Dr. A, a 36-year-old married male internist, was referred to G.O.G. from his state’s physicians’ health program for a comprehensive diagnostic assessment, including evaluation of his ability to practice medicine. The referring physician had concerns about a personality disorder and the potential for a mood disorder, in addition to known substance abuse, which precipitated surrender of Dr. A’s medical license in two states.
Dr. A’s mood symptoms had begun when he experienced feelings of depression as a child. His mood seemed to improve as he aged, but in college he had periods of depression for a few days at a time. His first treatment for depression occurred when he was a first-year medical student. One of his clinical teachers offered him samples of fluoxetine, which he took for 6 months. During his internship, he tried another antidepressant for a period of time as well. When he was in his last year of his internal medicine residency, he was treated with nefazodone, which he ultimately discontinued on his own. In recent months, he prescribed venlafaxine for himself. Dr. A had no history of clear episodes of mania, but he had some symptoms of hypomania when his venlafaxine dose had been increased to 300 mg/day.
Dr. A’s history of excessive drinking began at age 17, when he started experiencing blackouts every 3 to 4 months during binge drinking episodes. He continued to drink heavily intermittently throughout college and medical school. By the time he was a resident, his drinking had escalated to the point at which he was intoxicated most weekends when he was not on call or moonlighting. He became worried that he would become an alcoholic like his father, and he entered a residential rehabilitation program midway through his residency that led to approximately 3 years of abstinence from alcohol.
After the completion of his residency, he worked as a solo practitioner in a private office and was reasonably successful. However, he developed tension headaches, so he prescribed tramadol for himself. Over a period of 8 months, he increased his dose of tramadol to approximately 1.25–2.5 g/day. He had been attending Alcoholics Anonymous (AA) meetings regularly and had remained abstinent from alcohol. However, as he increased his tramadol dose, he experienced euphoria, and he realized he no longer needed to attend AA meetings in order to feel good. During the next 2 years, he increased his tramadol dose to an average of 5 to 6 g/day. On one occasion, this dose of medication induced a grand mal seizure while he was performing a minor medical procedure. Dr. A then medicated himself with primidone to prevent further seizures and to control an intention tremor in his hands.
For 3 years, he continued to abuse tramadol but remained abstinent from alcohol. After he was accused by his hospital medical staff of incompetent management of a complicated hospital patient, he relapsed and drank on three separate occasions. He left the hospital and worked at two different emergency rooms, requiring long hours away from his wife and an erratic sleep schedule.
Dr. A moved to another state to take over the internal medicine practice of a retiring physician. His wife stayed behind because of employment commitments. He reported being lonely in this new setting, and he started drinking heavily again while continuing to abuse tramadol. At this point, he was spending $1,000 a month to sustain his habit. He eventually came to the attention of the licensing board of this new state when he was arrested for being drunk and disorderly. He again entered a residential rehabilitation program, where he stayed for 3 months. He dreaded going home on weekend passes because of the stress of his marriage. He reported that his wife yelled and screamed at him, and he felt he was reexperiencing the violence of his childhood. On one occasion while he visited his wife on a pass, they had a fight, and she threatened divorce.
He was then transferred to a halfway house, where he stayed for only a few days. While there, he asked one of the female patients to have sex with him, thinking that sex would elevate his mood after his wife threatened divorce. He was also involved in a physical altercation with a male patient, although Dr. A described the incident as follows: "I was on the phone, and this guy came up and hit me." Dr. A was hauled off by the police and then thrown out of the halfway house. He found another halfway house, where he was able to maintain abstinence and was able to visit his wife on weekends.
Dr. A’s developmental history is notable for its extensive trauma. He was the oldest of four children, with one younger sister and two younger brothers. His father was a truck driver who was on the road most of the time. While he was away, his mother had multiple affairs; Dr. A reported that he witnessed his mother having sex with a number of different lovers. He also witnessed extreme violence between his mother and her lovers. Eventually, social services removed him from his mother’s household, and his father then raised him. Dr. A described his father as an alcoholic, but he said they had a reasonably good relationship, and Dr. A felt supported by him. He has an extensive family history of alcoholism that included both paternal grandparents and a maternal grandmother.
Dr. A was aware that he had a strong inclination to heavy drinking, so he deliberately chose a small religious college where he would not be tempted to spend each weekend partying. Nevertheless, his pattern of alcohol abuse continued. He married his wife during his residency, and the marriage gradually deteriorated when his wife discovered that he was using tramadol extensively. They repeatedly engaged in verbal arguments and occasionally became involved in physical altercations. Many of Dr. A’s relapses occurred after these fights. He and his wife did not seek consistent marital therapy for their difficulties, in part because his wife felt that the problems were his. Dr. A acknowledged his fear of abandonment by women and realized that he may unintentionally drive away women before they abandon him first.
Dr. A’s medical history is positive for mild hypertension, intention tremor, allergic rhinitis, obesity, tinnitus, gastroesophageal reflux, and irritable bowel syndrome. At the time of the evaluation, his medications included venlafaxine (sustained release) 225 mg/day; divalproex, 500 mg b.i.d.; doxepin, 75 mg/day at bedtime; and hydrochlorothiazide, 12.5 mg/day. He had not taken tramadol for approximately 4 months at the time of his evaluation.
At his mental status examination, Dr. A was a casually dressed, moderately obese man who appeared to be his stated age. He was oriented to place, person, time, and situation. He was cooperative and polite throughout the evaluation. He had good eye contact and revealed no motoric abnormalities. His thought processes were well organized but slightly slowed and somewhat overinclusive in the number of details he provided. He showed no evidence of a thought disorder or a flight of ideas. His mood appeared euthymic. His affect was constricted, with only rare expressions of emotion. He reported no suicidal ideation.
Dr. A’s history and presenting symptoms required psychological assessment for bipolar disorder (1) and/or severe character pathology (e.g., borderline). Hence, both projective (the Rorschach , the Thematic Apperception Test, the Incomplete Sentences Test ) and self-report personality tests (the MMPI-2, the Personality Assessment Inventory ) were needed. Neuropsychological measures for potential cognitive impairments resulting from chronic alcohol and substance abuse (4) were also indicated. Cognitive tests were additionally important to establish his fitness to function mentally at a level commensurate with the complexities of medical practice. The tests included the WAIS-III, the logical memory and visual reproduction subtests from the Wechsler Memory Scale—Revised (5), the California Verbal Learning Test (6), Trail Making Tests A and B (7), the Stroop Color-Word Interference Test (8), the Controlled Oral Word Association Test (9), and the Hooper Visuospatial Organizational Test (10). An author (E.S.) administered these tests under my supervision.
Dr. A appeared to be a pleasant and cooperative man. During the assessment, he was easy to engage in conversation, and rapport was established with little effort. Although Dr. A’s performance on the WAIS-III was superior overall, with a full-scale IQ of 126, his verbal skills were significantly higher than his nonverbal skills. His verbal IQ of 131 fell into the very superior range, whereas his performance IQ of 114 was in the high-average range. In score discrepancy distributions on the WAIS-III in the adult normative population, this occurs less than five times in 100 cases (p<0.05). His poorest performance was on measures of psychomotor speed (index score=103). While such an IQ disparity favoring verbal over performance IQ is not unusual in very bright individuals, marked differences of this kind are also known concomitants of chronic alcohol/drug abuse (4). This difference could also reflect the influence of his depression and/or prolonged tramadol use. Overall, Dr. A’s memory abilities were unimpaired. All estimates of his executive cognitive functioning were within the normal range.
Projective testing with the Rorschach, the Thematic Apperception Test, and the Incomplete Sentences Test was unrevealing for significant indications of a thought disturbance. There were no indications of a flight of ideas, grandiosity, or the pressured thoughts seen in bipolar disorder.
However, on both the MMPI-2 and the Personality Assessment Inventory personality scales, Dr. A scored in the clinically significant range for affective instability. His MMPI-2 profile score, in particular, was elevated on the hypomania scale. For individuals who are not genuinely bipolar, this finding can reflect a propensity for overexpansive or irritable-changeable behavior patterns, vulnerability to boredom, or a low tolerance for frustration. Our clinical experience has been that individuals with such characteristics may gravitate toward using substances to control mood or behavior disturbances. Dr. A’s score elevation on the Personality Assessment Inventory substance abuse index more directly identified drug abuse/addiction proneness as a primary diagnostic consideration. These self-report-generated impressions certainly corresponded to Dr. A’s known history of substance misuse.
Overall, the test findings suggested comorbid axis I and axis II pathology and a complex clinical syndrome necessitating a guarded prognosis. The conditions appeared treatable with no absolute contraindications to return to the practice of medicine. Considering his problematic history at the time of evaluation, these findings presented potentially encouraging news for this individual’s rehabilitation potential.
G.O.G. asked me to consult on Dr. A to assess the possibility of a bipolar affective disorder. This case raises several important issues. Comorbid mood disorders and substance abuse are common, and both disorders warrant aggressive treatment. In consideration of a possible mood disorder diagnosis, Dr. A clearly described depressive periods starting in adolescence, before his extensive abuse of substances. As such, a mood disorder may be one of his primary diagnoses but not necessarily the primary cause of his impairment.
Dr. A’s depressive episodes are characterized by increased sleep, increased appetite, decreased energy, decreased concentration, irritability, psychomotor retardation, and thoughts of death. Although not pathognomonic, major depressive episodes associated with reverse vegetative signs (e.g., increased sleep), psychomotor retardation, and an early age at onset warrant consideration of bipolar disorder (11). Dr. A has exhibited reckless behavior and impulsivity, which are indeed common in the manic phase of bipolar disorder. However, his impulsive behaviors had not occurred in the context of a manic episode. Indeed, Dr. A had not had a clear episode that would meet current diagnostic criteria for mania. This is an important distinction, and it is imperative not to allow the potential diagnosis of bipolar disorder to preclude consideration of other causes of maladaptive behavior. In this individual, complex psychodynamic issues and substance abuse appear to be more closely related to the problems that have compromised his ability to practice medicine.
The issue of responsibility for one’s actions is especially difficult in the context of bipolar disorder. I have found that this is an increasingly difficult dilemma, as both the medical and nonmedical communities are embracing broader definitions of bipolar disorder. While in the midst of a classic manic episode, it is not unusual for patients to behave in ways that are highly uncharacteristic of their baseline personality and that may be associated with negative consequences. In such instances, particularly first episodes in undiagnosed individuals, a psychiatrist often helps these patients understand that their actions were the result of their disorder. However, even in classic type I bipolar disorder, the issue of responsibility is germane when mania and associated maladaptive behaviors are precipitated by nonadherence to treatment. This is more difficult in the case of bipolar spectrum disorders. Although the concept of bipolar spectrum is sometimes of clinical utility, as in the case of Dr. A, increasingly broad definitions of bipolar disorder are potentially detrimental. Of particular concern is application of the term "bipolar spectrum" to all patients with mood lability. We have seen many patients who inappropriately grasp on to the diagnosis of bipolar disorder to avoid taking responsibility for acts of impulsivity that are more probably part of another etiology—for example, personality disorders. The issue is perhaps even more complicated because of the frequent comorbidity between bipolar disorder, personality disorders, and substance abuse.
However, refinement of Dr. A’s mood disorder diagnosis is relevant to providing optimal pharmacological management. Dr. A described episodes that are phenomenologically consistent with hypomania, including increased speech, flights of ideas, increased energy, and increased confidence. These episodes have lasted for no longer than several days at a time and appear to be isolated to times when he is either abusing prescription drugs and/or taking antidepressants. According to the current nomenclature (DSM-IV), hypomanic periods that occur only in the context of antidepressant treatment are not considered bipolar disorder. Unfortunately, in the case of ongoing depression in a patient with antidepressant-induced hypomania or cyclic moods, there is a tremendous paucity of data to guide treatment decisions. Most commonly, these patients are considered to have a bipolar spectrum disorder and are treated with mood stabilizers on the basis of extrapolations from data on the treatment of bipolar depression. When Dr. A was treated with divalproex, 500 mg b.i.d., in addition to an antidepressant, he continued to experience cyclic moods, even during periods of sobriety. Higher doses of divalproex were associated with intolerable sedation. Lower antidepressant doses resulted in depression. Given his ongoing mood symptoms despite sobriety and the fact that the onset of mood symptoms preceded substance use, ongoing pharmacotherapy is essential.
Antidepressants may induce hypomania or rapid cycling in a substantial percentage of patients with bipolar disorders (12), and they appear to do so in this patient, even when he is not using substances. As such, I suggest minimizing the use of antidepressants in Dr. A. Because he tends to experience depression with divalproex monotherapy, the use of lithium or another anticonvulsant is a reasonable consideration for this patient. Lamotrigine has been demonstrated to be more effective than placebo as monotherapy in patients in the depressed phase of type 1 bipolar disorder, with no greater risk of treatment-induced mania or hypomania (13). A separate 6-month study of patients with rapid-cycling bipolar disorder (14) indicated that lamotrigine monotherapy was more effective in preventing relapse than placebo, especially in bipolar II disorder. Both lithium and lamotrigine are considered first-line treatments for bipolar illness and can be added to divalproex. If the patient’s mood is stable with one of these combinations, consideration can be given to tapering divalproex. When lamotrigine is added to divalproex, lamotrigine levels are increased, mandating a particularly slow titration to minimize the likelihood of lamotrigine-induced rash or Stevens-Johnson syndrome. Finally, I would also move his entire dose of divalproex to bedtime in order to minimize daytime sedation and taper him off of doxepin.
G.O.G. asked me to consult on Dr. A to provide an evaluation of his substance abuse and suggest optimal rehabilitation strategies. Dr. A has an extensive family history of alcoholism, and he has had a serious alcohol problem himself since adolescence. He uses tramadol to the point of seizure induction. He has an extensive history of self-prescribing. He satisfies the traditional definition of an impaired physician in that he is unable to practice medicine with reasonable skill and safety to patients because of excessive use or abuse of drugs and alcohol (15). The most accurate diagnosis from a substance abuse standpoint would be polysubstance dependence. Despite efforts at recovery and rehabilitation, he is still relapsing regularly and must be considered in the early stages of recovery.
Physicians who have not been able to sustain recovery are generally not safe to practice medicine. He needs to maintain sobriety for 12 months before we can confidently allow his reentry into medical practice. During that time, he needs to attend at least three AA meetings per week and work with a sponsor. He will need to be monitored by the physicians’ health organization of his state and have at least twice-weekly urine drug screenings as part of that process. In my evaluation of him, marital problems emerged as major triggers of his relapse, so attention to the marriage in an ongoing marital therapy process is essential. Dr. A has also fallen into the habit of prescribing for himself, and he must establish himself with a primary care doctor who will be responsible for all of his medications. He must agree to do no self-prescribing whatsoever.
Dr. A’s diagnostic picture is one of complex comorbidity. While the term "impaired physician" has come to mean chemical dependency in many quarters, detailed evaluation frequently reveals more complex psychopathology. Dr. A certainly meets criteria for polysubstance abuse, and this problem area needs to be addressed as a major focus of treatment. Psychological testing suggests the possibility that his extensive alcohol and drug abuse may be involved in the disparity between verbal and performance IQs, although we cannot be certain.
L.M. points out that while Dr. A’s hypomanic periods do not clearly meet strict criteria and have occurred in the context of antidepressant treatment, his pattern of cycling in response to antidepressant medication in conjunction with a clear history of depression suggest that treatment is warranted. We would probably consider his diagnosis as part of the bipolar spectrum, with an official diagnosis of bipolar disorder not otherwise specified.
The diagnosis of personality disorder is problematic when someone is abusing substances or cycling between hypomania and depression. However, Dr. A manifests signs of cluster B personality disorder even while sober and euthymic. These features include impulsivity, poor judgment, problematic interpersonal relationships, and identity diffusion. Although some question was raised by the referring physician about the presence of antisocial personality disorder, Dr. A lacks a history of frequent violence, arrest, cruel and sadistic behavior, and inability to feel remorse that accompany that diagnosis. He has gotten into fights, approached women with sexual intent in a highly inappropriate manner, and had difficulty integrating his professional role with his personal behavior. Patients with this degree of trauma history frequently begin to compartmentalize aspects of themselves associated with trauma and attempt to compensate for those experiences by becoming achievement oriented and attempting to prove themselves professionally. Despite his condition, Dr. A is deeply devoted to the practice of medicine. In the case of Dr. A, his zeal to heal others may be partially determined by an unsuccessful effort to cure his mother of her suffering when he was a child. As a result of this defensive strategy, his self is fragmented, and he finds it difficult to integrate the diverse parts of himself in a way that allows continuous and smooth day-to-day functioning. A dedicated physician self coexists side by side with an aspect of himself that is impulsive, irresponsible, and self-destructive. The two remain unintegrated and cause him great difficulty in understanding who he is.
Preliminary data suggest that this trauma-related compartmentalization may relate to hemispheric lateralization. Auditory-probe-evoked potential attenuation was measured as an index of hemispheric activity in 10 subjects with a history of childhood trauma and 10 matched comparison subjects without such history while they recalled a neutral memory and then a traumatic memory (16). Abused children used their left hemisphere when thinking about neutral memories and their right hemisphere for frightening memories. The comparison group used both left and right sides equally, regardless of the memory content.
This vertical splitting (17) between the two disparate aspects of Dr. A’s personality makes it difficult for him to see that he is an active agent in re-creating the traumatic environment of his youth. For example, he sees the fight in the halfway house as a matter of self-defense rather than as an event he had an active role in creating. Like many adults who were traumatized as children, he has difficulty mentalizing (i.e., recognizing that his behavior is generated by internal factors such as beliefs, feelings, and intentions) (18). The best axis II diagnosis for him would probably be personality disorder not otherwise specified because he has a mixture of borderline, antisocial, and narcissistic features.
Dr. A is in the early stages of recovery. He is not safe to practice medicine at this point. However, we are not prepared to say that he will never be able to practice again. Even though he has a guarded prognosis because of his complex comorbidity, he appears to be motivated to make significant changes in his life. We suggest that he embark on a comprehensive rehabilitation program and not consider working in medicine again until he has maintained sobriety for 12 months.
We recommend the following components to his rehabilitation and treatment plan:
He should attend three AA meetings per week and continue working with his sponsor.
He should have at least twice-weekly urine drug screenings.
He needs continued treatment in his halfway house for the time being.
He and his wife should seek weekly marital therapy, because many of the triggers that cause him to use substances are related to marital issues.
He should establish himself with a primary care doctor who prescribes his medication.
He should do no self-prescribing whatsoever.
He should establish himself with a psychiatrist who can prescribe mood stabilizers. We suggest moving his divalproex to bedtime and tapering his doxepin. Trials of lithium and perhaps lamotrigine should also be considered. Whoever is responsible for his psychopharmacological treatment should feel free to call L.M. to discuss with her his optimal treatment.
He should also be in long-term individual dynamically oriented psychotherapy to deal with the way that his traumatic childhood has affected his personality. He needs to focus on integrating the fragments into a more coherent identity, anticipating the consequences of his actions, increasing his capacity to mentalize, and finding ways to process feelings without using substances. Although we do not see him as having a sexual disorder, he certainly needs to focus as well on his relationships with women in the course of his psychotherapy.
If the rehabilitation plan goes smoothly and he remains free from relapse, he could begin work in an institutional setting. An ideal situation would be a state hospital, a prison setting, a Department of Veterans Affairs hospital, or similar institutions. He should be careful to keep regular hours so that his sleep cycle returns to normal. He should have only one job and avoid any moonlighting. He should also continue the same monitoring process that we have recommended under the auspices of his state’s physicians’ health program.
Received July 31, 2002; accepted Dec. 1, 2002. Based on an evaluation conducted at Baylor Psychiatry Clinic, Baylor College of Medicine. Address reprint requests to Dr. Gabbard, Department of Psychiatry, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030; firstname.lastname@example.org (e-mail).