Are Health Workers More Prone to Alcohol-Related Death?

Alcohol-related deaths were not significantly higher for U.S. health workers compared with workers in the general population, with the lowest rates observed among physicians, a cross-sectional study found.

In an age- and sex-standardized analysis, rates of alcohol-related death were significantly lower among physicians and other diagnosing or treating clinicians, at 4.2 and 11.6 per 100,000 person-years, respectively, versus 18.1 per 100,000 person-years for non-healthcare workers, reported Mark Olfson, MD, MPH, of Columbia University in New York City, and colleagues.

However, after adjusting for a host of other factors -- including race/ethnicity, marital status, education, and income -- no significant differences were seen for physicians (adjusted HR 0.65, 95% CI 0.32-1.34), diagnosing or treating clinicians (adjusted HR 1.13, 95% CI 0.75-1.06), or any other health worker group.

Among the other health worker groups examined, the highest age- and sex-standardized rates of alcohol-related death were recorded in healthcare support workers, at 19.5 per 100,000 person-years, followed by registered nurses (16.8), social and behavioral health workers (16.5), and health technicians (13.8), according to the findings in JAMA Network Open.

"The risks of alcohol-related deaths among healthcare workers did not exceed those of non-healthcare workers," the researchers wrote. "However, this finding does not diminish the importance of improving management of alcohol-related problems among healthcare workers."

Olfson told MedPage Today that health workers are sometimes assumed to be at higher risk for problematic alcohol use, given that they "are routinely engaged in stressful tasks of caring for severely ill individuals and managing heavy workloads with often little control over patient outcomes."

The findings diverge from studies that have found higher risks for drug-overdose death in some health workers compared with workers in non-healthcare settings.

"The alcohol-related mortality results suggest healthcare workers do not have a general underlying liability to substance-related deaths," Olfson and coauthors wrote. "Specific occupational factors, such as access to controlled medications, may pose drug overdose risks that do not extend to alcohol-related deaths."

Issues affecting the general public tend to affect physicians less, Manish Sood, MD, of the University of Ottawa in Ontario, told MedPage Today via email, but they "receive considerable attention because when a physician is incapable of working due to these issues it affects many people in the community (their patients), so maintaining physician health is an important public health issue as it keeps communities healthy."

That said, "For other healthcare workers the takeaway is they are people too, same as non-healthcare workers," noted Sood, who was not involved in the study. "Everyone should reduce their alcohol intake based on recent emerging data and guidelines that there is no 'safe' amount," he added.

For the study, the researchers linked data from the 2008 American Community Survey (ACS) to National Death Index records from 2008 to 2019. They excluded people who were unemployed or younger than 26, leaving 1,838,000 workers in total, the bulk of which were non-healthcare workers (n=1,666,000).

The median age of the cohort was 44 years, and 47.7% were female. The health workers included 42,000 registered nurses, 39,000 healthcare support workers, 32,500 technicians, 27,000 social and behavioral workers, 22,500 other diagnosing or treating clinicians (e.g., dentists), and 13,000 physicians.

The adjusted analysis accounted for occupational group, age, sex, race and ethnicity, marital status, educational level, urban or rural residence, and income group.

Limitations included that the dataset did not extend into the COVID-19 pandemic, noted Olfson and colleagues, and they said that misclassification of alcohol-related deaths may have occurred. They also had no data on alcohol-related risk factors, like family history of alcohol use, and could not measure occupational transitions throughout the study period.

Comment