Published 2024
In this module:
- Primary care physicians as a source of women’s health services
- Primary care physicians as a source of behavioral health services
In most medical specialties, care is provided by teams that often include physicians, nurse practitioners, physician assistants, and a variety of other occupations. Separate modules document how the Health Workforce Simulation Model (HWSM) models supply and demand for physicians, advanced practice providers, behavioral health providers, and other health occupations. Health workers often are defined by both their occupation and specialty.
In this module we discuss two topics not covered in other modules:
- The role of primary care physicians as a source of women’s health services
- The role of primary care physicians as a source of behavioral health services
The focus on physicians is in large part due to data limitations with quantifying the role of other care providers (e.g., nurse practitioners and physician assistants) in delivering care outside their typical scope of practice.
Primary care physicians as a source of women’s health services
Both primary care physicians (PCPs) and obstetrician-gynecologists (OB-GYNs) offer a wide range of similar women’s health services, including birth control, cervical cancer screening, prenatal care, treatment of vaginal or urinary tract infections, various health screenings, and management of chronic conditions during pregnancy.1 Both specialties are also capable of referring patients for mammograms and coordinating care with other specialists as needed.
While OB-GYNs specialize in women’s health, many patients choose to receive these services from their PCPs for several key reasons:
- Convenience: Women's health services can often be addressed during the same visit as other primary care needs.
- Trust: Patients may have established trusting relationships with their PCP and feel comfortable discussing their women’s health concerns with them.
- Access: Limited availability of OB-GYNs may make PCPs the most accessible option for women’s health services.
All board-certified PCPs receive basic obstetrics and gynecology training during their residency programs, with some family physicians pursuing more advanced obstetrics training. However, the number of family physicians receiving additional obstetrics training has been declining.2 3 Among family physicians who recently completed residency and expressed an intention to include obstetrics in their practice, 69% reported that they currently do so, with these practices more commonly located in small, rural, or isolated communities.3 In a survey of rural hospitals (n=185), family physicians delivered babies in 67% of facilities, and in 27% of those hospitals, they were the sole providers of maternity care.4
These findings emphasize the vital role that family physicians play in providing obstetric care, particularly in rural and underserved areas. Additionally, PCPs are crucial in addressing health disparities faced by women. This analysis seeks to quantify the percentage of time PCPs devote to women’s health services, convert this into FTE, and assess the implications for projecting future demand for PCPs.
The updated analysis in 2024 integrates national data from the Medical Expenditure Panel Survey (MEPS) with insights from the 2015-2016 National Ambulatory Medical Care Survey (NAMCS), focusing on metropolitan statistical area (MSA) and non-MSA comparisons. MEPS, a nationally representative survey of the civilian non-institutionalized population, provides estimates of healthcare utilization and expenses across the U.S. To increase the robustness of our findings, we pooled three years (2019-2021) of the most recent MEPS data, allowing us to analyze patterns in women’s health services provided by primary care specialties. While MEPS offers valuable insights into visits by physician specialties and non-physician providers, its utility for analyzing advanced practice providers is limited due to data constraints. Notably, the public MEPS data does not include MSA/non-MSA identifiers, though this information is available in the restricted MEPS file, which could be accessed in future studies.
We assessed the number and proportion of office visits related to women’s health conditions, identified through ICD-10 diagnosis codes5 from the MEPS medical conditions file. This approach allowed us to estimate visits by provider specialty and flag those that involved women’s health services. Additionally, any visit involving a mammogram was included in the women’s health category. However, unlike NAMCS, MEPS lacks other indicators for women’s health-related exams and screenings, potentially leading to an underestimation of women’s health services provided during primary care visits. When comparing MEPS data to the 2015-2016 and 2018-2019 NAMCS, we estimate that MEPS captures 68-72% of women’s health services, suggesting an underestimation of about 28-32%. Accordingly, the FTE estimates of PCPs providing women’s health services have been adjusted to reflect this undercount.
Using the combined data, we aggregated the total number and percentage of office-based visits for any women’s health condition by primary-care physician specialty. Summing visits across all females and using MEPS sample weights, we calculate that, overall, 2.5% of PCP visits by women and female adolescents have the flag for a women’s health diagnosis (Exhibit IX-1). This figure encompasses visits solely for women’s health as well as those involving a combination of women’s health and other categories of care.
The proportion of women’s health visits varies by specialty, with osteopathy within family medicine having the highest percentage at 10.3%, pediatrics at 1.4%, and family medicine and internal medicine ranging from 2.8% to 2.2%, respectively. When considering all PCP visits, including those for male patients, about 1.4% of total visits had a flag for women’s health services. The proportion varies by type of primary care specialty with a high of 10.3% care attributed to osteopathy among the family medicine specialties, 1.4% to pediatrics and around 2.8% -2.2% to family medicine and internal medicine respectively. Across all visits (including visits by men), about 1.4% of total visits to PCPs are for women’s health services.
The analysis further breaks down women’s health visits for female patients by specific primary care specialties and patient age, as illustrated in Exhibit IX-2. The age distribution of women’s health visits varies across specialties, with family medicine providing a higher proportion of women’s health services to younger patients (under 18 to 44 years old) compared to internal medicine. These findings are consistent with our earlier analysis based on the 2015-2016 NAMCS data.
We calculated FTE estimates for PCPs providing women’s health services by applying the above proportions to family medicine and general internal medicine to physician supply estimates in HWSM, applying estimates from the 2018-2019 NAMCS on the proportion of time spent on women’s health services during visits flagged for providing at least one women’s health service, and adjusting for the underestimation of women’s health services reported in MEPS. In particular, the underutilization scalar is multiplied by the percent of visit time spent on women’s health services to create an overall adjustment scalar. For example, the NAMCS analysis indicates that for visits to PCPs where a women’s health service was performed, on average 42% of the visit time was spent to provide women’s health services. This adjustment is reflected in our estimates of primary care FTE providing women’s health services. Based on these calculations, we estimate that in 2022, approximately 1,060 FTE family physicians and 690 FTE general internists provided women’s health services (Exhibit IX-3). This 1,750 FTE estimate accounts for approximately 1% of total (201,180 FTEs) physicians in family medicine or general internal medicine.
To estimate the breakdown of FTE by metropolitan status, we used findings from the 2015-2016 NAMCS analysis. In total, we estimate that around 1,420 primary care FTEs (family physicians and general internists combined) provided women’s health services in metropolitan areas, while approximately 330 primary care FTEs delivered women’s health services in nonmetropolitan areas.
Given the demand for obstetrical care in nonmetropolitan areas and the significant role family medicine physicians play in providing this care, continued training in obstetrics and the management of complex women’s health issues will remain a critical component of family medicine residency programs.6 7
A key consideration in modeling supply and demand projections is whether the estimated 1,750 FTE PCPs providing women’s health services should be factored into the projections for PCPs and/or OB-GYNs, as well as the starting year shortfall for PCPs. Similar to our analysis in the past, no adjustments were made to the HWSM projections to account for PCPs providing women's health services. This approach assumes that the demand projections for PCPs already account for services currently provided by them, including women’s health services. Additionally, one could argue that any extra demand for PCP time to provide WH services is already captured in the HPSA-derived starting year shortfall for primary care.
Primary care physicians as a source of behavioral health services
Primary care physicians are playing an increasingly important role in delivering behavioral health services.8 9 10 Due to the frequency of patient interactions, PCPs are often the first point of contact for individuals seeking care, placing them in a unique position to identify and address mental health concerns early on. Our analysis examines the role of PCPs in providing behavioral health services, which encompasses the following key activities:
- Screening: Primary care is frequently the entry point into the healthcare system. PCPs are pivotal in screening patients for behavioral health issues and determining the need for further treatment. The U.S. Preventive Services Task Force (USPSTF) recommends that primary care providers screen children, adolescents, and young adults for behavioral health conditions, including depression and anxiety.11 12 Additionally, USPSTF guidelines recommend screening and counseling for adults regarding excessive alcohol use and depression.13 14
- Treatment: In areas with limited access to behavioral health providers, particularly in medically underserved regions, PCPs might assume responsibility for direct treatment. This may include counseling, prescribing medications for mental health conditions, and managing opioid use disorder through medication-assisted treatment.10
- Collaboration within multidisciplinary teams: The integration of behavioral health into primary care practices is becoming more common. Increasingly, PCPs work within multidisciplinary teams that include behavioral health specialists, improving the delivery of comprehensive care for patients with mental health needs.15 16
Similar to our analysis of PCP time dedicated to women’s health services, this analysis aimed to estimate the proportion of PCP time spent treating patients with behavioral health disorders and to examine the implications for workforce modeling of both primary care physicians and psychiatrists. Due to data limitations, we were unable to estimate the proportion of time that primary care nurse practitioners and physician assistants spend delivering behavioral health services. Another objective was to explore whether the time PCPs spend on behavioral health care varies by MSA/non-MSA location.
The current analysis includes recent national information combining 3 years of the most recent MEPS (2019-2021) data to understand patterns in patient use of behavioral health services provided by primary care specialties. MEPS data allow us to explore visits by detailed primary physician specialties. The publicly available MEPS data does not include any MSA/non-MSA identifier, however, so we use information from prior analyses with the NAMCS to understand the MSA/non-MSA breakdown.
We assessed the number and proportion of visits in office-visit settings for any behavioral health-related condition. Behavioral Health related conditions were identified based on ICD-10 diagnosis codes17 available in the MEPS medical conditions file. We flagged all PCP office visits that could be linked to a behavioral health diagnosis code.
Aggregating the Visits data to national totals, using MEPS sample weights, we estimate that 5.6 % of visits to a primary care physician are flagged as involving behavioral health services. For comparison, the NAMCS analysis indicated that 12.0% of visits had a flag for behavioral health services provided. We think that under-reporting in MEPS of behavioral health services provided during visits to a primary care provider is caused by the construction and purpose of the MEPS data. While the purpose of NAMCS is to better understand what type of care is being provided, MEPS is focused more on healthcare utilization patterns and associated expenditures.
MEPS has a Conditions file that indicates whether the patient has a diagnosis for a medical condition (e.g., anxiety, depression). The purpose of this Conditions file is primarily to estimate disease prevalence. While the Conditions file is linked to the Visits file (as presence of a diagnosed condition is linked to having healthcare use to treat that condition), it is unclear to what extent all visits where a behavioral health service was provided are accurately linked to the Conditions file (particularly for patients who might have had multiple visits to their primary care provider during the year and received behavioral health care at more than one visit). The NAMCS data, on the other hand, should be accurate because each visit record in NAMCS is associated with data extraction of the information recorded in patient records of services provided during that visit. To adjust for this under-reporting, we scale up the behavioral health-flagged visits in MEPS by 2.14 (12.0%÷5.6%).
Exhibit IX-4 shows the distribution of visits for behavioral health related condition by detailed primary care physician type.
Exhibit IX-5 shows the unscaled distribution of behavioral health visits by primary care specialty and sex. When we apply scaling for under-reporting, we lack data to know if a different scalar should be applied to different patients or primary care specialties. Overall, the proportion of primary care visits associated with behavioral health condition is higher for females (54.9%) compared to males (45.1%).
In the MEPS, we cannot directly identify the proportion of the visit or amount of time during the visit that was spent providing behavioral health services. To account for this limitation, we used prior estimates from NAMCS of the average amount of visit time providing behavioral health services given that the visit has a flag for at least one behavioral health service provided—which average 45% of the visit time. This adjustment scalar is used to revise primary care FTE currently providing behavioral health services in 2022 (Exhibit IX-6).
We use the metropolitan status area breakdown from the previous NAMCS analysis (2018 & 2019 combined files) to back-out the physician FTE demand by metropolitan and nonmetropolitan areas. We applied the ratios to FTE estimates of primary care physicians from the 2022 American Medical Association PPD and calculated total primary care physician FTE required to provide behavioral health services in 2022. The results shown in Exhibit IX-6 indicate that overall, an estimated 18,820 primary care physicians FTEs are needed to provide behavioral health services (or about 6.9% of primary care physician time is spent providing behavioral health services). This breaks down to approximately 16,780 primary care physician FTEs in metropolitan areas and 2,040 primary care physician FTEs in nonmetropolitan areas.
The estimates of primary care physicians providing care for behavioral services are compared to the FTE supply of psychiatrists in 2022 and the results suggest that primary care physicians are providing approximately 26% of all physician-supplied behavioral health services in metropolitan areas. In non-metropolitan areas, they are providing about 46% of the physician delivered behavioral health care. This is in line with our previous findings with the NAMCS data which also found a disproportionate level of behavioral health care in the non-metropolitan areas is provided by the primary care providers. This puts excess burden on primary care physicians in additional to providing recommended preventive services and increases their workload.
There is a well-documented shortfall of behavioral health providers, and therefore for modeling purposes we add the 18,820 FTEs to the estimated demand for primary care physicians. This reflects the increased workload taken on by PCPs due to limited access to behavioral health providers resulting in an upward shift of the demand curve. Previously, the total increase in demand for PCPs was distributed between family medicine and general internal medicine. Based on the provision of behavioral health services in Exhibit IX-4, the increased demand for primary care physicians equates to additional 13,990 FTE demand for family physicians, 1,920 FTE demand for general internists, 30 FTE demand for geriatricians, and 2,880 FTE demand for pediatricians.