Secondary Logo

Journal Logo

Research Report

Family Practice Training over the First 26 Years

A Cross-Sectional Survey of Graduates of the University of Washington Family Practice Residency Network

Kim, Sara PhD; Phillips, William R. MD, MPH; Stevens, Nancy G. MD, MPH

Author Information
  • Free

Abstract

Primary care is essential to the health of North Americans. Family practice (FP) is an important component of primary care, providing a broad scope of services to patients, including those in urban and rural areas and underserved communities.1–4 Since 1969, FP has been recognized as a specialty with defined residency training requirements and board certification. As primary care and family medicine (FM) face challenges in today’s health care environment,5,6 information about the training and practice of family physicians is important. Such knowledge is important to understand manpower needs, to recruit doctors to (and sustain doctors already practicing in) communities of need, and to prepare family physicians with the knowledge, skills, and attitudes they will need for their future roles.7,8 Since 1997, the numbers of medical students selecting residency programs in FP and other primary care specialties has decreased.1 The future of the specialty and the provision of care to many communities across North America may depend upon understanding the realties of primary care practice.

Since 1972, the University of Washington Family Practice Residency Network (the network) has trained almost 2,000 family physicians. The network consists of 16 residency programs that serve both rural and urban areas across a four-state region: Washington, Alaska, Montana, and Idaho (WAMI). Of the 16 programs in the network, 12 are community based, two are in academic institutions, and two are in the military. Previous reports on network graduates through 1990 focused on the practice differences between graduates in rural and urban locations9 and on their career patterns.10 We report results from the most recent survey of network graduates, conducted in 2000, that focused on their location of practice, scope of practice, and adequacy of residency training.

With data on this generation of residency-trained family physicians, we asked several questions about the success of the network’s programs in meeting these important goals:

  • to produce family physicians who serve urban and rural underserved communities
  • to train family physicians to provide a broad scope of services
  • to provide services in diverse practice settings, and
  • to prepare graduates to care for all patients, with all problems, using a wide range of clinical and procedural skills.

Method

We conducted a cross-sectional survey of all physicians who graduated from the 16 FP residency programs affiliated with the network. The survey instrument included items based on the interests of the 16 residency program directors and experience from previous graduate surveys. The instrument asked about demographic information (age, gender, graduation year, practice location, organization), practice patterns (percentage of full-time, hospital patient care, maternity care, outpatient care, procedures), career satisfaction, and adequacy of residency training for practice in 40 topic areas. Overall, it contained 120 items and was five pages in length. The questionnaire was accompanied by a cover letter from the graduate program director stating the purpose was to evaluate and improve training. The questionnaire was mailed in February 2000, with a self-addressed, stamped return envelope, to all 1,498 individuals who graduated from network residency programs from 1973 through 1999. Addresses were drawn from the network’s database, which is updated regularly. If questionnaires were returned as undeliverable, current addresses were sought from the program and from the American Academy of Family Physician’s (AAFP) membership database. A follow-up survey was sent to all nonrespondents two months later. Respondents were given the option of an anonymous response, but only six chose not to share their names. The University of Washington Human Subjects Review Board granted this study an exemption. All names and information that would identify individual respondents were removed from the data before analysis.

All physicians’ responses were reviewed by a family physician investigator to identify the nature of their current practices. Respondents were classified as “currently in active FP” if they provided patient care to adults and children, did not practice solely in an emergency medicine or urgent care setting, and did not practice solely in another specialty.

Results

Graduate Profile

We received completed questionnaires from 983 of 1,498 graduates. A total of 114 questionnaires could not be delivered (five graduates had since died, 109 addresses were unavailable). The adjusted response rate was 71%. Approximately 70% of these 983 graduates practice in the four-state WAMI region. The response rate across the 16 residency programs ranged from 52.4% to 100%. The adjusted response rate was 50% or greater for each graduation year, with the exception of 1973 graduates (n = 1). Among the 983 graduates, 870 (89%) were currently in FP, 46 graduates (5%) practiced only emergency medicine or urgent care, and 67 graduates (6.8%) worked in other medical fields or had left medicine. Among the 870 who were active family practitioners, 843 graduates (97%) were currently certified by the American Board of Family Practice. Thirty-three graduates (4%) also held Certificates of Added Qualification in geriatrics and 16 graduates (2%) held certificates in sports medicine. Ninety-five respondents (9.7%) had graduated from one of the two military programs affiliated with the network. Of the active family physicians, 523 graduates (53%) practiced in the state of Washington, and 674 (69%) practiced in the states served by the network–Washington, Alaska, Montana, Idaho. Eight graduates (0.8%) practiced outside the United States.

All results reported below are based on the 870 graduates who were in active FP. Table 1 lists their overall profile by graduation year, gender, practice community size, and practice organization. We stratified graduation year into cohorts based on five-year intervals, with the exception of the 1973–1975 and 1996–1999 cohorts. Among the 870 family physicians, approximately 15% graduated during the 1970s, 30% during the 1980s, and 55% during the 1990s. The proportion of female physicians has increased steadily over the 30-year span (χ2 = 60.5, p < .001), with women representing half of the 1996–1999 graduation cohort. The location of graduates’ practice sites was classified by three community sizes based on conventions from previous studies: small (population of ≤ 25,000), medium (population of 25,001 to ≤ 100,000), and large (population of > 100,000). Thirty-seven percent of the graduates practiced in small communities, 23.2% in medium-sized communities, and 37.6% in large communities. A higher proportion of women than men practiced in communities over 100,000 (48.2% versus 31.2%, p < .001). A greater proportion of recent graduates practiced in small communities (approximately 60% of the graduates of the 1990s). Conversely, fewer family physicians from the recent cohorts practice in communities with a population of more than 100,000. The proportion of family physicians practicing in medium-sized communities is much the same for graduate cohorts over the years (approximately 16–26%). Network graduates practicing in the WAMI region were more likely than graduates practicing outside the region to practice in large communities (41% versus 29.5%), and they were less likely to practice in small communities (31.5% versus 49.1%) and in federally designated health care sites (27.1% versus 33.6%) (p < .05 for all).

Table 1
Table 1:
Profile of Active Family Physicians from One Family Practice Residency Network by Graduation Year, Gender, Practice Location, and Practice Organization, 2000

More family physicians practiced in single specialty (36.8%) or multispecialty (31.7%) group practices than in residency or university settings (10%), partnerships (two-physician practices) (8.0%), or solo practices (6.7%). Twenty-nine percent of family physicians reported that they practiced at least half their time in federally designated health care sites, such as one or more of the following: health-professional shortage areas (9.3%), community health clinics (8%), medically underserved areas (6.6%), or rural health clinics (5.3%).

Work Patterns

We first describe the percentage of full-time family physicians, stratified by gender and graduation year. The definition of full-time and the estimate of the percentage of full-time worked are based on self-reports. Seventy-three percent of graduates reported that they worked full-time (87.6% of men and 49.4% of women). These numbers include physicians in teaching settings for whom full-time work may not all be clinical work. Among female family physicians, a higher percentage of the most recent cohort worked full-time (63% among 1996–1999 graduates) and practiced in small communities (62% in areas of 25,000 or fewer people). Only 3% (21) of the respondents worked less than 50% time. There was no significant trend in percentage of time worked across the graduation cohorts.

Most family physicians provided care to patients in hospitals, both adults and children (see Table 2). Seventy-nine percent of family physicians provided hospital care to adult medical patients, 54% provide care to patients in adult intensive care units or cardiac care units (ICU/CCU), and 71% provided hospital care to children. A higher proportion of male physicians compared with female physicians provided inpatient care to adult medical patients (88% versus 72.1%), ICU patients (60% versus 45.1%), and children (76% versus 62.9%) (p < .05 for all). Family physicians practicing in small and medium-sized communities cared for similar proportions of their own hospitalized patients. Of family physicians who provided hospital care for each patient group, they cared for 93.2% of their adult patients, 75.6% of their adults in the ICU/CCU, and 89.6% of their pediatric patients. Family physicians in large communities provided hospital care to most of their patients, but that proportion was smaller than the proportion cared for by their colleagues in smaller communities. Fifteen percent of the respondents reported serving as hospitalists for other physicians, a proportion similar across community size.

Table 2
Table 2:
% (No.) of 870 Active Family Physicians from One Family Practice Residency Network Providing Care of Hospital Patients, 2000

We summarize the practice patterns of family physicians in other important areas of practice–namely, maternity care, outpatient services, and clinical procedures–and stratify these patterns by practice community size. For each activity, family physicians rated the adequacy of their residency training in preparing them for current practice.

Family physicians provided a full spectrum of maternity care (see Table 3). Most respondents provided prenatal care (67.2%), performed deliveries (62.8%), and assisted at cesarean sections (57.5%). Fewer family physicians performed cesarean sections as primary surgeons (12.4%). A greater proportion of respondents in small communities provided maternity care than did those in larger communities. A higher proportion of the family physicians in the 1996–1999 cohort than in other cohorts provided prenatal care (80%), delivered babies (75.6%), and assisted at cesarean sections (66.3%) (p < .001 for all). A higher proportion of female physicians than men tended to provide prenatal care (72.7% versus 64%), deliver babies (65.6% versus 61.3%), and assist at cesarean sections (58% versus 57.2%), although these differences were not statistically significant. Over 95% of the respondents reported that they were adequately or well prepared during residency training for these aspects of maternity care. Of the 12.4% who performed cesarean sections, most report that their residency prepared them well.

Table 3
Table 3:
% (No.) of Active Family Physicians from One Family Practice Residency Network Providing Care for Maternity Patients, 2000

When examining practice patterns in outpatient care, the majority of the physicians provided ambulatory care to adults (96.3%) and children (97.2%), and gynecology (96.6%), psychiatry and psychiatric medicines (95.5%), and orthopedics (95.7%). The proportion of physicians providing care in these areas was similar across community size. There were no significant differences in outpatient practice patterns by physicians’ gender or graduation cohorts. Over 80% of physicians reported their residencies prepared them adequately or well for providing care in each of these areas. Orthopedics was the only area in which many (20%) of graduates reported that they felt underprepared by their residency training.

The family physicians who responded performed a wide variety of clinical procedures (see Table 4). Almost all reported interpreting electrocardiograms (93.4%) and performed minor surgery (95.5%). Large percentages performed colposcopy (42.3%), flexible sigmoidoscopy (53%), vasectomy (44.6%), endotracheal intubation (36.2%), and dilation and currettage (D&C) (34.0%). About 20% performed treadmill stress tests, ventilator management, and placement of arterial and central lines. Fewer than 10% of the physicians performed LEEP/LETZ excision procedures (8.7%), colonoscopy (2.6%), upper gastrointestinal endoscopy (4.0%), and surgical abortion (6.3%). The following procedures were more often performed by physicians in small communities: colposcopy (49.5%), flexible sigmoidoscopy (58.9%), vasectomy (57.3%), endotracheal intubation (59.8%), and D&C (49.2%).

Table 4
Table 4:
% (No.) of Active Family Physicians from One Family Practice Residency Network Reporting Practice of and Adequacy of Training for Procedures, 2000

A slightly lower proportion of graduates practicing in the WAMI region than those practicing outside the region took care of their ICU patients (73.8% versus 80.1%; nonsignificant). A higher proportion of graduates practicing in the WAMI region provided prenatal care (69% versus 63.5%, p < .05), delivered babies (66.5% versus 54.6%, p < .005), and assisted with cesarean sections (62.8% versus 45.8%, p < .001). The procedures performed less frequently by graduates practicing in the WAMI region were treadmill stress test (20.6% versus 30.6%), intubation (32.7% versus 44.3%), ventilator management (16.8% versus 26.6%), and arterial line (12.9% versus 23.6%) (p < .05 for all). The only procedure they performed more often than graduates practicing outside the WAMI region was D&C (37.5% versus 26.2%, p < .005).

A significantly higher proportion of male than female respondents performed selected procedures: flexible sigmoidoscopy (66.2% versus 31.3%), treadmill stress test (31.9% versus 10.1%), vasectomy (62.5% versus 14.7%), intubation (43.9% versus 23%), ventilator management (23.1% versus 14.1%), arterial line (20.3% versus 9.5%), central line (22.3% versus 8.3%), and D&C (39.3% versus 25.5%) (p < .05 for all). Colposcopy was the only procedure performed by a higher proportion of female than male physicians (52.5% versus 36.0%, p < .001). Colposcopy was performed by 55% of the 1996–1999 cohort, compared with 18.3% of the 1973–1980 cohort (p < .001). Compared with physicians from the 1996–1999 cohort, physicians from the 1973–1980 cohort were more likely to perform flexible sigmoidoscopy (68.3% versus 45.6%) and vasectomy (61.1% versus 37.4%), (p < .001 for both).

For all procedures, the majority of physicians who performed them reported that their residencies prepared them adequately or well. However, for colposcopy, flexible sigmoidoscopy, and treadmill stress test, fewer than 75% of the physicians reported that they felt residency provided adequate preparation. Many of these physicians graduated before the introduction of these procedures and, therefore, received no specific training during residency. Family physicians who graduated after 1990 were much more likely to report that residency prepared them adequately or well for practice of these procedures: 86% for colposcopy, 83% for flexible sigmoidoscopy, and 78% for treadmill stress test.

Fifty percent of the respondents reported that they teach medical students, residents, and other health professionals in their practices.

Discussion

This is the largest published study describing the training and practices of FP residency graduates. Although these family physicians trained in one regional network, both their residency program experiences and their current practices vary widely, and they serve a diversity of communities spread across a great geographic expanse. Many serve the poor and underserved, both in urban and rural settings. FP residencies are modeled to prepare primary care physicians to meet the needs of all patients in all communities.4 Our data document the success of this model in producing and sustaining family physicians to fulfill these roles in practice.

We note several limitations of this study. We surveyed graduates from a large but regional network. Most graduates still practice in this region. This was a cross-sectional survey, not a longitudinal study. We report on differences between cohorts of physicians by their year of residency graduation, but we cannot examine trajectories of individual physicians over time. All results are based on physicians’ self-reports. We asked if they performed procedures but did not inquire about the frequency of their performing procedures. We also did not assess the quality or outcomes of those services.

Continuity of care is a key attribute of FP, and our data document that family physicians provided care to their patients in the hospital when they were acutely ill and at their sickest. Although we identified differences in the services family physicians provided according to their gender, graduate cohort, and location of practice, across all community sizes, the majority of these physicians practiced a broad spectrum of FP. Even in the largest cities, over 67% took care of adult patients in the hospital. Those who cared for patients in the hospital provided such care to almost all of their patients, regardless of age. More than half of physicians in the largest cities provided hospital care to their pediatric patients. In small communities, more than 85% of family physicians provided care to both adult and pediatric hospitalized patients. The majority of family physicians provided care for their hospitalized adults in the ICU/CCU in small and medium-sized towns, and almost 40% do so in the largest cities. There are increasing pressures in some systems to limit the role family physicians play in hospital care; however, our data document that most family physicians play an active role in this important aspect of care.

Concerns about maternity care as part of FP present challenges for the future and may be greater problems in other regions of the country. Malpractice liability, fear of lawsuits, interference with lifestyle, and time demands have been cited as barriers contributing to a decreasing trend in maternity care provided by family physicians.11–14 However, for the family physicians in our study, maternity care is still a vital part of FP. Most provided maternity care. Over 58% of practitioners in large cities delivered babies, and up to 69% in small communities did so (over 27% in small communities also did their own cesarean sections). These figures are higher than the proportion of family physicians providing maternity care reported in other studies.11,12 A majority delivered babies, regardless of graduation cohort, gender, size of community, or organization of practice.

The breadth of FP requires family physicians to acquire and maintain a large variety of clinical skills across the age, gender, and problem spectra of their patients.15,16 Some procedures, such as the interpretation of EKGs and minor office surgery, are provided by almost all family physicians. Others are provided frequently, such as flexible sigmoidoscopy, vasectomy, colposcopy, and D&C. Other procedures are offered by smaller numbers of family physicians, presumably depending upon the nature of their patients and needs of their communities. Respondents in smaller communities performed all procedures except LEEP/LETZ, interpretation of EKG, and surgical abortion, in higher proportions.

Our study suggests some regional variation in family physicians’ practice patterns. Sixty-three percent of all network graduates delivered babies, compared with 31% of family physicians generally across the United States, according to an AAFP national survey (p < .001).17 We cannot explain this difference solely by physicians’ gender or the sizes of communities in which our graduates practice. The Pacific Northwest has a reputation for attracting and supporting family physicians who choose to make maternity care part of their practices. Among family physicians who graduated from the network and practiced in the WAMI region, 67% delivered babies (397/597) versus 55% of those who later left the region to practice (148/271, p < .01). Both groups of graduates were more likely to deliver babies than were family physicians across the United States in the AAFP study (31%, p < .001). Further comparisons with AAFP national data showed that a higher proportion of our network graduates performed flexible sigmoidoscopy (53% versus 39%), vasectomy (45% vs. 30%), and treadmill stress testing (24% versus 14%), (p < .05 for all).18

Family physicians who completed their residency training in the network and remained in the WAMI region to practice were less likely to work in rural and underserved areas, provide hospital care, and perform a broad range of procedures than were graduates who left the region to practice. Those practicing in the WAMI region were more likely to provide maternity care. These findings suggest that the broad range of FP is sustainable across the country, not just in the Pacific Northwest region.

One challenge of FP residency training is to provide experience and instruction to adequately prepare physicians for a wide variety of practices. Although all FP residencies are required to follow-up with graduates to assess the adequacy of training, these data are not available at the national level for comparison. Our survey showed that the great majority of family physicians felt their residency programs prepared them for the needs of their current practices. For example, 98% felt their residency training prepared them to provide maternity care and deliver babies. Those family physicians who chose to perform cesarean sections likewise felt their residency training prepared them well for this task (97% for assistant and 82% for primary surgeons). This is a particularly important finding because there is evidence that the level of preparation during residency training influences the likelihood of practicing maternity care.13 In ambulatory care, the network’s graduates also felt that their residency training prepared them well for the care of adults, children, gynecology patients, psychiatry problems, the management of psychiatric medicines, and orthopedic care. Of all these areas, the graduates identified only orthopedics as an area where they might have benefited from improved training. FP training includes a wide variety of procedures, some of which have been introduced since many of the respondents to our study had finished their residency training and entered practice. Continuing education provides opportunities for physicians to add procedures and clinical skills to those they acquire during residency training.

Conclusions

This study provides a picture of FP among one large group of physicians who graduated from a network of residency programs. Studies from other training programs in other regions of the country will be needed to complete the picture. In further research, we plan to study the methods these graduates used to gain additional training following graduation. The variety and evolution of medical practice, community needs, and physicians’ interest require such strategies for adding new competencies. The changing demands of FP require continued study of services, procedures, and the care of special populations. Future studies could help explain changes and predict patterns by tracking the trajectory of physicians in their practices with longitudinal data analysis. Further research will be needed to measure the effectiveness of educational strategies, using objective measures of physicians’ performance10 and the assessment of patient outcomes.

The graduates in our study have demonstrated that residency training prepares family physicians to meet the demands of current practice and the needs of diverse communities. Individual programs in the network have focused on specific practice settings and emphasized selected clinical areas. These data and future studies can aid educators in designing residency curricula, teaching strategies, and evaluation mechanisms. The task is no longer to prove that residency training provides adequate preparation for FP. The challenge for the future will be to identify which elements of comprehensive care will be the core of FP and which will be enhancements offered by individual residency programs.

References

1.Pugno P, Schmittling GT, McPherson DS, Kahn NB. Results of the 2001 National Resident Matching Program: family practice. Fam Med. 2001;33:594–601.
2.Kahn NB, Schmittling GT, Ostergaard D, Graham R. Specialty practice of family practice residency graduates, 1969 through 1993. JAMA. 1996;275:713–15.
3.Green LA, Fryer GE, Dill DR. What do family medicine residency graduates do? J Am Board Fam Pract. 1996;9:149–51.
4.Phillips WR, Haynes DG. The domain of family practice: scope, role, and function. Fam Med. 2001;33:259–67.
5.Geyman JP, Bliss E. What does family practice need to do next? A cross-generational view. Fam Med. 2001;33:259–67.
6.Green LA, Fryer GE. Family practice in the United States: position and prospect. Acad Med. 2002;77:781–89.
7.Cantor JC, Baker LC, Hughes RG. Prepared for practice: young physicians’ views of their professional education. JAMA. 1993;270:1035–40.
8.Blumenthal D, Gokhale M, Campbell EG, Weissman JS. Preparedness for clinical practice: reports of graduating residents at academic health centers. JAMA. 2001;286:1027–34.
9.Baldwin LM, Hart G, West PA, Norris TE, Gore E, Schneeweiss R. Two decades of experiences in the University of Washington Family Practice Residency Network: practice differences between graduates in rural and urban locations. J Rural Health. 1995;11:60–72.
10.West PA, Norris TE, Gore EJ, Baldwin LM, Hart G. The geographic and temporal patterns of residency-trained family physicians: University of Washington Family Practice Residency Network. J Am Board Fam Pract. 1996;9:100–108.
11.Balaban DJ, Rosenthal MP, Ungemack JA, Carlson BL, Zervanos NJ. Obstetric care among family physicians in Pennsylvania: trends, association with residency training, and policy implications. J Fam Pract. 1990;31:281–86.
12.Sakornbut EL, Dickinson L. Obstetric care in family practice residencies: a national survey. J Am Board Fam Pract. 1993;6:379–84.
13.Kruse J, Phillips D, Wesley RM. Factors influencing changes in obstetric care provided by family physicians: a national survey. J Fam Pract. 1989;28:597–602.
14.Greenberg DM, Hochheiser LI. Family practice residents’ decision making regarding future practice of obstetrics. J Am Board Fam Pract. 1994;7:25–30.
15.Norris TE, Felmar E, Tolleson G. Which procedures should be taught in family practice residency programs? Fam Med. 1997;29:99–104.
16.Phillips WR. Diagnostic and screening procedures in family practice: past, present and future use. Arch Fam Pract. 1993;2:1051–57.
17.American Academy of Family Physicians. Facts about family practice: practice profile of family physicians by family practice residency completion, January 1, 2003. http://aafp.org/x763.xml. Accessed 2 June 2003. Leawood, KS: American Academy of Family Physicians, 2003.
18.American Academy of Family Physicians. Facts about family practice: diagnostic procedures performed and interpreted in family physicians offices by residency training, May 2002. http://aafp.org/x822.xml. Accessed 2 June 2003. Leawood, KS: American Academy of Family Physicians, 2002.     
© 2003 Association of American Medical Colleges