Abstract: Biases in the collection of blood alcohol data likely impact the surveillance of alcohol-related injuries. Routine alcohol testing after major trauma is needed to accurately inform epidemiology and the subsequent implementation of strategies for reducing alcohol-related injuries.
Introduction
Acute alcohol use is a leading risk factor for major trauma, yet hospital screening remains inconsistent. When clinicians only test patients they suspect are intoxicated, they introduce "selection bias." This creates a feedback loop: if testing is prioritized for specific demographics based on clinical suspicion, the resulting data will "confirm" those groups as the primary risk, potentially missing broader trends and opportunities for intervention.
This study, published in the Emergency Medicine Journal, investigates the systemic biases inherent in how hospitals collect blood alcohol concentration (BAC) data from major trauma patients within the Victorian State Trauma System (VSTS).
Methods
The researchers conducted a retrospective cohort study using data from the Victorian State Trauma Registry (VSTR), a population-based registry collecting data from all 138 trauma-receiving hospitals in Victoria, Australia. The study analyzed records for 14,221 adult major trauma patients (age $\ge$ 18) injured between January 2018 and December 2021.
The team employed multivariable logistic regression models to identify predictors of receiving a blood alcohol test. A key feature of their methodology was a sensitivity analysis comparing two different hospital environments:
- MTC1: A Major Trauma Centre where alcohol testing is part of a routine, automated panel of laboratory tests for trauma patients.
- The rest of the system: Hospitals where testing is left to "clinician discretion."
Key Results
The study found that only 32.1% (4,563 of 14,221) of major trauma patients had a BAC test recorded. However, the testing rate was highly dependent on institutional policy: testing occurred in 85.2% of cases at the routine-testing site (MTC1) compared to just 7.8% at non-major trauma sites.
In discretionary settings, significant demographic and clinical biases were observed:
- Stereotyping in the ER: Patients were significantly more likely to be tested if they fit a "high-risk" profile: younger, male, current smokers, or those with a history of substance use.
- The Invisible Elderly: Patients aged 65+ were significantly less likely to be tested (aOR 0.57), suggesting a clinical blind spot regarding alcohol use in older populations prone to falls.
- Language Barriers: Patients whose preferred language was not English had 38% lower odds of being tested, highlighting potential inequities in care.
- Socioeconomic Bias: Patients from the most disadvantaged areas had 2.3 times higher odds of being tested than those from the least disadvantaged areas in the adjusted model.
- Clinical Indicators: Lower Glasgow Coma Scale (GCS) scores were associated with higher odds of testing, as clinicians sought to differentiate between head injury and intoxication.
Discussion and Limitations
The findings demonstrate that policy dictates data quality. Implementing routine, opt-out testing protocols effectively eliminates most demographic biases, providing a clearer picture of alcohol-related injuries. Discretionary testing leads to "clinical suspicion" biases related to age, language, and socioeconomic status, which can skew public health data.
The authors acknowledge several limitations:
- First Hospital Focus: The study only looked at testing at the first hospital a patient reached.
- The "Time" Factor: The study did not account for the time elapsed between injury and the blood draw, which is critical for accurate BAC readings as alcohol metabolizes quickly.
- Age Exclusion: The study excluded patients under 18.
The paper strongly advocates for universal, routine alcohol screening across all trauma centers to improve the surveillance of alcohol-related injuries and the subsequent design of prevention strategies.
Summary updated