Background
Contraception care is essential to providing comprehensive healthcare; however, little is known nationally about the contraception workforce. Previous research has examined the supply, distribution, and adequacy of the health workforce providing contraception services, but this research has faced a series of data limitations, relying on surveys or focusing on a subset of practitioners and resulting in an incomplete picture of contraception practitioners in the United States.
Objective
This study aimed to construct a comprehensive database of the contraceptive workforce in the United States that provides the following 6 types of highly effective contraception: intrauterine device, implant, shot (depot medroxyprogesterone acetate), oral contraception, hormonal patch, and vaginal ring. In addition, we aimed to examine the difference in supply, distribution, the types of contraception services offered, and Medicaid participation.
Study Design
We constructed a national database of contraceptive service providers using multiple data sets: IQVIA prescription claims, preadjudicated medical claims, and the OneKey healthcare provider data set; the National Plan and Provider Enumeration System data set; and the Census Bureau’s American Community Survey data on population demographics. All statistical analyses were descriptive, including chi-squared tests for groupwise differences and pairwise post hoc tests with Bonferroni corrections for multiple comparisons.
Results
Although 73.1% of obstetrician-gynecologists and 72.6% of nurse-midwives prescribed the pill, patch, or ring, only 51.4% of family medicine physicians, 32.4% of pediatricians, and 19.8% of internal medicine physicians do so. The ratio of all primary care providers prescribing contraception to the female population of reproductive age (ages, 15–44 years) varied substantially across states, with a range of 27.9 providers per 10,000 population in New Jersey to 74.2 providers per 10,000 population in Maine. In addition, there are substantial differences across states for Medicaid acceptance. Of the obstetrician-gynecologists providing contraception, the percentage of providers who prescribe contraception to Medicaid patients ranged from 83.9% (District of Columbia) to 100% (North Dakota); for family medicine physicians, it ranged from 49.7% (Florida) to 91.1% (Massachusetts); and for internal medicine physicians, it ranged from 25.0% (Texas) to 75.9% (Delaware). For in-person contraception, there were large differences in the proportion of providers offering the 3 different contraceptive method types (intrauterine device, implant, and shot) by provider specialty.
Conclusion
This study found a significant difference in the distribution, types of contraception, and Medicaid participation of the contraception workforce. In addition to obstetrician-gynecologists and nurse-midwives, family medicine physicians, internal medicine physicians, pediatricians, advanced practice nurses, and physician assistants are important contraception providers. However, large gaps remain in the provision of highly effective services such as intrauterine devices and implants. Future research should examine provider characteristics, programs, and policies associated with the provision of different contraception services.
Introduction
Contraception care is essential to providing comprehensive healthcare; however, little is known nationally about the healthcare workforce providing these services. The contraception workforce includes any type of provider that offers contraceptive care, whether reproductive health specialists (obstetrician-gynecologists, certified nurse-midwives, women’s health nurse practitioners, and women’s health physician assistants) or general primary care providers (family medicine physicians, internal medicine physicians, pediatricians, nurse practitioners, and physician assistants). Projections of obstetricians and gynecologists and “women’s health service providers” predict future workforce shortages in most states, with the South and Western regions of the United States facing the greatest shortages.
Why was this study conducted?
This study aimed to examine the difference in the composition and distribution of the health workforce that provides contraception, using a comprehensive national-level database with claims data and provider characteristics.
Key findings
We found a substantial difference in the geographic distribution, provider specialty and type, Medicaid acceptance, and service provision of the contraception workforce.
What does this add to what is known?
To our knowledge, this study is the largest and most comprehensive study of contraception providers to date and includes multiple specialties at a single time point. Our study strengthened previous studies by allowing a national examination of the contraception workforce, including county-level geographic distribution of reproductive health and primary care providers with demonstrated prescription, contraception service, and Medicaid patients, illustrated through claims data rather than self-reports of service.
Access to contraception services is dependent on both the local availability of healthcare providers and the types of contraception services offered by those providers. Reproductive health and primary care specialties vary in their provision of contraception care, particularly for highly effective methods such as intrauterine devices (IUDs) and implants. In surveys, nearly all obstetrician-gynecologists reported providing IUDs, whereas ≤40% of family medicine physicians reported inserting IUDs. , In addition, pediatricians recommend IUDs and implants at a low rate, with few reporting that they believe IUDs and implants are appropriate for their patients.
Research on the supply, distribution, and adequacy of the health workforce providing contraception services has largely relied on surveys, often with limited sample sizes and focused on a subset of providers, in particular the supply and demand of reproductive health or “women’s health” specialists. Even when primary care contributions are taken into account, they are often based on estimations of family medicine physicians’ time spent on women’s health services. Primary care providers are important contraception providers, but little is known about this workforce at the national level. Given these data limitations, an incomplete picture of the contraception workforce in the United States persists.
In addition to provider distribution and provision of contraception services, provider choices, such as Medicaid participation, affect the access to contraception care. Medicaid and the Children’s Health Insurance Programs provide health insurance coverage to nearly 73 million Americans, with more than half of them being women. However, the acceptance rates of Medicaid are lower than that of Medicare and private insurance and vary by state policies, such as Medicaid expansion and state Medicaid payment rates. When healthcare providers accept Medicaid, patients are more likely to access contraceptive services ; conversely, when local healthcare providers do not accept Medicaid, the access to contraception care is limited further, especially in underserved communities.
This study aimed to describe the contraception workforce in the United States and address the gaps in previous literature. We combined multiple national data sources to construct a comprehensive database of the contraception provider workforce and examine the difference in supply, distribution, the types of contraception services offered, and Medicaid participation.
Materials and Methods
Data sources
Our main data sources came from IQVIA, a proprietary health information company that aggregates data from multiple sources for healthcare analysis and business solutions. We used 3 IQVIA data sets: prescription claims, preadjudicated medical claims, and the OneKey healthcare provider data set. The prescription claims data set includes an estimated 92% of all US retail prescription claims. The medical claims data set includes an estimated 84% of American Medical Association (AMA) registered physicians nationally. We obtained full-year 2019 provider month-level counts of prescription claims for contraceptive products and procedure and diagnosis codes associated with contraceptive services, including depot medroxyprogesterone acetate (shot), IUDs, etonogestrel implant, and encounter for contraception management (Z30) for the year 2019. All 3 data sets included provider National Provider Identifiers, provider types, and addresses.
Using the 3 data sets, we constructed a database of contraceptive service providers. In addition, we used the National Plan and Provider Enumeration System (NPPES) data set (2020) to identify provider type and location (state and county) of all active providers and the 2019 American Community Survey to determine the number of reproductive-age women (ages, 15–44 years) at the county level. , We recognized that women are not the only people who need contraception services and sought to be inclusive of all genders. However, we chose to use the most comprehensive and readily available data, which use a binary gender indicator.
Data analysis
We identified all providers who prescribed the birth control pill, patch, or ring in the 2019 prescription claims. We excluded providers who prescribed <10 total contraception prescriptions during 2019, as they were rare contraception prescription providers, possibly providing contraception prescriptions outside of the scope of their normal practice. The proportion of each specialty and profession prescribing these contraceptive methods was calculated on the basis of the total number of providers by specialty and profession in the prescription claims.
We identified providers who had and did not have associated Medicaid contraception prescription claims and calculated the percentage of each state’s contraception providers who serve Medicaid patients. Medicaid patients had the following coverage: fee-for-service Medicaid; managed Medicaid; managed Medicaid, Medicare supplement, Medigap, and state assistance; and the Children’s Health Insurance Program.
We identified providers of in-person contraception by submitted claims for the shot, IUDs, and implants, using Current Procedural Terminology and Healthcare Common Procedure Coding System codes. For in-person contraception services, we did not apply a minimum number of services for inclusion as these procedures are less likely than prescriptions to be provided out of scope. We focused the analysis on a subset of 61,124 contraception providers in the medical claims; the subset included providers who had full or partial volume of claims (compared with low or minimal volume of claims) and who had billed a contraceptive management code for at least 1 visit in the calendar year 2019. There was no advanced practice nurse, nurse-midwife, or physician assistant in this subset of in-person contraception providers.
For all analyses, we grouped providers by provider type and specialty: obstetrician-gynecologists, nurse-midwives, family medicine physicians, internal medicine physicians, pediatricians, advanced practice nurses, and physician assistants. We excluded providers in US territories using NPPES data to determine the location.
All statistical analyses were descriptive, including chi-squared tests for groupwise differences and pairwise post hoc tests with Bonferroni corrections for multiple comparisons. Stata/MP (version 16.1; StataCorp LLC, College Station, TX) was used for all analyses. This study was approved by the George Washington University Institutional Review Board.
Results
Prescription contraception
Of 1,797,086 providers in our specialties and professions of interest, we identified 289,926 prescription contraception providers. Compared with 51.4% of family medicine physicians, 32.4% of pediatricians, 25.2% of advanced practice nurses, 19.8% of internal medicine physicians, and 19.4% of physician assistants, nearly three-fourths of obstetrician-gynecologists (73.1%) and nurse-midwives (72.6%), prescribed the pill, patch, or ring ( Table 1 ). Although obstetrician-gynecologists and nurse-midwives were more likely to prescribe the pill, patch, or ring, the greatest numbers of contraception prescribers were family medicine physicians (72,725) and advanced practice nurses (70,115).
| Provider type | Pill, patch, or ring | Pill | Patch | Ring | Total providers | ||||
|---|---|---|---|---|---|---|---|---|---|
| n | % | n | % | n | % | n | % | ||
| Obstetrician-gynecologist | 36,887 | 73.1 | 36,839 | 73.0 | 15,166 | 30.0 | 24,721 | 49.0 | 50,489 |
| Family medicine physician | 72,725 | 51.4 | 72,529 | 51.3 | 6641 | 4.7 | 13,176 | 9.3 | 141,455 |
| Internal medicine physician | 28,527 | 19.8 | 28,314 | 19.6 | 1110 | 0.8 | 3514 | 2.4 | 144,155 |
| Pediatrician | 23,080 | 32.4 | 23,027 | 32.4 | 1242 | 1.7 | 474 | 0.7 | 71,167 |
| Other physicians | 24,478 | 3.9 | 24,152 | 3.8 | 747 | 0.1 | 1404 | 0.2 | 633,415 |
| Total physicians | 185,697 | 17.8 | 184,861 | 17.8 | 24,906 | 2.4 | 43,289 | 4.2 | 1,040,681 |
| Advanced practice nurse | 70,115 | 25.2 | 69,874 | 25.1 | 10,635 | 3.8 | 15,901 | 5.7 | 278,695 |
| Nurse-midwife | 6552 | 72.6 | 6545 | 72.5 | 2230 | 24.7 | 3627 | 40.2 | 9029 |
| Physician assistant | 25,464 | 19.4 | 25,388 | 19.4 | 2860 | 2.2 | 5190 | 4.0 | 130,926 |
| Other health professions | 2098 | 0.6 | 2061 | 0.6 | 37 | 0.0 | 165 | 0.0 | 337,755 |
| Total advanced practice clinicians | 104,229 | 13.8 | 103,868 | 13.7 | 15,762 | 2.1 | 24,883 | 3.3 | 756,405 |
| Total | 289,926 | 16.1 | 288,729 | 16.1 | 40,668 | 2.3 | 68,172 | 3.8 | 1,797,086 |
In addition, contraception prescribing patterns differed by specialty. Although 73.0% of obstetrician-gynecologists prescribed the pill, 30.0% and 49.0% prescribed the patch or ring, respectively. This general pattern was observed across other specialties and professions, but at lower rates. Although 51.3% of family medicine physicians prescribed the pill, 4.7% prescribed the patch and 9.3% prescribed the ring. Pediatricians were the exception in that 32.4% prescribed the pill, 0.7% prescribed the ring, and 1.7% prescribed the patch ( Table 1 ).
Mapping the contraception workforce revealed a substantial difference within and across states. The ratio of all primary care providers prescribing contraception to the female population of reproductive age (ages, 15–44 years) varied from 27.9 providers per 10,000 population in New Jersey to 74.2 providers per 10,000 population in Maine ( Figure 1 ). We included the following as primary care providers: obstetrician-gynecologists, family medicine physicians, internal medicine physicians, pediatricians, advanced practice nurses, nurse-midwives, and physician assistants. The Supplemental Table in the Appendix shows the provider density for all states. Across counties, the bottom and top quartiles of counties ranged from <27.9 to >51.34 providers per 10,000 women of reproductive age. A further difference in county-level contraception providers was seen across provider specialty and profession: 1411 of 3006 US counties had no contraception that was provided by obstetrician-gynecologists or nurse-midwives. However, 674 of 1411 counties had family medicine physicians, advanced practice nurses, or physician assistants as contraception providers (county-level counts of contraception providers are available to interested researchers on request).
The proportion of prescription contraception providers (n=263,350) who accepted Medicaid varied by state from 55.5% in Texas to 89.1% in Vermont ( Figure 2 ). In addition, the proportion of providers that accepted Medicaid varied by specialty within states ( Figure 3 ). For example, the proportions of family medicine contraception providers that accepted Medicaid varied from 49.7% in Florida to 91.1% in Massachusetts. Obstetrician-gynecologists providing prescription contraception who accepted Medicaid ranged from 83.9% in the District of Columbia to 100% in North Dakota; internal medicine physicians ranged from 25.0% in Texas to 75.9% in Delaware; pediatricians ranged from 63.0% in Wyoming to 97.4% in Idaho; and advanced practice nurses and nurse-midwives ranged from 55.9% in Texas to 91.8% in Montana. The Supplemental Table in the Appendix shows the Medicaid acceptance rates for all states.
In-person contraception
In a subset of 61,124 physicians who billed a contraception procedure code or contraception management diagnosis code, there were large differences in the proportion of providers offering the 3 different contraceptive method types by provider specialty. Approximately half of obstetrician-gynecologists and family medicine physicians (50.2% and 52.2%, respectively) who billed for contraception management provided the contraceptive shot compared with 34.7% of internal medicine physicians and 34.1% of pediatricians ( Table 2 ). There was a wider distribution in the proportion of IUD and implant provisions by specialty. Although 92.8% of obstetrician-gynecologists provided IUDs, the next highest proportion of physicians that provided IUDs was 16.4% among family medicine physicians, followed by 2.6% of internal medicine physicians and 0.6% of pediatricians. Similarly, 56.2% of obstetrician-gynecologists provided the implant, compared with 13.7% of family medicine physicians, 1.8% of internal medicine physicians, and 4.0% of pediatricians. Pediatricians are more likely to provide the implant than the IUD, whereas the reverse is true for other specialties.
