Physicians and nurses with substance use disorders


  • Matthew F. Shaw PhD,

  • Mark P. McGovern PhD,

  • Daniel H. Angres MD,

  • Purva Rawal MA

Matthew Shaw,
Yale University School of Medicine,
333 Cedar Street,
New Haven,
CT 06510,


Background.  The literature addressing substance use patterns among medical professionals suggests that specialty, gender, age, familial substance abuse, and access/familiarity with prescription drugs are associated with particular chemical dependencies. These studies have rarely compared nurses and physicians directly, thereby making if difficult to tailor interventions to the potentially unique needs of each group.

Aim.  This paper reports a study to compare the initial clinical presentations, service utilization patterns, and post-treatment functioning of nurses and physicians who received services in an addiction treatment programme.

Method.  This exploratory study combined data collected through retrospective record reviews and prospective questionnaires. There were three types of dependent variables: initial clinical characteristics, treatment utilization patterns, and post-treatment functioning. The independent variable was membership of either professional group. Time both in treatment and between discharge and follow-up were covariates.

Results.  Nurses and physicians showed comparable results in most domains. Among the statistically significant differences between groups, a subset was particularly noteworthy. Prior to participating in the programme nurses showed significantly less personality disturbance than physicians, although they tended to work and live in environments with more triggers to relapse, such as other substance users. After the index hospitalization, nurses received less primary treatment, worked longer hours, and were more symptomatic than physicians. Furthermore, nurses reported more frequent and severe work-related sanctions as a consequence of their behavioural disorders.

Conclusion.  In most areas of study, nurses and physicians demonstrated comparable results; however, a series of statistically significant differences suggest that these groups may have unique clinical needs. The policy implications of these findings are discussed.