Comparison of Intended Scope of Practice for Family Medicine Residents With Reported Scope of Practice Among Practicing Family Physicians

Importance  Narrowing of the scope of practice of US family physicians has been well documented. Proposed reasons include changing practice patterns as physicians age, employer restrictions, or generational choices. Determining components of care that remain integral to the practice of family medicine may be informed by assessing gaps between the intended scope of practice of residents and actual scope of practice of family physicians.

Objective  To compare intended scope of practice for American Board of Family Medicine (ABFM) initial certifiers at residency completion with self-reported actual scope of practice of recertifying family physicians.

Design and Participants  Cross-sectional data were collected from a practice demographic questionnaire completed by all individuals applying to take the ABFM Maintenance of Certification for Family Physicians examination. Initial certifiers reported intentions and recertifiers reported actual provision of specific clinical activities. All physicians who registered for the 2014 ABFM Maintenance of Certification for Family Physicians examination were included: 3038 initial certifiers and 10 846 recertifiers.

Exposures  Initially certifying physicians vs recertifying physicians.

Main Outcomes and Measures  The Scope of Practice for Primary Care score (scope score), a psychometric scale, was calculated for each physician and ranged from 0 to 30, with higher numbers equating to broader scope of practice. Recertifiers were categorized by decades in practice.

Results  The final sample included 13 884 family physicians and, because the questionnaire was a required component of the examination application, there was a 100% response rate. Mean scope score was significantly higher for initial certifier intended practice compared with recertifying physicians’ reported actual practices (17.7 vs 15.5; difference, 2.2 [95% CI, 2.1-2.3]; P < .001). Compared with recertifiers, initial certifiers were more likely to report intending to provide all clinical services asked except pain management; this included obstetric care (23.7% vs 7.7%; difference, 16.0% [95% CI, 14.4%-17.6%]; P < .001), inpatient care (54.9% vs 33.5%; difference, 21.4% [95% CI, 19.4%-23.4%]; P < .001), and prenatal care (50.2% vs 9.9%; difference, 40.3 [95% CI, 38.5%-42.2%]; P < .001). Similar differences from initial certifiers were present when comparisons were limited to recertifiers in practice for only 1 to 10 years.

Conclusions and Relevance  In this study of family physicians taking ABFM examinations, graduating family medicine residents reported an intention to provide a broader scope of practice than that reported by current practitioners. This pattern suggests that these differences are not generational, but whether they are due to limited practice support, employer constraints, or other causes remains to be determined.

Family physicians are trained broadly to provide comprehensive continuing health care and are not limited by patient age, sex, or care setting. Robust primary care systems include comprehensiveness as a core component and these systems have demonstrated better health outcomes.^1,2 Comprehensiveness is a key attribute of primary care³ that aids its ability to achieve the triple aim of improving the experience of care, improving the health of populations, and reducing per capita costs of health care.⁴

Despite the known benefits of comprehensive primary care, research has documented narrowing in the scope of practice of family physicians. From 2000 to 2010, the percentage of family physicians providing maternity care,⁵ pediatric care,⁶ and women’s health services⁷ declined at least 10%. A single-state survey of the most recent 5 years of graduating family medicine residents conducted in 2004 and 2010 found that recent family medicine residency graduates reported providing less inpatient care (46% vs 68%) and nursing home care (22% vs 45%).⁸ No definitive cause for the narrowing of scope of practice has been determined, although possibilities include training, individual preference, or multiple external factors such as poor insurance reimbursement, employment contract restrictions, market incentives, or time constraints as a result of increasingly complex patients.^5-8 Most research has documented decreases in individual services, but few studies have examined multiple components of scope of practice.^8-10

The reduction in comprehensiveness of care has generated further discussion of whether certain components of care are still integral to the practice of family medicine.^2,5-8,11 A common assumption is that a narrow scope of practice is a result of lifestyle choices, especially of younger physicians,^5,8,11,12 yet little has been published about the intended practice of new family physicians.^13,14 Therefore, the aim of this study was to compare the intended scope of practice for initial board certifiers with the American Board of Family Medicine with the actual scope of practice reported by recertifying family physicians.

We used data supplied by physicians registering to take the American Board of Family Medicine Maintenance of Certification for Family Physicians examination in 2014. A practice demographic questionnaire was a required part of the examination application, creating a cross-sectional census of all initial certifiers and recertifiers each year. All questions within the questionnaire were required. The examination registration questionnaire included questions on race and ethnicity. Initial certifiers completed the questionnaire in their last year of residency or just after residency, and thus questions were asked about intentions to provide care, as well as future practice organization (if known). Recertifying physicians reported their current scope of practice, practice characteristics and organization, and practice locations. Intentions to practice and scope of practice questions included a series of individual clinical activities that represented the breadth of family medicine training. Some questions were asked with slight differences in each questionnaire, but variables were coded to reflect common constructs. For example, intention to provide maternity care was a yes/no question for initial certifiers, and recertifying physicians were asked how many deliveries they performed (eTables 1 and 2 in the Supplement). After answering common questions, recertifying physicians were randomly assigned 1 of 4 question sets: electronic health record functionality, patient-centered medical home practice features, payment models, or whether they perform specific procedures. All initial certifiers were asked identical questions about intentions to perform these procedures.

The American Academy of Family Physicians Institutional Review Board approved this study as exempt research.

To characterize the overall scope of practice for a given individual, the Scope of Practice for Primary Care score (scope score) was used.¹⁵ The score was calculated according to the sum of the number of clinical activities provided; the sum was then converted to a scaled score that was created with the Rasch model.¹⁶ The scope score ranged from 0 to 30, with a lower number indicating a smaller scope of practice. Practicing physicians who did not provide direct patient care were not asked practice characteristic questions and were excluded from the study. If a physician registered for both the spring and the fall examination, data from the fall questionnaire were used.

We characterized the sample with descriptive statistics. The main analysis compared reported scope items by years in practice, which were categorized in 10-year increments up to 30 or more years, with initial certifiers coded as having 0 years in practice. χ² Tests were used to compare demographic information between the cohorts and to test for differences in percentages reporting intentions to practice or actual practice of the clinical activities that comprised the scope score between physicians in various years of practice. Analysis of variance and t tests were used to compare the mean scope score between these cohorts. Overall percentages and means of all recertifiers vs those of initial certifiers were also compared. Cohen d was calculated to estimate the effect size for the difference in scope score between all recertifiers and initial certifiers.¹⁷ The study also tested a hypothesis that the intended practice patterns of older initial certifiers would be similar to those of practicing physicians. To test this, initial certifiers were categorized into physicians younger than 38 years and those aged 38 years or older, which represented the 90th percentile of the age distribution and roughly matched the average age of practicing physicians with 1 to 10 years of practice. Because employment contracts traditionally detail the scope of practice expected of the physician, it was also hypothesized that initial certifiers who already had a contract may have had a lower intended scope of practice than those without a contract because they may already have known they would not be providing certain services. All analyses were conducted with SAS version 9.3. All testing was 2-sided and analyses were corrected for multiple comparisons with the Hochberg procedure.¹⁸

The final sample included 13 884 family physicians, and because the questionnaire was a required component of the examination application, there was a 100% response rate. Of the respondents, 22% (N = 3038) were initial certifiers and 9% (N = 1223) had been in practice longer than 30 years (Table 1). Younger physicians and initial certifiers were more likely to be women, doctors of osteopathy, and international medical graduates. Younger physicians were also more likely to be in hospital-based practice settings and to be employed, rather than owners of their practice.

Scope score was statistically significantly higher for initial certifiers than for all categories of practicing physicians (17.7, initial certifiers; 15.5, all practicing physicians; difference, 2.2 [95% CI 2.1-2.3]; P < .001) (Table 2). The effect size for this difference, as measured by Cohen d, was 0.70. Among recertifying physicians, there was little variation in the mean scope score across the years of practice, with those 1 to 10 years in practice having the same scope score as those with 31 or more years in practice. For all individual clinical activities, initial certifiers reported a higher intention to provide these services than recertifying physicians, with the exception of pain management. In comparisons between initial certifiers and all practicing physicians, the largest differences between reported intentions to practice and actual practice included prenatal care (50.2% vs 9.9%; difference, 40.3% [95% CI, 38.5%-42.2%]), home visits (44.1% vs 9.3%; difference, 34.8% [95% CI, 33.0%-36.7%]), nursing home care (38.1% vs 16.3%; difference, 21.8% [95% CI, 19.9%-23.6%]), inpatient care (54.9% vs 33.5%; difference, 21.4% [95% CI, 19.4%-23.4%]), and obstetric care (23.7% vs 7.7%; difference, 16.0% [95% CI, 14.4%-17.6%]) (P < .001 for all comparisons).

The sample of recertifiers asked the procedure module questions had characteristics similar to those of other recertifiers (eTable 3 in the Supplement). Initial certifiers reported a greater intention to provide each of the listed procedures compared with practicing physicians (Table 3). In comparisons between initial certifiers and all practicing physicians, differences were observed in reported intentions to perform musculoskeletal ultrasonography (24.1% vs 3.3% [95% CI, 22.5%-25.6% vs 2.6%-3.9%]), insertion of long-acting reversible contraception (66.9% vs 10.2% [95% CI, 65.2%-68.5% vs 9.0%-11.1%]), and prenatal ultrasonography (20.6% vs 3.9% [95% CI, 19.2%-22.0% vs 3.2%-4.7%]) (P < .001 for all comparisons).

There were no statistically significant differences in mean overall scope score between initial certifiers younger than 38 years vs those aged 38 years or older (17.7 vs 17.6; difference, 0.1 [95% CI, −0.5 to 0.4]; P = .80) (Table 4). However, there were significant differences in individual clinical activities. Younger initial certifiers reported higher intentions to provide sports medicine (76.1% vs 63.8%; difference, 12.3% [95% CI, 6.8%-17.8%]), newborn care (81.2% vs 71.9%; difference, 9.3% [95% CI, 4.2%-14.4%]), and pediatric care (91.8% vs 87.8%; difference, 4.0% [95% CI, 0.2%-7.7%]), whereas older initial certifiers reported higher intentions to provide nursing home care (36.5% vs 51.3%; difference, 14.8% [95% CI, 8.9%-20.5%]), home visits (43.1% vs 52.5%; difference, 9.4% [95% CI, 3.6%-15.2%]), and inpatient care (54.0% vs 62.5%; difference, 8.5% [95% CI, 2.9%-14.2%]) (P < .05 for all comparisons). Comparisons of older initial certifiers with recertifiers in 1 to 10 years of practice indicated that the former still intended a broader scope of practice, as measured with the scope score (17.6, >38 years old; 15.3, 1-10 years in practice; difference, 2.3 [95% CI 1.9-2.7]; P < .001) (eTable 4 in the Supplement).

Of initial certifiers, 62% reported they had secured employment after residency (Table 5). Known future employment was not associated with a significantly narrower overall scope of practice (17.6 with employment contract; 17.8 with no contract; difference, −0.19 [95% CI −0.5 to 0.1]; P = .17). However, significant differences were observed in specific clinical activities, the largest being in prenatal care, home visits, and obstetric care. Comparisons of physicians with an employment contract and those with 1 to 10 years of practice showed that initial certifiers with an employment contract still intended a broader scope of practice (17.6 with employment contract; 15.3 with 1-10 years in practice; difference, 2.3 [95% CI 2.1-2.5]; P < .001), with the exception of pain management and pediatrics (eTable 5 in the Supplement).

According to data from all certifying and recertifying family physicians in 2014, initial certifiers reported intentions for a broader scope of practice than that reported by recertifying physicians across multiple clinical activities. In particular, they reported substantial differences for prenatal care, inpatient care, nursing home care, home visits, and women’s health procedures. To our knowledge, this is one of the first large-scale evaluations comparing the intentions of newly trained physicians with the services reportedly provided by practicing physicians. The benefits of family physicians providing a broader scope of practice may include lower overall health care costs and reduced hospitalizations,¹⁹ as well as increased availability of services in physician shortage areas. These findings suggest that graduating family medicine residents intend to provide a broad array of care commensurate with their training.

It is unclear whether reported intentions for broader scope of practice among initial certifiers indicate a desire for return to more comprehensive care or whether this intention has been long-standing but continuously unsupported by the available practice options after training. In a study conducted in 1992, family medicine residents planned to practice obstetrics at considerably higher rates than community physicians.¹³ The authors suggested that this heralded a resurgence in the practice of obstetrics by family physicians, but subsequent studies demonstrated continual declines,⁵ indicating these intentions either did not translate into practice or attrition was greater among already practicing physicians. Our findings that recently trained family physicians intend to have a broad scope of practice suggest that narrowing scope may be more strongly associated with the posttraining work environment than initial personal desire to limit certain clinical services. However, our comparisons of physicians with and without employment contracts showed that employment was associated with only small decreases in intentions to provide many services, which are likely balanced with other factors such as job location, work hours, and stress of the work environment.

If intentions for broad scope of care are not translating into more comprehensive care, the reasons must be elucidated. One possibility is that initial certifiers’ responses may have been influenced by residency training requirements, which are broad in scope but may not reflect the options available to employed physicians. A second possibility is that initial certifiers are overly aspirational in reporting practice intentions. Residents may be unlikely to report that they do not intend to provide a service they have spent 3 years learning. This possible reporting bias may vary by clinical service. With home visits as an example, training requirements necessitate the completion of a minimum of 2 visits²⁰; it seems unlikely that this would be a commitment sufficient enough that 44% of initial certifiers would report intentions to provide this service solely according to invested time. The residency training standards for family medicine necessitate that residents be exposed to a broad practice model; it is likely that most graduating residents consider themselves prepared and do intend to practice such a model after graduation.¹¹ Although residents may still be overly aspirational, the difference between intentions and the reported practice patterns of early-career physicians would be difficult to explain by this phenomenon alone.

Given that practicing physicians report a narrower scope of practice than graduating physicians intend, these new physicians may not find support in providing broader-scope practice in the workplace. New physicians may model their practice according to practicing physicians’ behaviors. Previous studies have associated faculty role modeling of broad-scope practice with increased student interest and performance, and there is reason to believe these influences extend beyond residency training.^21,22 Thus, even if payment or employment policies change sufficiently to support broad-scope practice, new physicians may still face the challenge that practicing physicians may not be willing or able to increase their scope of practice and support these intentions. Obstetric, inpatient, or emergency care can impinge on lifestyle, but having experienced colleagues to share call and offer guidance may make providing such services more tolerable. In light of these possibilities, there is no guarantee that if these graduating residents began employment where they could practice broadly they would enjoy doing so. One study of high-performing primary care practices showed that physicians can find “joy in practice” with task delegation and work-flow redesign, but the authors did not address whether this extended to practice areas beyond the ambulatory setting.²³

This study advances the literature on scope of practice by moving beyond individual clinical services and procedures to a global measure. Previous research either studied items individually^8-11 or created an ad hoc scope scale by summing the services queried.^19,24 The methodology of the scope scale used in our study allows flexibility to include different clinical activities and recalibrate the scale to provide a comparable measure of scope of practice.

Three areas of evidence support that the 2-point difference we found between initial certifiers and practicing physicians is meaningful. First, the calculated effect size for this difference was 0.70, which is generally considered medium to large.¹⁷ Second, in measures of quality of life across multiple scales, a minimally important difference was consistently found with a change of half a standard deviation.²⁵ Half the standard deviation of the scope score is 1.5, which supports that 2 points may be an important difference. Third, using an ad hoc scope scale, Bazemore et al¹⁹ found lower overall Medicare costs when family physicians reported performing more clinical activities. Although direct comparisons between their scale and the one used in our study are difficult, their findings suggest that a 2- to 3-point increase in our scale could be associated with lower-cost care. Moreover, underlying the 2-point difference, there were large and meaningful differences between certifiers’ intentions to provide maternity care and inpatient care, as well as treat patients in the emergency department and in their homes, compared with what recertifiers reported providing.

This study had several limitations. The data from the American Board of Family Medicine are self-reported and may be subject to recall or social desirability bias. However, data from the board’s questionnaires have provided results comparable to those of validated national surveys.²⁶ Second, our data are from only 1 year, and family physicians in other testing cohorts may differ. Initial certifiers, however, represent more than 90% of the entire national graduating class of family medicine residents. Third, our study does not address the preparedness of physicians to perform the activities indicated.

Further research should continue to examine subsequent years of initial certifiers to determine whether intentions to have a broad scope of practice remain. Tracking these intentions may make it possible to correlate them with health system reforms or alterations in family medicine residency training requirements. It will be important to follow these new family physicians’ practice patterns to determine whether their intentions were realized and, if not, why. Strengthening relationships between practicing physicians and certifying boards offers the opportunity to monitor training outcomes and individual practice activities over time.²⁷

In this study of family physicians taking American Board of Family Medicine examinations, graduating family medicine residents reported intentions to provide a broader scope of practice than that reported by current family physicians. This pattern suggests that these differences are not generational, but whether they are due to limited practice support, employer constraints, or other causes remains to be determined.

Corresponding Author: Lars E. Peterson, MD, PhD, American Board of Family Medicine, 1648 McGrathiana Pkwy, Fifth Floor, Lexington, KY 40511 (lpeterson@theabfm.org).

Author Contributions: Dr Peterson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Coutinho, Phillips, Peterson.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Coutinho, Peterson.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Coutinho, Cochrane, Peterson.

Administrative, technical, or material support: Cochrane, Phillips.

Study supervision: Coutinho, Phillips, Peterson.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Drs Phillips and Peterson are employees of the American Board of Family Medicine (ABFM). Ms Cochrane was an employee of the ABFM during the study.

Funding/Support: Dr Coutinho’s work at the ABFM was sponsored by the ABFM Foundation.

Role of the Funders/Sponsor: The ABFM Foundation had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.