods (linear regression, Tobit, and response mapping).…”

Mapping the FACT-B Instrument to EQ-5D-3L in Patients with Breast Cancer Using Adjusted Limited Dependent Variable Mixture Models versus Response Mapping

“…Independent variables can influence component membership and/or each of the components in the mixture model part. This model has been used with success in previous mapping studies [4], [6], [7], [10], [11], [12].…”

Cost-effectiveness analysis of tranexamic acid for the treatment of traumatic brain injury, based on the results of the CRASH-3 randomised trial: a decision modelling approach

“…It also impacts on the economy due to healthcare expenditure and productivity loss. 2 Around 8% of all TBIs are classified as severe (Glasgow Coma Scale (GCS) score of [3][4][5][6][7][8], which are more likely to result in death or long-lasting disability, compared with mild (GCS of [13][14][15] or moderate (GCS 9-12) head injury, for which the risk of death is lower, but still considerable. 1 3 Tranexamic acid reduces bleeding by inhibiting the breakdown of blood clots.…”

“…A systematic review and EQ-5D utility mapping study was identified, which reported utility values for patients with TBI, based on their level of disability, as defined by the Glasgow Outcome Scale (GOS). 15 In the absence of GOS outcomes in the CRASH-3 trial, we used the DRS scores of CRASH-3 trial patients to estimate the proportion of patients in each GOS category, using a qualitative estimation involving a clinical expert. This resulted in an estimated 3094 patients with a good recovery, 1288 with moderate disability, 677 with severe disability and 124 in a vegetative state.…”