There has been a long-standing and ongoing debate about how many physicians, in general, and psychiatrists, specifically, we need in the United States (1,2). Several models have been proposed to determine what this need may be (3,4). However, there are many complicating factors in developing models of physician (and psychiatrist) supply and demand. First, determining the medical need for psychiatrists might be useful in the abstract but is not practical. This is because the medical need is likely to be far greater than the supply of psychiatrists that can be supported in the economic and health care environment. Economic factors must be considered in any model of psychiatrist supply and demand. Additionally, mental health needs of the public are met by physicians other than psychiatrists and by other mental health professionals. Thus, the contributions of these sectors of the workforce must also be considered.

Before considering the factors that add complexity to psychiatric workforce analysis, it may be useful to begin with some basic data about the physician workforce as a whole and the relative place of psychiatrists in the total workforce. In deciding how many psychiatrists are needed or can be afforded, it is useful to start with how many there are at present, who they are, what they are doing, and comparing these data to historical trends.

Determining these numbers is not as obvious as might be thought. In this paper, we will use two major data sources that contain historical and current data on the U.S. physician and psychiatrist population: 1) the Physician Characteristics and Distribution in the U.S. published by the American Medical Association (AMA), a series of annual reports that began in 1963 with the Distribution of Physicians; and 2) data from the 2002 National Survey of Psychiatric Practice (NSPP) conducted by the American Psychiatric Institute for Research and Education.

Data from the AMA are updated through a yearly Physician's Practice Arrangement Questionnaire sent to physicians in the AMA Physician Masterfile. Data collected and published in the Physician Characteristics and Distribution include: 1) physician characteristics such as age, sex, school, year of graduation, international medical school graduate (IMG) status, etc.; 2) physician distribution by geographic location; 3) analysis of professional activity, including specialty; 4) primary care specialties of family medicine, general practice, internal medicine, obstetrics and gynecology, and pediatrics; and 5) physician trends from 1970 through 2000.

Although these data constitute the universe of physicians, an important limitation of this database is that a physician's specialty is self-designated. The AMA makes no attempt to verify any training or qualifications of individual physician specialty designation. A physician could, presumably, choose any self-designated specialty, and there would be no attempt to check training or any other credentials.

The questionnaire lists 179 specialties. In addition to general psychiatry, there are six subspecialties listed: addiction psychiatry, child and adolescent psychiatry, internal medicine (psychiatry), forensic psychiatry, geriatric psychiatry, and psychoanalysis. For statistical purposes, the AMA uses only 39 of the 179 self-designated specialties. Only child psychiatry and psychiatry are listed in the tables in the report. Despite its clear limitations, the AMA Masterfile is the best national database of physicians. There is no national medical license, and while state boards list physicians for each state, physicians may have licenses in more than one state. Moreover, the list of specialists certified by the American Board of Medical Specialties obviously doesn't include physicians who are not certified.

The 2002 NSPP was a self-administered questionnaire mailed to about 2000 randomly selected psychiatrists listed in the AMA Physician Masterfile. The final response rate was 57%, resulting in a final sample of 1,189. Responses were then poststratified for race and ethnicity to reflect the distribution of the AMA Physician Masterfile, resulting in a nationally representative sample of routine psychiatric practice in the U.S. While the NSPP data are limited in that the data do not represent the universe of psychiatrists, these data do provide information not otherwise available in the AMA data, particularly concerning the characteristics of routine psychiatric practice in the United States.

Total Numbers and Demographics

According to the AMA's 2002—2003 Physician Characteristics and Distribution in the U.S., there were 813,770 physicians listed in the Physician Masterfile in 2000. Psychiatry is the fourth largest specialty in the United States when child and adolescent psychiatry is included (45,615 physicians). Internal medicine is the largest specialty, followed by family medicine and pediatrics (see t1).

If child psychiatry is not included, psychiatry is the fifth largest specialty (39,457). However, for specialty identification in this paper, we include the subspecialty of child and adolescent psychiatry as well as all the new subspecialties in psychiatry.

Of all physicians, 76% were male, and 24% were female. In 1980, female physicians constituted 11.6% of physicians in the United States (5). Thus, the percentage has more than doubled in 20 years. Nearly two-fifths (39%) of residents were female in 2000, compared to 21.5% of the resident total in 1980 (5). Psychiatry has the fifth largest number of women in the workforce, behind internal medicine, pediatrics, family practice, and obstetrics/gynecology. These rankings remain the same for female physicians under the age of 35 (5). According to data from the 2002 NSPP, women constitute more than 30% of the psychiatric workforce, as compared to 19% in a similar study completed for 1988 to 1989 (6).

International medical graduates (IMGs) represented 24.2% of the total physician population in 2000. In 1980, IMGs accounted for 20.9% of total physicians and were more highly represented in nonsurgical specialties (3). Psychiatry has the third highest number of IMGs (12,009), following internal medicine (42,762) and pediatrics (17,115) (5). If general practice is included with family practice, psychiatry is the fourth largest specialty for IMG physicians (5).

Changes in Workforce Size

Using a ratio of physicians in patient care per 100,000 members of the total population, there has been a significant increase in physicians to population since 1965. The ratio has increased from 132 in 1965 to 234 in 2000, an increase of 77.3% (5).

For all physicians, the increase in percent change from 1970 to 2000 has been even greater, (143.6%) than from 1965 to 2000. This is probably due to the significant increase in class size, the number of U.S. medical schools in the late 1960s and early 1970s, and changes in immigration laws, which allowed more foreign physicians into the United States.

The percent growth in psychiatry has also been positive during the past 30 years, but there is a significant difference between child psychiatry and general psychiatry. Psychiatry has increased 86.7%, while child psychiatry has increased 194.6%. Thus, child psychiatry has grown more than the number of total physicians, while general psychiatry has grown at a slower rate (5).

The same relationship has held true over the past decade. In the years 1990 to 2000, the percent change in the number of total physicians has been 32.2%. Psychiatry has grown 12.7%, while child psychiatry has increased by 41.8% (5).

Comparing psychiatry to several other selected specialties over the past decade, one will find that obstetrics and gynecology has increased 19.4%, internal medicine 36.8%, and general surgery 4.7% (5). Thus, child psychiatry has increased more than many other specialties, but general psychiatry has not.

During the period of 1970 to 2000, the percent change for IMGs for total physiciansF was 244.2%. Again, child and adolescent psychiatry increased at a greater rate of 283.5%, and psychiatry (IMG physicians) increased at a slower rate of 139.0% (5).

Overall, there are 294 total physicians per 100,000 members of the population in the United States. This includes patient care physicians and those in administration. Patient care physicians constitute 234 per 100,000 members of the population. In the year 2000, the physician/population ratio for the total number of psychiatrists was 14.3 per 100,000 and 2.2 per 100,000 for child psychiatrists. Altogether, psychiatry has 16.5 per 100,000 members of the population in the United States (and possessions) (5).

As best as can be determined, it appears that the United States has a higher ratio of specialists to population than any other country. There is little published data in this area, but by contacting the National Psychiatric Association (and Royal College) in various countries, it appears that Canada and Holland have ratios closest to that of the United States, but other countries have lower ratios (see t2).

Although the ratio of physicians to general population can provide a rough estimate of the relative size of the workforce, it can be a misleading figure because the distribution of physicians across the country can be uneven (see t3). In fact, the geographic distribution of psychiatrists throughout the United States is highly variable. While New York and California have the most psychiatrists, Massachusetts (31.1) has the highest ratio of psychiatrists to population, and New York (28.2) has twice the ratio as California (14.9). Idaho (4.6) and Wyoming (5.9) have the lowest ratios. Additionally, there is a wide distribution of specialists, including psychiatrists within the states, but the level of analysis required to determine the distribution of these data are beyond the scope of this report.

Physician workforce measures usually include ratios of physicians to the general population and an analysis of how ratios have changed over the years. These changes have often been inappropriately interpreted to indicate either physician oversupply or undersupply. Using these data to decide shortages is an inadequate measure and has led to false predictions because, among other reasons, the data do not account for demographic or socioeconomic factors that influence the demand for physicians (7).

A useful way to supplement these data and consider the status of the psychiatric workforce is to determine psychiatrists' work activities and how they have changed over the years. We did this by assessing data from the 2002 NSPP and comparing this information to data from the 1988 to 1989 Professional Activities Survey (PAS), a similar study described in a previous paper (6). Both studies were self-administered questionnaires sent to randomly selected psychiatrists who were drawn from the AMA Physician Masterfile. For both surveys, responses were weighted in order to provide nationally representative estimates of routine psychiatric practice in the United States.

Results of the 2002 NSPP indicate that, overall, psychiatrists worked fewer hours per week in 2002, as compared to 1988 to 1989. In 2002, psychiatrists worked an average of 42.5 hours per week, and they averaged 48.2 hours per week from 1988 to 1989. There could be several explanations for this drop in hours worked, but demographic explanations are most likely. As mentioned earlier, more women have been entering the psychiatric workforce during the last 13 years. Additionally, the psychiatric workforce has been aging. For example, in the 1988 to 1989 PAS, 24% of the psychiatric workforce was under the age of 40. This percentage dropped to 8% in the 2002 NSPP. Further, the percentage of psychiatrists above the age of 55 rose from 32% to 46% from 1988 to 1989 and in 2002. In fact, the average age of respondents in the 2002 NSPP was 55.7 years old. Both women and older members of the workforce are groups that traditionally have been likely to work more part-time hours. Thus, these changing demographics may help explain the decrease in the number of hours psychiatrists work each week.

Findings from the 2002 NSPP indicate that psychiatrists spend the largest percentage of their time in direct patient care activities (60%). However, this is less than the percentage of time spent in direct patient care in 1989 (67%). Additionally, the percentage of time spent in administrative activities was 20% in 2002 and 12% in 1988 to 1989. Although psychiatrists were working fewer hours on average and spending less of that time in direct patient care in 2002 versus 1988 to 1989, they were seeing an average of 41 patients per week, as compared to 35 patients per week in 1988 to 1989. Given the fewer hours worked and more patients treated, it is not surprising that psychiatrists spent an average of 34 minutes with each patient in 2002 and 55 minutes per patient in 1988 to 1989.

In addressing the debate about whether the nation's psychiatric workforce can meet the needs of the population, it is important to assess the current size of the workforce and in what professional activities it is engaged. The number of psychiatrists and the ratio of psychiatrists to the population continue to grow, with child psychiatry growing faster than the supply of other physicians. By any measure, psychiatry is a significant part of the physician workforce in American medicine.

Despite the continued growth in absolute numbers of the psychiatric workforce, there are some trends that provide cause for concern about its capacity to meet the needs of the population. First, the psychiatric workforce as a whole is getting older. The average age of psychiatrists is older than 55 years old, and the percentage of psychiatrists under the age of 40 continues to decrease. In the near future, we are likely to see an increase in retirements among psychiatrists. In order to maintain the capacity of the psychiatric workforce, efforts to recruit medical students into psychiatry will likely have to increase as well.

Moreover, while the psychiatric workforce is growing, psychiatrists are working fewer hours per week on average and spending less time in direct patient care, yet they are treating more patients. Thus, in the future, it is possible that it will take larger numbers of psychiatrists to produce the output that a smaller number can produce today. While factors contributing to this trend are not clear, if it continues, more psychiatrists will be needed in the workforce in order to meet the demands of the population.

Additionally, there are substantial differences in both the number and ratios of psychiatrists across states. These geographic differences may have an impact on access to services, and they have certainly been used in scope of practice arguments. Despite these potential pitfalls, psychiatry is a major medical specialty that is growing steadily. The question remains, however, whether its rate of growth will be able to keep pace with the rate of demand for services.