Physician Substance Abuse and Recovery: Title and subTitle BreakWhat Does It Mean for Physicians—and Everyone Else?

The 10% to 15% prevalence of substance use disorders among physicians is similar to that in the general population,^{1 - 2} but the quality and intensity of treatment given to physicians may far exceed that available to other individuals with these disorders.^{3 - 5} Recognition of the impaired physician began to emerge only in the 1970s⁶ and has led to the development of physician health programs (PHPs). These are now mature models, available in many states, usually through medical societies, as an alternative to monitoring by state government boards of registration in medicine.⁷ In many cases, physicians who voluntarily contract with a PHP may remain anonymous⁷ to the state medical board and the National Practitioner Data Bank, a feature designed to promote early intervention in the disease process, ie, before patients are harmed. Many PHPs now offer services to other health professionals also. Treatment in these programs is probably the most comprehensive available for the disease,⁸ likely to include a full continuum of care, longitudinal (1- to 5-year) management, contracting for treatment and mutual help group (eg, Alcoholics Anonymous) participation, frequent assessment, random urine testing with observed micturition, hair testing for abused substances, and workplace surveillance.⁹

Referral of an impaired physician is nonpunitive, imperative, and can be life-saving—for both patients and the impaired physician. Both the Joint Commission on the Accreditation of Healthcare Organizations¹⁰ and the American Medical Association¹¹ emphasize the ethical importance of reporting impaired and disruptive physicians. For any physician who is concerned about a colleague, regardless of the relationship, the key step is to call the regional PHP (http://www.fsphp.org) to anonymously request guidance.

Despite the differences in care for impaired physicians compared with other persons with substance use disorders, anything that adds to current knowledge of addiction in physicians potentially adds to the understanding of all human addiction. Hence there will be great interest in the article by Domino et al in this issue of JAMA¹² reporting 11 years of data from the Washington Physicians Health Program (WPHP), a posttreatment program monitoring physicians and other health professionals. The hypothesis was that anesthesiologists, reportedly overrepresented among impaired physicians^{13 - 14} and at markedly greater risk for death from this disease,¹⁵ would show the highest risk for relapse. Also, given that anesthesiologists have ready access to parenteral opioids, it would seem reasonable to hypothesize that opioids would be the agent with highest prevalence in relapse. Confirmed risk predictors can help improve outcomes in impaired physicians, although they might be rather specific for this subpopulation. Fortunately, broader questions about the disease were also asked in this study.

Addiction is a disorder of the brain’s reward system. Functional imaging shows the vulnerable circuitry for addiction originating in the paleocortex.¹⁶ Paradoxically, humankind’s greatest adaptive advantage, the neocortex, responsible for the phenomenon of consciousness, is at best only minimally protective from addictive disease and may pose a hurdle for recovery. Unlike most medical disorders, in addiction a net effect of supraphysiologic reward, impaired inhibition, or both paradoxically leads the limbic drive system to reinforce exposure to the disease vector. This is in direct violation of the principle of survival of the species. In individuals with underlying vulnerabilities, limbic drive progressively recruits neocortical function to protect continued access to abused substances, the polar opposite of self-preservation. Thus, when physicians, whom society selects for high-level cortical functioning, become alcoholic or addicted, they often manifest exceptionally rationalized denial and sophisticated resistance.

In the study by Domino et al,¹² 1 in 4 monitored physicians (total n = 292) relapsed at least once—risking license, livelihood, and identity. To those physicians who enjoy freedom from this vulnerability, relapsing under monitored, sanctioned conditions might seem to be a remarkably poor behavioral choice. The fact that well-trained (and not infrequently otherwise successful) physicians do so is at least circumstantial evidence that the drive to relapse originates far from the realm of conscious intent.

It is notable that the vast majority of physicians who have substance use disorders seem to do surprisingly well in recovery.^{7 ,9 ,12 - 14 ,17} This is an unusual context, however. Monitored physicians receive an optimal treatment model that assumes primary medical responsibility for a disease that is inherently self-destructive. It combines empathic support with the high level of structure—restrictiveness, actually—of close monitoring and sanctions. Physicians’ treatment appears to be matched according to need,¹⁸ but for the general public, this is only a recent development, strongly championed by the American Society of Addiction Medicine,¹⁹ and with evidence for validity in a range of populations,²⁰ although implementation is only now emerging.

The data in the study by Domino et al¹² suggest a particular pattern to physician relapse. It is a pattern that is neither attributable to chance nor to volition. What is left, if not accident or intent?

Family history, for one thing. As found in prior general population studies showing strong specific genetic risks for alcoholism²¹ and among physicians surveyed for substance use,²² nearly three fourths of these physicians had a family history of substance use disorder, and this more than doubled the likelihood of a relapse (hazard ratio [HR], 2.29; 95% confidence interval [CI], l.44-3.64). The original primary agent of abuse at the time of monitoring did not have an impact on relapse rates, challenging prior assumptions about the elevated risk of opioid use, although agreeing with more recent findings.²³ Instead, use of a major opioid increased relapse rates more than 5-fold but only in physicians with a coexisting psychiatric disorder (HR, 5.79; 95% CI, 2.89-11.42). More than a third of these physicians had a coexisting psychiatric disorder, but only 7% had any personality disorder. In physicians with all 3 factors (major opioid use, dual diagnosis, and family predisposition), the risk of relapse was elevated 13-fold (HR, 13.3; 95% CI, 5.22-33.6).

This finding regarding dual disorders suggests that psychiatric evaluation is an important aspect of assessing impaired physicians and that a final diagnosis should be made only following an extended period of monitored abstinence, as was done in this study. The doubling of the rate of dual diagnoses in the second half of the enrollment period is noteworthy. A similar increase was reported by Angres et al.²⁴ Attributed to better psychiatric staffing and psychometric assessment, this improved detection is potentially valuable in improving outcomes.

It remains to be seen how improved detection and better-matched recovery planning will address those with the doubled relapse risk of dual diagnosis and the multifold risk of the triple-threat: dual diagnosis, opioid dependence, and family history. These data suggest that analyzing the trajectories of recovering physicians may improve the knowledge base for anticipating and matching the needs of physicians entering recovery. But retrospective cohort analyses from single states using clinically derived data, while better than no data at all, are inadequate in this millennium. Better data are needed, such as multistate data from prospective studies with research quality instrumentation. It is time for PHPs to become formal research programs, or better yet, to form a national research program.

Individualized monitoring plans and treatment contracts that take into account various risk loadings should improve outcomes for patients with substance use disorders. Success from substance use disorder treatment should be sought and expected not just for physicians but for every patient—but only if the conditions available for physician recovery can be provided to all.¹⁷ These “ifs” are pivotal: if intervention occurs early, if structure is provided as well as support, if treatment resources are provided as if a life and career matter, and if close monitoring and treatment matching are provided with active treatment intervention and escalation to meet the clinical need. Surely this type of care is costly. Why should such high-quality care be provided? Because a brain disease that subverts self-preservation is a disease nonetheless, and helping patients recover from this disease can save lives, families, and productive careers.