In 1991, a physician satisfaction survey indicated that 13% to 19% of Northwest Permanente physicians had symptoms of burnout. However, the Medical Group's hospital-based Impaired Physician Committee was seeing only about two clients per year. Most of these clients had advanced substance abuse disorders, confirming the presence of a great unmet need for counseling.

In 1993, therefore, Northwest Permanente established an internal employee assistance program, the Physician Advocate Resource (PAR), a confidential counseling program by and for physicians that was designed to overcome their general reticence to seek mental health care. The present study examined the caseload of the PAR during a four-year period to characterize the ongoing need for such a physician counseling program in a large, group-model HMO.

During the study period, July 1993 through June 1997, the PAR saw 229 new clients, of whom approximately 70% were physicians and 22% were family members of physicians. The most frequent initial complaints (in 45% of clients) were stress, anxiety, and depression, equally divided between job-related and non-job-related causes; 24% of clients had marital or other family problems as their primary complaint. Most clients (58%) were self-referred to PAR. Physician clients were referred most frequently by general internal medicine departments and the least frequently by surgical departments. Physician clients were a mean 44.7 years of age and had worked a mean 9.2 years at the Medical Group. Male and female physicians were referred with equal frequency.

In-house counseling programs should be available for physicians, whose general reticence to seek help can be overcome if the program is confidential, physician-focused, and conducted in a supportive environment.

Introduction
In 1985, Northwest Permanente, the physician group associated with the Kaiser Permanente (KP) Northwest Division, developed a special committee to help physicians having psychoemotional problems and substance abuse problems. Service on this committee was a hospital staff function, and it operated as the Impaired Physician Committee, a format that had gained popularity at the time. The committee consisted of a chairman and four physician volunteers who offered support to peers with psychoemotional problems. When indicated, referrals were made for appropriate treatment.

Regular meetings were held before or after work to educate committee members and to discuss cases. Committee members were highly dedicated, but the committee's work faltered because of limited time for travel and meetings. Overriding clinical obligations of committee members also interfered. In its seven years of existence, the committee saw only 17 physician clients among a mean Medical Group population of 426 physicians. Eleven of these 17 clients had alcohol problems that were evident to peers. None of the 17 were self-referred.

This small yield of cases (about two per year) might have indicated that our physicians had few psychoemotional problems. However, results of a representative survey conducted among the 526 physicians in our Medical Group in 1991 (response rate of survey, 85%) suggested otherwise.¹ The survey found that 13% of our physicians could be considered "burned out" as measured by the Tedium Index, a well-established measure of burnout.² The survey also included questions asking whether the respondents believed themselves to be burned out. In response, 19% of physicians perceived they were "burned out" or "burning out."¹

Clearly, our committee was not meeting the emotional needs of our physicians, particularly in the area of work-related stress and burnout. However, to meet these needs, we would have to overcome a characteristic common among physicians: reticence to seek help. Self-sacrifice and "noble" stoicism appear to be norms of medical culture--norms which deny real needs and disallow healthy self-interest.³ We were determined to create an environment that recognized physicians' needs and encouraged physicians to seek necessary help. Our resource would be proactive, strictly confidential, and physician-focused.

Designing a Solution: The Physician Advocate Resource
With these criteria in mind, the Physician Advocate Resource (PAR) was established in June 1993 as an entity by and for physicians. To support this goal, the PAR functions within the Medical Group directly instead of being part of the broader hospital administration. Moreover, the PAR is not a volunteer effort that depends on time donated by busy practitioners; instead, it consists of salaried employees of the Medical Group. A critically important feature of the PAR is that one of the four PAR employees (0.5 FTE) is a therapist with a master's degree and experience in counseling physicians, employee assistance, treatment of mental and substance abuse disorders, and family therapy. Other staff includes a part-time physician-director experienced in addiction medicine and a part-time psychiatrist acting as assistant director. The two physicians chosen were long-time Medical Group members who are experienced in treating physicians, and these qualifications engender trust. The PAR clinicians share an on-call schedule and are accessible by pager, phone, and electronic mail. They are supported by a half-time confidential secretary.

The PAR is thus an employee assistance program whose function is to reach out to physicians, educate them, evaluate those in need of help, and refer them to identified competent counselors or programs. The nature of the PAR's services varies widely: In practice, a client may require only information or brief advice or instead may require extended counseling and long-term monitoring, especially for substance use disorders.

Administration of the PAR
From an organizational standpoint, the PAR was given particular legitimacy by being incorporated as a part of the Physician Health Committee, a standing committee of the Medical Group's Board of Directors: this structure established mental health as a component of physician well-being as well as a legitimate concern of the Medical Group. It tacitly gave
permission to ask for help. To address the potential conflict of interest created by the PAR being both part of the employer structure and acting as therapist, the PAR from its inception has been understood by the Northwest Permanente Board of Directors to represent physician clients primarily.

To maintain confidentiality, the PAR office and its records are situated away from main clinical and patient flow areas. PAR records are privileged and confidential by Oregon and Washington law under peer-review privilege statutes* in addition to other privileges which may be available, including psychotherapist-patient privilege, physician-patient privilege, or clinical social worker-patient privilege. PAR records are available only to PAR staff.

Initiating the PAR
Armed with additional legitimacy, confidentiality, and staff for the PAR, we initiated an outreach effort in June 1993 by issuing a letter of introduction describing the PAR and including a questionnaire seeking physicians' input, concerns, and needs. To get the message to the entire family, the letter was sent to physicians' homes and was addressed to "John [or Jane] Doe, MD and family." Other outreach efforts included presentations at departmental and staff meetings and at individual orientation meetings with new physicians.

Response to the PAR
Within two weeks of the initial mailing, the PAR added 11 clients to its caseload. Referrals to the PAR continued at a brisk rate during the subsequent four years, beginning in June 1993 (Table 1). During this period, physicians comprised about 70% of referrals; 22% were members of physicians' families (Table 1). Physicians seeking PAR services were aged a mean 44.7 years and had been employed by Northwest Permanente for a mean 9.24 years. The percentage of women physicians seeking PAR services did not differ substantially from the percentage of male physicians seeking these services.

Reasons for Referral to PAR
Reasons for referral to PAR included mostly stress, anxiety, and depression; of clients seen for these complaints, half perceived their symptoms to be job-related (Table 2). Marital and family issues were the next most frequent complaint; combined with the anxiety/stress/depression category, these accounted for almost 70% of cases (Table 2). The "Other" category included inquiries (eg, about insurance coverage, sources of outside therapy, and extent of confidentiality) or requests from physician administrators for advice regarding difficult physicians. Because evaluation by the PAR was brief, we did not categorize all cases according to the DSM-IV.⁴ Moreover, definite diagnosis was often made after referral to outside resources. We sampled the diagnoses of a clinical psychologist, to whom we referred 18 clients between 1995 and 1997. These clients were referred for "talk therapy" but required neither substance abuse therapy nor pharmacotherapy. The cases were predominantly categorized by the therapist as Adjustment Disorder (DSM IV 309.0) with or without anxiety and depression.

The counselor suggested that the categorization might be more straightforward if it identified problems as Marital (8 cases), Family (3 cases), Work (6 cases), or Other (1 case).

Sources of Referral to PAR
We have been especially pleased that most clients seen in the four years were self-referred (Table 3), and we have noted an increasing trend toward self-referral (70% of clients seen in the fourth year were self-referred). We believe this indicates an increasing confidence in acceptance of the PAR and a departure from the physicians' traditional reluctance to seek help. Administrative referrals originate from department chiefs, usually as a response to excessive patient complaints or to unacceptable physician behavior. The PAR's role in these cases is to identify treatable conditions and to recommend therapy that might preclude the need for disciplinary measures. Physicians named in liability cases were referred by the Region's Medical-Legal Department to the PAR for evaluation and stress counseling as needed. Other sources of referral were other health care professionals such as primary care practitioners and mental health practitioners. Family members and physician peers also make referrals, although these are few.

Specialty of Physicians Referred to PAR
Primary care physicians (ie, those in pediatrics, general internal medicine, family practice, and obstetrics and gynecology) accounted for the largest proportion of physicians referred to PAR, compared with physicians from other departments and based on total department population at risk (Table 4). The high numbers of physicians from the general internal medicine department is consistent with that department's high scores on burnout measures as reported in the 1991 physician survey. Least represented among those seeking PAR service were members of the surgery and surgical specialty departments (neurology, orthopedics, head and neck surgery, urology, ophthalmology). Indeed, members of surgical specialties scored lowest on the 1991 Tedium Index.

Effectiveness of Monitoring and Other Follow-up Activities
During the four-year period, many of our 229 clients were distressed, but only 13 had impairment that necessitated cessation of practice. Among these 13, impairment was caused by untreated alcohol and drug problems (6 clients), medical reasons (4 clients), or mental health problems (3 clients). The PAR acted as intermediary between therapists and the Medical Group by assuring that treatment would continue and by coordinating an appropriate return-to-work date. For some clients, PAR physicians prescribed and monitored use of psychoactive drugs. Of the 13 clients designated as impaired, six returned to work; seven remained permanently disabled. Coordination with the health professionals programs of the licensing bodies in Oregon or Washington has been a positive part of PAR's monitoring of clients impaired by dependence on alcohol or drugs.

Discussion
Despite the myth of invincibility--believed by patients and physicians alike--physicians do have human problems that require therapy. Some evidence suggests that the need among physicians might be especially great, given the nature of medical practice and the perfectionistic characteristics of those who choose a career in medicine.^3,5-7 Practice in a health maintenance organization may inherently carry a potential for burnout by permitting less physician autonomy than in other milieus,¹ but stress and burnout are found in diverse medical settings.^8-12 The number of clients seen at the PAR suggests that this newly available service is addressing a previously unmet need. However, the number of physicians who sought care outside the PAR during and before its four-year existence period is unknown. The need for the service may have recently become considerably more urgent because of the enormous recent changes in medical care delivery systems.

The large number of marital and family problems seen in this study population is similar to the number observed by others.^5-7 With this phenomenon in mind, future therapeutic and preventive efforts should focus on empowering physicians in their work environment and on strengthening their marital and family relationships.

Conclusions
The PAR filled an important need in the Medical Group by making counseling services available, particularly during this era of stress in medical practice. Whereas anxiety, depression, and burnout were the conditions most frequently noted at intake, our experience showed that physicians' traditional reticence to seek help can be overcome by outreach and by providing a confidential, physician-focused service. This observation is supported by the predominance of self-referral among the physician clients served by the PAR.

The number of physician clients varied greatly among departments. The reasons are unclear but could include circumstances of practice, differences in personality or other unknown factors. Understanding the reasons for these differences might be useful in determining solutions to stress and burnout among physicians.

Outcomes of the PAR experience to date are measured only in terms of physicians' satisfaction with the process and with the therapy received. Scientifically designed, controlled research is greatly needed to measure clinical outcomes of the PAR approach and to study other aspects of physician health.

*ORS 41.675;RCW4.24.250

References
1. Schmoldt RA, Freeborn DK, Klevit HD. Physician burnout: recommendations for HMO managers. HMO Pract 1994 Jun;8(2):58-63.
2. Pines A, Aronson E, Kafry D. Burnout: from tedium to personal growth. New York: Free Press; 1981.
3. Gabbard GO. The role of compulsiveness in the normal physician. JAMA 1985;254:2926-9.
4. American Psychiatric Association Task Force on DSM-IV. Diagnostic and statistical manual of mental disorders. DSM-IV. 4th ed. Washington, DC: American Psychiatric Association; 1994.
5. Vaillant GE, Sobowale NC, McArthur C. Some psychologic vulnerabilities of physicians. N Engl J Med 1972;287:372-5.
6. Krakowski AJ. Stress and the practice of medicine. II. Stressors, stresses, and strains. Psychother Psychosom 1982;38:11-23.
7. Gabbard GO, Menninger RD. The psychology of postponement in the medical marriage. JAMA 1989;261:2378-81.
8. Quill TE, Williamson PR. Healthy approaches to physician stress. Arch Intern Med 1990;150:1857-61.
9. Deckard GJ, Hicks LL, Hamory BH. The occurrence and distribution of burnout among infectious diseases physicians. J Infect Dis 1992;165:224-8.
10. Lemkau J, Rafferty J, Gordon R Jr. Burnout and career-choice regret among family practice physicians in early practice. Fam Pract Res J 1994;14:213-22.
11. Lloyd S, Streiner D, Shannon S. Burnout, depression, life and job satisfaction among Canadian emergency physicians. J Emerg Med 1994;12:559-65.
12. Fields AI, Cuerdon TT, Brasseux CO, Getson PR, Thompson AE, Orlowski JP, et al. Physician burnout in pediatric critical care medicine. Crit Care Med 1995;23:1425-9.