Minority physicians, who are more likely to serve poor, underinsured, and uninsured populations, improve access to care for minority groups.1–4 They are very important, therefore, in states such as California, which will become “majority minority” beginning in 2000.5 Ironically, since 1995, when California prohibited its state universities from using affirmative action, there has been a stunning drop in underrepresented minority applications, admissions, and matriculations in both public and private medical schools.6
California's 10.4 million Latinos are, by far, the largest of the state's “minority” populations.5 Concerned about the future supply of Latino physicians, the recently formed California Latino Medical Association (CLMA) asked the Center for the Study of Latino Health and Culture, of the Division of General Internal Medicine and Health Services Research at the School of Medicine, UCLA, to develop a method to gain an overall picture of Latino physician supply.
Using a composite method, we identified around 90% of the Latino physicians licensed to practice in California, explored their training sources, analyzed their presence in minority communities, and projected the state's future supply of Latino physicians. We discovered that California suffers from a major shortfall of Latino physicians, one that will worsen over the next 20 years unless the state quickly makes policy changes in undergraduate and graduate medical education. At the end of the article, we suggest policies and programs that may help California meet its need for Latino physicians.
In 1940, California had only four allopathic medical schools: the University of California (UC) Berkeley (later to become the independent health sciences campus of UC San Francisco), Stanford, the University of Southern California (USC), and Loma Linda. Between 1940 and 1960, the state's population nearly tripled7 as postwar Americans from other states (mostly white non-Hispanic) moved west to California and subsequently had lots of babies (the baby boom). But the numbers of medical schools and medical school graduates did not keep pace. (UC Los Angeles (UCLA), established in 1954, did not greatly affect the overall physician supply until after 1960.) Despite a surge in “imported” physicians—doctors who, after being educated in other states, joined the westward trek—policymakers still feared a physician shortage,8 and the University of California system established three new medical schools (UC Davis, UC Irvine, and UC San Diego) during the 1960s. To further increase the physician supply, California residency programs accepted more and more international medical graduates (IMGs).
After the baby boom came the baby bust, and by the mid-1970s, California experienced a sudden, sharp decline in white non-Hispanic population growth. Concerns shifted from physician shortage to physician oversupply.8,9 Policymakers talked of closing one of the UC medical schools, but never did.* They did, however, begin reducing the numbers of residencies open to IMGs.
In 1969, while Latinos comprised 10.5% of the state's population, they accounted for less than 1% of the state's medical school enrollments. In the fall of that year, new affirmative action policies resulted in the first sizable group of Latino students beginning their medical studies. Between 1969 and 1979, the number of Latino students ramped up quickly; by the end of the decade they comprised 8.4% of medical school enrollments.10 While Latinos were still underrepresented (Latinos had grown to comprise 19.2% of California's population in the same decade), the growth rate in their numbers had been so spectacular, far outstripping population growth, that it appeared that Latinos were on the path to population parity, at least in medical school enrollments.
Two dynamics changed that optimistic scenario. First, the medical school enrollment growth rate slowed considerably. Between 1980 and 1990, Latino medical school enrollments increased only marginally, up to 9.4% of medical school enrollments.10 At the same time, the state's Latino population grew extraordinarily. Fueled by high fertility and increased levels of immigration, the Latino presence grew to represent 24.6% of the state's total population. By 1999, the 10.4 million Latinos were 30.4% of the state's population, yet Latino representation in the state's medical schools had not materially changed from the 1979 level.
Thus, while the Latino population continues to grow, the supply of California-educated Latino physicians has failed to keep pace and, as we shall document, the number of foreign-educated Latino physicians immigrating into the state has actually fallen. And now, in the post—affirmative-action era, the number of Latino students entering the state's medical schools has shrunk. How will the medical needs of Latino Californians be served?
An accurate analysis of the Latino physician supply is difficult because the major sources of data used to identify Latino physicians are incomplete.
The American Medical Association Physician Masterfile lists every physician in California (and throughout the country). However, the masterfile reports the ethnicity of fewer than half of the physicians recently graduated.11 The California Medical Association (CMA) does provide data on ethnicity, but only of its members, and membership in the CMA, mirroring national trends, is low.11 It is estimated that fewer than 40% of California's physicians join the CMA, and among younger and minority physicians, the membership rates may be even lower, perhaps only 10% to 20%. Thus, using CMA membership lists to identify physicians may severely undercount Latino physicians.
The California Department of Consumer Affairs, Medical Board of California, does keep data on all licensed physicians in the state, but does not track ethnicity. California's medical schools do keep records of the ethnicity of their more recent graduates (as do other medical schools in the country), but these data generally do not cover physicians who graduated before 1975 or who were educated outside the United States. Finally, local Latino physician groups throughout the country keep lists of their members, but because membership in these groups is voluntary, the lists are incomplete.
Identifying Latino Physicians and the Places They Trained
Presented with these incomplete sources, we developed an algorithm to identify Latino physicians. From the Department of Consumer Affairs, we obtained the electronic file of all 74,345 physicians licensed in California (as of February 1999). After dividing the list into two groups—U.S. medical graduates (USMGs) and international medical graduates (IMGs)—we matched the USMG physicians to the Census Bureau's “heavily Hispanic surnames,” a list of 12,215 Spanish surnames, both common and not, that have been value-checked to correlate highly with Hispanic ethnicity.12
We then sorted the IMGs by country and medical school. Those who had graduated from medical schools in Spain and Latin America, including Brazil, were considered Latino, regardless of surname. An exception was made for ten medical schools in Latin America that are known to graduate significant numbers of U.S.-born students (e.g., the Autonomous University of Guadalajara and Ross University in the West Indies). Graduates of this small group of schools (labeled USIMGs, for U.S. international medical graduates) also had to be matched by Spanish surname to be counted as Latino.
After developing this initial list, we searched the small membership records of California's two separate Latino physician organizations and one Latino medical student alumni listing to identify Latino physicians whom we had not identified using the algorithm.
This method has its strengths and weaknesses. Graduates from Spain and Brazil may be considered Latino for some purposes but not for others. Spanish physicians communicate fluently and share many cultural similarities with their Latino patients. We also know, anecdotally, that many Brazilian physicians working in California's Latino communities learn enough Spanish to communicate fairly effectively. Since CLMA's goal was to identify physicians with cultural competence in medical care delivery, we preferred to err on the side of inclusion in the case of these two countries usually not considered “Latino.” In any case, excluding physicians from Spain and Brazil would have reduced the Latino physician total by only 4.6%.
While the Census Bureau estimates that its “heavily Hispanic surname” list captures about 90% of the Spanish-surname population,12 relying on the list may both underestimate and overestimate the population. On the one hand, not all Latinos in the United States have Spanish surnames. In particular, Latina women who marry non-Latinos may take their husbands' surnames. On the other hand, this list omits a number of certain Spanish surnames that are also common in different national origin groups (e.g., Italian, Portuguese, French, or even Romanian), called “moderately Hispanic” or “occasionally Hispanic” surnames. The apportioning algorithms that the Census Bureau has developed for these “moderately” and “occasionally” Hispanic surnames are useful only for estimating the number of Latinos present in a population of a particular surname, not for identifying the ethnicity of specific individuals. Because we felt that too much error would be introduced by adding all persons with moderately or occasionally Spanish surnames, we excluded them altogether, realizing we would exclude Latinos with these surnames.
An additional overall counting error is introduced because licensing data do not indicate whether a physician is practicing, in full-time administration or research, or retired. Without using a physician survey (which was beyond the scope of our project), we could not distinguish practicing from non-practicing and retired Latino physicians. Because we based our analyses on active licenses, we necessarily overestimated the number of practicing physicians.
In short, there is no way to count with 100% accuracy practicing Latino physicians. We believe that the method developed for this project was an efficient, economical way to identify around 90% of California's licensed Latino physicians, and that this was enough to give us some insight into the trends in Latino physician supply.
Analyzing Physicians' Locations
To analyze California physicians' locations, we looked at the ethnic composition (percentage Latino) of the zip codes listed by physicians for the purpose of receiving license-renewal information. The physicians do not specify whether this is a practice or a home address; any definitive statement about that would require a survey, which is beyond the scope of our research.
We limited the analysis of location to Los Angeles County to use relatively recent population data at the zip-code level. Although 1990 census data provided ethnic composition by zip code for the whole state, only Los Angeles County had later data—1998.13 Because there was so much Latino population growth and movement in the 1990s (an increase of 2.8 million, equivalent to the entire state of Iowa), we felt that a one-year lag in comparison data (1999 physicians and 1998 population in Los Angeles County) would introduce less error than would a nine-year lag (1999 physicians and 1990 population in the state of California). Additionally, given that the population of Los Angeles County is approaching 50% Latino (4.3 million Latinos out of a population of close to 10 million), physician location patterns can be effectively studied in this single geographic entity.
There are 276 zip codes in Los Angeles County, which we grouped into three categories of Latino presence: low Latino (0%-19% Latino), moderately Latino (20%-39%), and heavily Latino (40%-99%). We grouped the 20,895 physicians in Los Angeles County in 1999 into four categories: Latino USMG, Latino IMG, non-Latino USMG, and non-Latino IMG.
Projecting the Future Supply of Latino Physician
The baseline projection of Latino physician supply is fairly straightfoward because the number of Latino medical students in California and out-of-state medical schools has been virtually constant for nearly 20 years. We also assumed stable entry of Latino IMGs. We factored into this nearly constant number of graduates the number of Latino physicians (USMG and IMG) leaving practice. We have assumed that withdrawal from active practice occurs 40 years after graduation from medical school, approximately at age 65. This accords with previous estimates.1
The period 1986 to 1995 provides the baseline figures for average annual estimates of: (1) Latino USMG physicians graduating from U.S. medical schools (annual average = 86.1, of which 55.1 were from California medical schools); (2) Latino USMG physicians withdrawing from active supply (occurring 40 years after graduation from medical school); (3) Latino IMGs graduating from medical school (annual average = 24.6); and (4) Latino IMGs withdrawing from active supply (occurring 40 years after graduation from medical school).
The worst-case projection is based on recent trends (1994 to 1999) that may or may not be permanent: a sharp drop of IMGs entering the state to five annually, a 32% drop in Latino California USMGs (following a 32% drop in first-year matriculations6), and a 19% drop in Latino out-of-state USMGs (following a 19% drop in Latino first-year matriculations nationally14).
RESULTS AND DISCUSSION
Number of Latino Physicians
Among the 74,345 physicians licensed to practice in California in February 1999, we identified 3,578 Latino physicians. Thus, Latinos comprised 4.8% of all physicians in the state. By way of comparison, the 10,352,763 Latinos comprised 30.4% of the state's total population in 1999.
One way to put these numbers into perspective is to look at the physician-to-population ratio. In 1999, there were 70,767 non-Latino physicians and 23,719,715 non-Latinos in the general population, or one doctor for every 335 potential patients. In comparison, the ratio of Latino physicians to Latino population was one doctor for every 2,893 potential patients.
Because nearly half of all Latinos in California are immigrants from Latin America, it is instructive to compare this physician-to-population ratio with those of Latin American countries.15 Some Latin American countries have physician-to-population ratios close to that of non-Latino California: Cuba at 1:226, Uruguay at 1:268, and Argentina at 1:364. Mexico has a ratio that is nearly twice that of non-Latino California, at 1:593. Latin America's overall ratio is 1:649. California's Latino physician-to-population ratio, at 1:2,893, exceeds the ratio of every Latin American country, including Paraguay (1:1,463), Bolivia (1:2,156), and Nicaragua (1:2,247). Only one country in the western hemisphere, Haiti (a non-Latino country), exceeds the California Latino physician-to-population ratio, at 1:11,532.
Sources of Physicians
As described above, California's physicians come from three sources: they train in Californian medical schools, they train in other states' medical schools, or they train abroad and arrive in California as IMG residents. In 1999, only 25.7% of non-Latino physicians and 32.3% of Latino physicians had been educated in California. Despite a medical education policy, in place since the 1960s, to reduce California's dependence on physicians educated elsewhere, the state continues to rely heavily on “imported” physicians. The sources of importation are different, however. In 1999, over half (52.5%) of California's non-Latino physicians had been educated in other states (Table 1). But while one fifth (20.2%) of California's Latino physicians came from other states (primarily Massachusetts, Illinois, New York, Pennsylvania, and Texas), nearly half (47.6%) were IMGs, mostly from Mexico, Argentina, Peru, Puerto Rico, Spain, Colombia, the Dominican Republic, and Chile (Table 2).† In contrast, only 21.8% of the non-Latino physicians were IMGs (primarily from India, the Philippines, Canada, Iran, Mexico (for USIMGs), Taiwan, Korea, and Egypt).
This discrepancy between the supplies of Latino and non-Latino physicians has important implications. As mentioned above, when California decided to favor homegrown physicians, it chose to severely limit IMGs' access to residencies. This reduction in IMGs (which, interestingly, has taken place against the backdrop of a steep rise in IMGs throughout the rest of the country16) has had a disproportionate effect on the supply of Latino physicians.
Figure 1 shows the numbers of Latino USMGs (trained in California and out-of-state) and Latino IMGs who graduated each year from 1948 to 1996 and who were licensed to practice medicine in California in 1999. The date of graduation, obviously, is not the same as the date of first state licensure, but it provides a good picture of a critical stage of the Latino physician pipeline. We begin the figure in 1948 because so few Latino physicians entered the pipeline before then.
From 1948 to 1972, Latino IMGs outnumbered Latino USMGs by nearly four to one. But then two things changed California's reliance on IMGs as the major source of Latino physicians, the first positive, the second not. First, the number of Latino USMG physicians (from Californian and out-of-state schools) began to increase from an average annual graduation of four in the 1960s to an average of 70 by the early 1980s. For the period 1977 to 1982, there were about equal numbers of Latino USMG and IMG physicians entering the supply pipeline. In fact, the supply of all Latino physicians (USMG and IMG) into the pipeline was then at its peak, reaching a high of 163 physicians graduated in 1982. But then the state put into place the policy that began reducing the number of IMGs. The annual number of Latino IMG graduates fell sharply from a high of 93 in 1982 to only four in 1996. In the 1990s, Latino IMGs represented less than 10% of the Latino physicians entering the pipeline to the state. Although the numbers of non-Latino IMGs also dropped during this period, their decline has been masked by the large numbers of out-of-state USMGs entering the state. In contrast, the loss of Latino IMGs has not been compensated for by increases in either California- or out-of-state—educated Latino physicians.
Location of Physicians
We grouped the 20,895 physicians in Los Angeles County into four categories: Latino USMG, Latino IMG, non-Latino USMG, and non-Latino IMG. The address locations of the four groups are given in Table 4.
Non-Latino USMG physicians were the most likely (55.6%) to list a licensing address in a low-Latino zip code (0%–19% Latino), followed by non-Latino IMGs (43.6%). By contrast, Latino physicians were far less likely to list addresses in low-Latino zip codes (35.2% of Latino IMGs; 32.5% of Latino USMGs). As predicted by numerous studies of the intentions of fourth-year medical students to practice in minority areas, higher percentages of Latino than non-Latino physicians listed addresses in the heavily Latino zip codes: 41.6% of Latino USMGs and 40.8% of Latino IMGs, compared with 20.0% of non-Latino USMGs and 27.0% of non-Latino IMGs. Though the licensing address is not always a practice location, these data suggest that, by increasing the number of Latino physicians in the state, one increases the number of physicians likely to reach Latino patients.
Projections of Latino Physician Supply
There has always been a discrepancy in the number of Latino physicians in California relative to the Latino presence in the general population, and under current policies and conditions, that discrepancy is likely to grow even greater soon due to three trends: (1) the number of Latino IMGs entering the pipeline has been sharply reduced; (2) the number of Latino USMGs has been virtually stagnant and may fall up to 32% if recent matriculation trends continue; and (3) the Latino population will continue to grow from its current level of 10.2 million to 14.0 million by 2010 and 17.8 million by 2020. To see the interacting effects of these three trends, we project the supply of Latino physicians to the years 2010, 2015, and 2020.
Baseline projection. The baseline projection based on the estimated components of change described above shows a slow build-up in the supply of Latino USMG physicians, rising from 1,876 in 1999 to 2,786 by 2010 and 3,269 by 2020. This 74.3% gain in Latino USMG physicians, however, is moderately offset by a decline in the number of Latino IMGs. Increasingly, more Latino IMGs withdraw from than enter into the Latino physician supply. The number of Latino IMGs drops from 1,702 in 1999 to 1,422 by 2010, before it finally falls to 1,126 by 2020—a 33.8% drop in the Latino IMG supply. Overall, under these assumptions, the Latino physician supply (both USMG and IMG) will grow over the next 20 years by nearly 30%, from 3,578 in 1999 to 4,209 by 2010 and 4,396 by 2020 (Figure 2).
This modest gain in Latino physician supply, however, is more than offset by the projected 77.3% growth in the Latino population over the same time period, from 10.3 million in 1999 to 14.0 million by 2010 and to 17.8 million by 2020.5 Expressed as a physician-to-population ratio, the current ratio of one Latino physician to 2,893 Latino potential patients will worsen to 1:3,317 by 2010 and to 1:4,044 by 2020.
Worst-case post-affirmative-action projection. Between 1995 and 1998, there was a 32% drop in underrepresented minority matriculations in California's medical schools.6 In addition, recent AAMC data for 1995 to 1999 indicate a 19% drop in Latino first-year matriculations nationally.14 And, while the baseline assumed a modest drop of Latino IMGs, the trend of the 1990s has been to almost “zero out” Latino IMGs. If this is a harbinger of matriculation levels to come (and not a temporary trend), the future looks bleak for the state's supply of Latino physicians. Assuming that 32% fewer California-educated Latino USMGs, 19% fewer Latino out-of-state USMGs, and an average annual five Latino IMGs enter the pipeline from 1999 onwards, the total projected Latino supply in 2020 will be 3,448, actually less than the current supply of 3,578. Factoring in the projected Latino population growth, by 2020 the Latino physician-to-population ratio will rise to 1:5,157, far worse than that in any Latin American country today.
CONCLUSIONS AND IMPLICATIONS
Latino Physician Shortfall
In an ideal world, we would not worry about the number of Latino physicians, for there would be enough physicians, well dispersed and well trained, to care for all Californians. But the world is not ideal. The distribution of doctors throughout the state is uneven—non-Latino physicians tend not to practice in heavily Latino areas. And very few physicians are trained in cultural competence—non-Latino physicians, for example, are rarely fluent in Spanish.
For the Latino population to enjoy the same physician-to-population ratio enjoyed by the non-Latino population, there should have been 30,887 Latino physicians active in the state in 1999. However, there were only 3,578, a shortfall of 27,309 Latino physicians. While this “Latino physician shortfall” is strictly a heuristic tool, it provides an order-of-magnitude scale for the type of policy efforts that need to be addressed immediately. One may argue that a more rational system requires fewer than one physician for every 335 potential patients, but in the absence of other indicators, a movement towards parity between Latino and non-Latino ratios is a useful direction to take.
Quality of Care
While non-Latino physicians can successfully provide care to Latino patients, an increase in the Latino physician supply will most likely lead to an improvement in the quality of their care in a number of ways. First, given that Latino physicians are more likely to report license addresses in heavily Latino zip codes, an increase in the Latino physician supply will lead to a greater increase of physicians in underserved zip codes than could be expected from the current physician supply composition. Second, as the general trend is for Latino fourth-year medical students to lean more heavily towards primary care than do their non-Hispanic white classmates, an increase in the Latino physician supply will be likely to lead to more primary care physicians. Third, given a greater (although not without room to be improved) communication ability in Spanish found in Latino physicians, there will probably be improved linguistic access for the Spanish-speaking segment of the Latino patient population. These three expected outcomes of an increased Latino physician supply are also characteristics of medical care delivery that are likely to lead to increased access, greater rates of acceptance of medical protocols, higher rates of compliance, and possibly higher rates of return visits. These are, of course, preliminary speculations, but they appear to be logical consequences of an increased Latino physician supply.
Implications for Medical Education Policy
At least three major policies need to be implemented: (1) increase the supply of Latino USMGs; (2) increase the supply of Latino IMGs; and (3) increase the cultural effectiveness of non-Latino physicians.
Increase the supply of Latino USMGs. Grumbach's recent analysis of medical school patterns from 1990 to 19986 shows a drastic decline in underrepresented minority applications, admissions, and matriculations since 1995. While Proposition 209 and the regental rulings known as SP-1 and SP-2 (which forbade allocation of resources for reasons of ethnicity, gender, or other criteria) were cited as the main cause of this drop, the generally anti-Latino public sentiment as expressed in 1994's Proposition 18717 and 1998's Proposition 227 (a measure that banned bilingual education in the state) may well have played a role in discouraging Latino interest in higher education in general and medical careers in specific.
Past efforts in recruitment and admission were successful, but were not sufficient to keep pace with Latino population growth. New efforts at various levels need to be developed and implemented. Experience from 1969 to 1979, when USMG Latino graduates increased tenfold, shows that Latino applications can be greatly increased by focused recruitment and outreach efforts. Because of the early success, many of these efforts have been neglected in recent decades. These efforts need to be reinvigorated and expanded to include the community college and state university systems, where the bulk of Latino premedical students are found.
Admission policies need to be developed that can increase Latino admissions without running afoul of anti-affirmative action legislation. For example, the state, needing more physicians who speak Spanish (irrespective of ethnicity), can give an edge to applicants who speak Spanish (again, irrespective of ethnicity). Likewise, students with experiences in Chicano/Latino studies, Mexican studies, or Latin American studies could be given preference, irrespective of ethnicity. The achievement of academic success in these areas is open to all.
As one of the greatest predictors of practice location is residency location, more residencies involving heavily Latino populations need to be developed, implemented, and supported.
Increase the supply of Latino IMGs. Latino physician supply has been, in the past, heavily dependent on Latino IMGs. One obvious way to increase Latino physician supply is to reincorporate Latino IMGs into the pipeline. This needs to be approached with a certain amount of caution; the state has a responsibility to educate “its own,” and should not lean upon an increase in the Latino IMG supply in lieu of increasing the supply of California Latino USMGs. However, the realities of the Latino physician shortfall are so great that a judicious use of Latino IMGs might be in order.
Two possible ways to increase the number of Latino IMGs are targeting the “Latino-serving institutions” located in Mexico, Argentina, Peru, Puerto Rico, and Chile (that is, filling more of the existing residency slots with IMGs from Latin American medical schools) and temporarily increasing the number of IMG residency positions. Such an increase could be limited and heavily regulated, but nonetheless have an enormous effect. In Mexico, for example, where there is little private medical practice, the public-sector organizations (e.g., Seguro Social, Salud Publica) have been downsizing for over a decade due to recurrent economic crises. As a result, Mexico has a considerable supply of underutilized physicians. While not all of them are sufficiently trained, there are probably thousands who could meet California's standards. Preparing merely the top 10% of those physicians to work in California would nearly double the Latino physician supply in a matter of four to five years.
Increase the cultural effectiveness of non-Latino physicians. Success in treating Latino patients may have as much to do with the physician's linguistic and cultural understanding as with specific medical training. While Latino physicians often have a head start, they are generally of the opinion that cultural effectiveness is a learnable skill; virtually none believe that only Latino physicians can, or should, treat Latino patients. In fact, a recent book, Healing Latinos: Realidad y Fantasia. The Art of Cultural Competence in Medicine,18 was written by Latino providers to introduce their non-Latino colleagues to the elements of cultural effectiveness.
Cultural effectiveness training need not be confined to non-Latino physicians. Latino physicians as well often benefit from increased work in cultural effectiveness. By virtue of their living and training in Latin America, nearly all Latino IMGs will have linguistic backgrounds appropriate for Latino patients. However, many Latino IMGs come from upper-class backgrounds, which might impede some forms of communication with a largely blue-collar Latino population. Latino USMGs may be more familiar with the Latino blue-collar experience, but may not be fluent in medical Spanish.
Despite these gaps, very little cultural effectiveness training is offered in medical school curricula. A great amount of curricular development in the area of cultural effectiveness remains to be done. The basic population-based patterns of disease are scarcely known, and the algorithms needed to manage disease states with Latino patient bases have yet to be developed, then incorporated into the core curricula.
With communication so important in the management of chronic diseases, an ability in medical Spanish is becoming an important tool. The effects of Proposition 227, which banned bilingual education in California, may have degraded programs designed to build upon childhood Spanish abilities, yet the increase in the Latino population requires an ever-greater pool of providers who can communicate in Spanish. Medical school may not be the best place to learn the basics of Spanish, but unless basic Spanish course work becomes rewarded in the admission process, it may be the place where many non-Latinos will begin their Spanish training.
The competition for admission to California medical schools remains high. Improvement in the cultural effectiveness of all aspiring physicians could be improved by placing weight on previous course work in Chicano/Latino studies at an undergraduate level or experiences working in Latino community settings or in Latin American countries.
In conclusion, to improve the delivery of health care to California's fast-growing Latino population, the state must review those policies that affect the number of Latino USMGs and IMGs entering practice and must also increase the number of non-Latino physicians who are prepared to work with Latino populations.