Treating Physicians for Addiction

If you are in a workroom with six to seven other doctors, it is statistically probable that one meets criteria for alcohol use disorder, and nearly 1 in 10 of those have thought about suicide in the last year. A recent study of 7,288 physicians asserts that 15.3% of the profession may meet DSM-IV criteria for alcohol abuse or dependence. These practitioners are more likely to experience burnout, screen positive for depression, and have lower quality of life, and 8.8% reported suicidal ideation in the preceding 12 months (1). This is not a new problem. In 1973, the American Medical Association published a seminal report that summarized the central issue of doctors with addiction: A lenient approach to impaired physicians could risk patient safety; a harsh approach, often resulting in financial ruin and public shaming, might frighten practitioners into hiding illness precisely when care was most needed. The report called for action noting that, “Suicide is generally accepted to be one of the major behavioral consequences of mental illness. About 100 physicians commit suicide annually, equivalent to the size of the average medical school graduating class (2).”

Available Treatment and Concerns

State-based organizations began to form around the 1970s, often being run by physicians with a personal history of successful recovery (3). Today, physicians with substance use disorders are referred to these physician health programs (PHPs) for care. Concerns have been raised that these programs are invasive, coercive, expensive for the participant, and may need organizational oversight (4). Even so, a series of analyses based on an observational study data set that enrolled 904 consecutive PHP participants in 16 states from 1995 to 2001, which followed each participant for 5 years, found exceptional long-term recovery rates compared with other approaches: 78% of participants were without a single positive screen for alcohol or drug use over 5 years compared with 40%–60% relapse rates after just 6 months with standard treatments; 72% were licensed and practicing after 5 years (5, 6). Furthermore, one study of the PHP in Florida revealed that 92.5% of participants would recommend participation to others (7). A state audit of the North Carolina PHP found an organizational structure that might allow malfeasance but noted that none had occurred. A sample of 10% of all participants in North Carolina from 2002 to 2012 revealed that there was “sufficient, appropriate evidence to support referral to a treatment center” in every case (8).

How PHPs Work

PHPs do not provide care; they provide case management. After referral by a loved one, colleague, or employer, physicians undergo assessment. If treatment is indicated, a PHP contract lasting 1–5 years is offered and generally includes “safe harbor” from prosecution or professional consequences pending successful completion. More than 88% run the full 5 years. These contracts usually require total abstinence, and 95% engage participants in 12-step-oriented treatment. Two-thirds of participants start by attending inpatient residential treatment for an average of 72 days. The rest attend intensive outpatient treatment initially. This is followed by close monitoring and 1–4 nights of contractual activities for the first year (e.g., 12-step meetings, case management). Contracts often require consent to contact family and ongoing access to all medical records. There are commonly duty-hour restrictions, unannounced work-site visits, and on-site monitors. On average, 48 random drug screens are collected in year 1, and around 20 are collected in year 5. Consequences for infractions, from refusing to provide a urine sample to being found intoxicated on the job, can result in further evaluation or treatment, a report to the licensing board, or more serious consequences (3). Fundamentally, “safe harbor” allows physicians to seek help without fear, while PHP intensity, invasiveness, and duration protects patients from impaired doctors.

Of all 904 physicians followed in the cohort mentioned above (5), 50% were referred primarily for alcohol use and 35% for opioid use. Thirty-one percent struggled with both drug use and alcohol use. Fourteen percent used intravenous drugs. One was prescribed methadone. Six percent were prescribed naltrexone, and 32% were prescribed an antidepressant for depression or anxiety. Twenty-two percent had one positive screen. Of these, 26% had a second positive screen. A total of 448 completed their contract. Eighty-nine extended their contract voluntarily. A total of 110 extended their contract involuntarily. Sixty-nine transferred to another state. Thirty-three were lost to follow-up after moving. Eighty-five withdrew, often retiring or surrendering their licenses. Forty-eight failed treatment and had their licenses revoked. Twenty-two died; of these deaths, two were substance-related and six were by suicide (6).

Hard to Scale

There is an ongoing effort to change the way addiction treatment is defined and measured to align with the PHP model (3, 9). But the length and expense, as well as the motivation inherent in the threat of license revocation, make it hard to apply this model to other populations. Evaluation can cost $4,500 and residential treatment $45,000 (8). One study attempted 1 year of PHP-style monitoring in a general population and showed moderate benefit, but the program could only be offered to self-pay patients (10). With such high costs, the coercive nature of participation can become a hardship for allied health care professionals and physicians early in training. The Michigan HPRP [Health Professional Recovery Program], which manages care for multiple health care career fields, is being sued in a class action lawsuit brought by three registered nurses and a physician’s assistant because, per the filing document, “failure to ‘voluntarily’ submit to unnecessary and costly HPRP treatment results in automatic summary suspension by the Bureau of Healthcare Services [sic] without a pre-deprivation hearing” [correctly the Bureau of Community and Health Systems] (11).

Late Intervention

No source included an accounting of how many referred physicians were evaluated and found not to need treatment. The implicit conclusion is that all need treatment. While this raises the concern that conflicts of interest may lead to inappropriate coercion, the findings of the North Carolina audit (8) and the Florida satisfaction study (7) provide clear evidence to the contrary. Another possible conclusion is that physicians are only referred for this kind of evaluation when they have reached such extremity that intensive, long-term treatment and monitoring are always needed. It may be that physicians struggle with how or if to approach the problem of an impaired colleague (12).

Conclusions

Many physicians are likely to suffer from substance use disorders (1), and JAMA recently published a compelling opinion piece about the problem (13). To allow physicians to access care without fear while also providing for the protection of the public, state-based PHPs coordinate evaluation, case-management, and “safe harbor” from certain consequences pending successful completion of a PHP contract (5). Concerns have been raised (4), but PHPs produce better results than standard treatment (6). In apparent support, the American Medical Association recently released draft legislation to facilitate formal codification of the PHP paradigm (14). Although physicians may certainly seek care on their own, PHPs provide an evidence-based, though intense and possibly invasive, route to rehabilitation and recovery.

Key Points/Clinical Pearls