Prevalence of substance use among US physicians

JAMA. 1992 May 6;267(17):2333-9.


Objective: To estimate the prevalence of substance use among US physicians.

Design: A mailed, anonymous, self-report survey that assessed use of 13 substances and permitted comparison with results of the National Household Survey on Drug Abuse. Rates of physician substance use were weighted to provide national prevalence estimates.

Participants: A national sample of 9600 physicians, stratified by specialty and career stage, and randomly selected from the American Medical Association master file. The response rate after three mailings was 59%. Demographic characteristics of respondents closely reflected those of the US physician population.

Main outcome measures: Subjects' self-reported use of 13 substances in their lifetime, the past year, and the past month; reasons for use; self-admitted substance abuse or dependence; and whether treatment was received. For controlled prescription substances, respondents were asked to report only use "not prescribed by another physician for a legitimate medical or psychiatric condition."

Results: Physicians were less likely to have used cigarettes and illicit substances, such as marijuana, cocaine, and heroin, in the past year than their age and gender counterparts in the National Household Survey on Drug Abuse. They were more likely to have used alcohol and two types of prescription medications--minor opiates and benzodiazepine tranquilizers. Prescription substances were used primarily for self-treatment, whereas illicit substances and alcohol were used primarily for recreation. Current daily use of illicit or controlled substances was rare.

Conclusions: Although physicians were as likely to have experimented with illicit substances in their lifetime as their age and gender peers in society, they were far less likely to be current users of illicit substances. The higher prevalence of alcohol use among respondents may be more a characteristic of their socioeconomic class than of their profession. A unique concern for physicians, however, is their high rate of self-treatment with controlled medications--a practice that could increase their risk of drug abuse or dependence. Uniform national guidelines are needed to sensitize medical students and physicians to the dangers of self-treatment with controlled prescription substances.

Publication types

  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Adult
  • Alcoholism / epidemiology
  • Anti-Anxiety Agents
  • Benzodiazepines
  • Female
  • Humans
  • Male
  • Middle Aged
  • Narcotics
  • Physician Impairment / statistics & numerical data*
  • Prevalence
  • Sampling Studies
  • Smoking / epidemiology
  • Substance-Related Disorders / epidemiology*
  • United States / epidemiology


  • Anti-Anxiety Agents
  • Narcotics
  • Benzodiazepines