ORIGINAL ARTICLE
The Contribution of Physicians, Physician Assistants, and Nurse Practitioners Toward Rural Primary Care Findings From a 13-State Survey Mark P. Doescher, MD, MSPH,* C. Holly A. Andrilla, MS,w Susan M. Skillman, MS,w Perri Morgan, PhD, PA-C,z and Louise Kaplan, PhDy
Background: Estimates of the relative contributions of physicians, physician assistants (PAs), and nurse practitioners (NPs) toward rural primary care are needed to inform workforce planning activities aimed at reducing rural primary shortages. Objectives: For each provider group, this study quantifies the average weekly number of outpatient primary care visits and the types of services provided within and beyond the outpatient setting. Methods: A randomly drawn sample of 788 physicians, 601 PAs, and 918 NPs with rural addresses in 13 US states responded to a mailed questionnaire that measured reported weekly outpatient visits and scope of services provided within and beyond the outpatient setting. Analysis of variance and w2 testing were used to test for bivariate associations. Multivariate regression was used to model average weekly outpatient volume adjusting for provider sociodemographics and geographical location. Results: Compared with physicians, average weekly outpatient visit quantity was 8% lower for PAs and 25% lower for NPs (P < 0.001). After multivariate adjustment, this gap became negligible for PAs (P = 0.56) and decreased to 10% for NPs (P < 0.001). Compared with PAs and NPs, primary care physicians were more likely to provide services beyond the outpatient setting, including hospital care, emergency care, childbirth attending deliveries, and afterhours call coverage (all P < 0.001). Conclusions: Although our findings suggest that a greater reliance on PAs and NPs in rural primary settings would have a minor impact on outpatient practice volume, this shift might reduce the From the *Peggy and Charles Stephenson Cancer Center, University of Oklahoma Health Sciences Center, Oklahoma City, OK; wWWAMI Rural Health Research Center, University of Washington, Seattle, WA; zDepartment of Community and Family Medicine, Duke University Medical Center, Durham, NC; and yDepartment of Nursing, Saint Martin’s University, Lacey, WA. Funded by the US Health Resources and Services Administration (HRSA), Federal Office of Rural Health Policy, through the Rural Health Research Center program, Grant #U1CRH03712. This University of Washington Rural Health Research Center study was supported by the Federal Office of Rural Health Policy, Health Resources and Services Administration, and Department of Health and Human Services. The authors declare no conflict of interest. Reprints: Mark P. Doescher, MD, MSPH, Peggy Charles Stephenson Cancer Center, University of Oklahoma, 800 Northeast 10th Street, Suite 5011, Oklahoma City, OK 73104. E-mail: [email protected]. Copyright r 2014 by Lippincott Williams & Wilkins ISSN: 0025-7079/14/5206-0549
Medical Care
Volume 52, Number 6, June 2014
availability of services that have more often been traditionally provided by rural primary care physicians beyond the outpatient clinic setting. Key Words: health care workforce, rural populations, primary care, outpatient visit quantity (Med Care 2014;52: 549–556)
H
istorically, rural locations in the United States have relied on physicians for primary care services, and in 2005 roughly half of all rural physicians were family physicians, general internists, or general pediatricians.1 Despite recent increases in the number of US medical school graduates, fewer medical students are from rural locations or intend to practice in rural locations.2 Thus, shortages of primary care physicians in rural locations are apt to intensify3,4 as demand for care increases because of an aging rural population and uptake of health insurance coverage through the Affordable Care Act.5 Because physician assistants (PAs) and nurse practitioners (NPs) contribute significantly to the rural primary care workforce,6 recent increases in their numbers may help meet this growing demand. The number of PAs in the US health care workforce increased over 100% between 2000 and 2010,7 and the number of NPs increased by over 75% between 2000 and 2011.8,9 In contrast, the US physician workforce increased 50% Don’t know Percent of patients with Medicaid coverage (%) None < 10% 10%–25% 26%–50% > 50% Don’t know
Physicians
PAs
NPs
788 53.6 (52.8–54.3) 67.5 20.9 (20.1–21.8) 18.4 (17.6–19.3)
601 46.3 (45.3–47.3) 39.3 14.2 (13.3–15.1) 11.2 (10.5–12.0)
918 50.5 (49.8–51.2) 6.2 11.8 (11.2–12.4) 11.0 (10.4–11.6)
70.6 29.4
75.0 25.3
71.7 28.3
55.2 8.3 16.8 19.7
31.6 10.9 34.0 23.5
33.3 12.4 21.1 33.1
47.6 29.4 21.1
37.6 33.4 24.7
43.0 25.7 23.5
11.5 5.1 22.0 35.9 20.9 4.6
4.1 4.9 23.1 40.8 21.9 5.1
10.6 10.5 23.6 33.0 16.6 5.6
4.9 14.2 34.1 22.5 19.4 4.9
3.8 8.6 25.4 31.4 25.0 5.7
6.0 8.8 22.0 32.1 25.4 5.7
P
Missing Cases
< 0.001 < 0.001 < 0.001 < 0.001 0.231
44 24 79 153 57
< 0.001 < 0.001 0.025 < 0.001 < 0.001
57
< 0.001 0.006 0.258 < 0.001
88
< 0.001
82
*Percentages may not total 100 because primary practice location for some providers with multiple practice locations was in urban area types. CI indicates confidence interval; NP, nurse practitioners; PA, physician assistants; RHC, rural health clinic.
and higher percentages of PAs practiced in small rural areas (P = 0.006). There were no significant differences among the types of providers practicing in isolated small rural areas (Table 1). The majority of physicians, PAs, and NPs provided services for patients with Medicare and Medicaid coverage. Among providers who answered the survey questions on the proportion of patients covered by Medicare or Medicaid, slightly over one fifth of physicians, physicians and PAs responded that more than half of their ambulatory care practice patients were covered by Medicare, but this proportion was slightly lower for NPs (Table 1) (P < 0.001). About one fifth of the physicians and over one quarter of PAs and NPs reported that more than half of their patients had Medicaid coverage (P = 0.004).
Average Weekly Outpatient Visits and Hours Worked In unadjusted analyses, the physician average of 84.6 outpatient visits per week was 6.5 visits higher than the 78.1 average for PAs and about 21 visits higher than the 63.5 figure for NPs (Table 2) (P < 0.001). Physicians and PAs also spent more hours per week in direct outpatient patient care than NPs (39.1 and 39.0 for physicians and PAs, respectively, compared with 34.9 for NPs) (P < 0.001), but NPs reported the fewest weeks per year spent not seeing patients r
2014 Lippincott Williams & Wilkins
(3.5 weeks for NPs compared with 4.4 and 4.0 for physicians and PAs, respectively) (P < 0.001). Physicians provided significantly more well-child visits per week than PAs and NPs (12.6 compared with 8.0 and 7.4, respectively) (P < 0.001), but the provider groups did not differ significantly in the provision of weekly prenatal visits (fewer than an average of 2.5 per week for all). PAs performed the most minor office procedures per week (6.7 on average for PAs compared with 5.4 and 5.1 for physicians and NPs, respectively) (P = 0.007). Multivariate regression analyses examined whether provider characteristics (provider type, sex, age category, geography, and type of practice facility) contributed to the observed differences in 3 dependent measures: weekly outpatient visit quantity, hours per week spent in direct patient care, and weeks per year spent not seeing patients (Table 3). In the adjusted analyses, PAs averaged about the same number of outpatient visits as physicians, whereas NPs averaged about 7 fewer visits per week than physicians (P < 0.001). Outpatient visit volume also was significantly lower for providers who were: female (19.0 fewer visits, P < 0.001); 55 years of age or older (12 fewer visits than those who were 45 to 54 y of age, P < 0.001); located in Census regions other than the South (24 fewer visits in the Northeast, 12 fewer in the Midwest, and 19 fewer in the West, all P < 0.001); or working in practice facilities other www.lww-medicalcare.com |
551
Medical Care
Doescher et al
Volume 52, Number 6, June 2014
TABLE 2. Average Weekly Number of Outpatient Visits and Types of Visits Among Rural Primary Care Providers Physicians n Mean Mean Mean Mean Mean Mean
weekly office visits (95% CI) weekly hours of direct patient care (95% CI) weeks not seeing patients (95% CI) weekly well-child visits (95% CI) weekly prenatal care visits (95% CI) weekly minor procedures (95% CI)
84.6 39.1 4.4 12.6 2.3 5.4
PAs
788 (81.4–87.7) (38.1–40.1) (4.1–4.8) (10.4–14.9) (1.3–3.3) (4.7–6.0)
78.1 39.0 4.0 8.0 1.1 6.7
601 (74.3–82.0) (38.0–39.9) (3.6–4.4) (6.8–9.2) (0.5–1.6) (5.8–7.6)
NPs 63.5 34.9 3.5 7.4 2.1 5.1
918 (60.8–66.2) (34.1–35.6) (3.4–3.7) (6.6–8.2) (1.5–2.7) (4.4–5.7)
P
Missing Cases
< 0.001 < 0.001 < 0.001 < 0.001 0.064 0.007
87 35 78 183 268 169
CI indicates confidence interval; NP, nurse practitioners; PA, physician assistants.
than private practice (6 fewer visits among those in community health centers, P = 0.042; 7 fewer visits in rural health clinics, P = 0.002; and 16 fewer visits for those in other nonprivate practice settings, P < 0.001). No significant difference between physicians, PAs, and NPs was observed for hours worked per week. However, the number of hours worked per week was significantly lower for providers who were: female (4.6 fewer hours, P < 0.001); 55 y of age or older (2.3 fewer hours per week than those who were 45 to
54 y of age, P < 0.001); and located in Census regions other than the South (5.1 fewer hours in the Northeast, P < 0.001; 1.9 fewer in the Midwest, P = 0.003; and 2.5 fewer in the West, P = 0.001). Physicians averaged 0.5 weeks less per year seeing patients compared with PAs (P = 0.014) and 0.8 weeks less compared with NPs (P < 0.001). Those located in Census regions other than the South spent significantly less weeks per year seeing patients (1.4 fewer weeks in the Northeast, 1.0 fewer in the Midwest and 1.1 fewer in the West, all
TABLE 3. Adjusted Average Numbers of Outpatient Visits per Week, Average Hours Spent in Direct Patient Care per Week, and Average Number of Weeks per Year Spent Not Seeing Patients by Rural Primary Care Provider Type and Other Characteristics
Characteristics
Average Number Outpatient Visits Per Week (b-Coefficient and CI) (N = 2040)
Constant Provider type PAs NPs Physicians (ref.) Age (y) < 45 45 to
The Contribution of Physicians, Physician Assistants, and Nurse Practitioners Toward Rural Primary Care Findings From a 13-State Survey Mark P. Doescher, MD, MSPH,* C. Holly A. Andrilla, MS,w Susan M. Skillman, MS,w Perri Morgan, PhD, PA-C,z and Louise Kaplan, PhDy
Background: Estimates of the relative contributions of physicians, physician assistants (PAs), and nurse practitioners (NPs) toward rural primary care are needed to inform workforce planning activities aimed at reducing rural primary shortages. Objectives: For each provider group, this study quantifies the average weekly number of outpatient primary care visits and the types of services provided within and beyond the outpatient setting. Methods: A randomly drawn sample of 788 physicians, 601 PAs, and 918 NPs with rural addresses in 13 US states responded to a mailed questionnaire that measured reported weekly outpatient visits and scope of services provided within and beyond the outpatient setting. Analysis of variance and w2 testing were used to test for bivariate associations. Multivariate regression was used to model average weekly outpatient volume adjusting for provider sociodemographics and geographical location. Results: Compared with physicians, average weekly outpatient visit quantity was 8% lower for PAs and 25% lower for NPs (P < 0.001). After multivariate adjustment, this gap became negligible for PAs (P = 0.56) and decreased to 10% for NPs (P < 0.001). Compared with PAs and NPs, primary care physicians were more likely to provide services beyond the outpatient setting, including hospital care, emergency care, childbirth attending deliveries, and afterhours call coverage (all P < 0.001). Conclusions: Although our findings suggest that a greater reliance on PAs and NPs in rural primary settings would have a minor impact on outpatient practice volume, this shift might reduce the From the *Peggy and Charles Stephenson Cancer Center, University of Oklahoma Health Sciences Center, Oklahoma City, OK; wWWAMI Rural Health Research Center, University of Washington, Seattle, WA; zDepartment of Community and Family Medicine, Duke University Medical Center, Durham, NC; and yDepartment of Nursing, Saint Martin’s University, Lacey, WA. Funded by the US Health Resources and Services Administration (HRSA), Federal Office of Rural Health Policy, through the Rural Health Research Center program, Grant #U1CRH03712. This University of Washington Rural Health Research Center study was supported by the Federal Office of Rural Health Policy, Health Resources and Services Administration, and Department of Health and Human Services. The authors declare no conflict of interest. Reprints: Mark P. Doescher, MD, MSPH, Peggy Charles Stephenson Cancer Center, University of Oklahoma, 800 Northeast 10th Street, Suite 5011, Oklahoma City, OK 73104. E-mail: [email protected]. Copyright r 2014 by Lippincott Williams & Wilkins ISSN: 0025-7079/14/5206-0549
Medical Care
Volume 52, Number 6, June 2014
availability of services that have more often been traditionally provided by rural primary care physicians beyond the outpatient clinic setting. Key Words: health care workforce, rural populations, primary care, outpatient visit quantity (Med Care 2014;52: 549–556)
H
istorically, rural locations in the United States have relied on physicians for primary care services, and in 2005 roughly half of all rural physicians were family physicians, general internists, or general pediatricians.1 Despite recent increases in the number of US medical school graduates, fewer medical students are from rural locations or intend to practice in rural locations.2 Thus, shortages of primary care physicians in rural locations are apt to intensify3,4 as demand for care increases because of an aging rural population and uptake of health insurance coverage through the Affordable Care Act.5 Because physician assistants (PAs) and nurse practitioners (NPs) contribute significantly to the rural primary care workforce,6 recent increases in their numbers may help meet this growing demand. The number of PAs in the US health care workforce increased over 100% between 2000 and 2010,7 and the number of NPs increased by over 75% between 2000 and 2011.8,9 In contrast, the US physician workforce increased 50% Don’t know Percent of patients with Medicaid coverage (%) None < 10% 10%–25% 26%–50% > 50% Don’t know
Physicians
PAs
NPs
788 53.6 (52.8–54.3) 67.5 20.9 (20.1–21.8) 18.4 (17.6–19.3)
601 46.3 (45.3–47.3) 39.3 14.2 (13.3–15.1) 11.2 (10.5–12.0)
918 50.5 (49.8–51.2) 6.2 11.8 (11.2–12.4) 11.0 (10.4–11.6)
70.6 29.4
75.0 25.3
71.7 28.3
55.2 8.3 16.8 19.7
31.6 10.9 34.0 23.5
33.3 12.4 21.1 33.1
47.6 29.4 21.1
37.6 33.4 24.7
43.0 25.7 23.5
11.5 5.1 22.0 35.9 20.9 4.6
4.1 4.9 23.1 40.8 21.9 5.1
10.6 10.5 23.6 33.0 16.6 5.6
4.9 14.2 34.1 22.5 19.4 4.9
3.8 8.6 25.4 31.4 25.0 5.7
6.0 8.8 22.0 32.1 25.4 5.7
P
Missing Cases
< 0.001 < 0.001 < 0.001 < 0.001 0.231
44 24 79 153 57
< 0.001 < 0.001 0.025 < 0.001 < 0.001
57
< 0.001 0.006 0.258 < 0.001
88
< 0.001
82
*Percentages may not total 100 because primary practice location for some providers with multiple practice locations was in urban area types. CI indicates confidence interval; NP, nurse practitioners; PA, physician assistants; RHC, rural health clinic.
and higher percentages of PAs practiced in small rural areas (P = 0.006). There were no significant differences among the types of providers practicing in isolated small rural areas (Table 1). The majority of physicians, PAs, and NPs provided services for patients with Medicare and Medicaid coverage. Among providers who answered the survey questions on the proportion of patients covered by Medicare or Medicaid, slightly over one fifth of physicians, physicians and PAs responded that more than half of their ambulatory care practice patients were covered by Medicare, but this proportion was slightly lower for NPs (Table 1) (P < 0.001). About one fifth of the physicians and over one quarter of PAs and NPs reported that more than half of their patients had Medicaid coverage (P = 0.004).
Average Weekly Outpatient Visits and Hours Worked In unadjusted analyses, the physician average of 84.6 outpatient visits per week was 6.5 visits higher than the 78.1 average for PAs and about 21 visits higher than the 63.5 figure for NPs (Table 2) (P < 0.001). Physicians and PAs also spent more hours per week in direct outpatient patient care than NPs (39.1 and 39.0 for physicians and PAs, respectively, compared with 34.9 for NPs) (P < 0.001), but NPs reported the fewest weeks per year spent not seeing patients r
2014 Lippincott Williams & Wilkins
(3.5 weeks for NPs compared with 4.4 and 4.0 for physicians and PAs, respectively) (P < 0.001). Physicians provided significantly more well-child visits per week than PAs and NPs (12.6 compared with 8.0 and 7.4, respectively) (P < 0.001), but the provider groups did not differ significantly in the provision of weekly prenatal visits (fewer than an average of 2.5 per week for all). PAs performed the most minor office procedures per week (6.7 on average for PAs compared with 5.4 and 5.1 for physicians and NPs, respectively) (P = 0.007). Multivariate regression analyses examined whether provider characteristics (provider type, sex, age category, geography, and type of practice facility) contributed to the observed differences in 3 dependent measures: weekly outpatient visit quantity, hours per week spent in direct patient care, and weeks per year spent not seeing patients (Table 3). In the adjusted analyses, PAs averaged about the same number of outpatient visits as physicians, whereas NPs averaged about 7 fewer visits per week than physicians (P < 0.001). Outpatient visit volume also was significantly lower for providers who were: female (19.0 fewer visits, P < 0.001); 55 years of age or older (12 fewer visits than those who were 45 to 54 y of age, P < 0.001); located in Census regions other than the South (24 fewer visits in the Northeast, 12 fewer in the Midwest, and 19 fewer in the West, all P < 0.001); or working in practice facilities other www.lww-medicalcare.com |
551
Medical Care
Doescher et al
Volume 52, Number 6, June 2014
TABLE 2. Average Weekly Number of Outpatient Visits and Types of Visits Among Rural Primary Care Providers Physicians n Mean Mean Mean Mean Mean Mean
weekly office visits (95% CI) weekly hours of direct patient care (95% CI) weeks not seeing patients (95% CI) weekly well-child visits (95% CI) weekly prenatal care visits (95% CI) weekly minor procedures (95% CI)
84.6 39.1 4.4 12.6 2.3 5.4
PAs
788 (81.4–87.7) (38.1–40.1) (4.1–4.8) (10.4–14.9) (1.3–3.3) (4.7–6.0)
78.1 39.0 4.0 8.0 1.1 6.7
601 (74.3–82.0) (38.0–39.9) (3.6–4.4) (6.8–9.2) (0.5–1.6) (5.8–7.6)
NPs 63.5 34.9 3.5 7.4 2.1 5.1
918 (60.8–66.2) (34.1–35.6) (3.4–3.7) (6.6–8.2) (1.5–2.7) (4.4–5.7)
P
Missing Cases
< 0.001 < 0.001 < 0.001 < 0.001 0.064 0.007
87 35 78 183 268 169
CI indicates confidence interval; NP, nurse practitioners; PA, physician assistants.
than private practice (6 fewer visits among those in community health centers, P = 0.042; 7 fewer visits in rural health clinics, P = 0.002; and 16 fewer visits for those in other nonprivate practice settings, P < 0.001). No significant difference between physicians, PAs, and NPs was observed for hours worked per week. However, the number of hours worked per week was significantly lower for providers who were: female (4.6 fewer hours, P < 0.001); 55 y of age or older (2.3 fewer hours per week than those who were 45 to
54 y of age, P < 0.001); and located in Census regions other than the South (5.1 fewer hours in the Northeast, P < 0.001; 1.9 fewer in the Midwest, P = 0.003; and 2.5 fewer in the West, P = 0.001). Physicians averaged 0.5 weeks less per year seeing patients compared with PAs (P = 0.014) and 0.8 weeks less compared with NPs (P < 0.001). Those located in Census regions other than the South spent significantly less weeks per year seeing patients (1.4 fewer weeks in the Northeast, 1.0 fewer in the Midwest and 1.1 fewer in the West, all
TABLE 3. Adjusted Average Numbers of Outpatient Visits per Week, Average Hours Spent in Direct Patient Care per Week, and Average Number of Weeks per Year Spent Not Seeing Patients by Rural Primary Care Provider Type and Other Characteristics
Characteristics
Average Number Outpatient Visits Per Week (b-Coefficient and CI) (N = 2040)
Constant Provider type PAs NPs Physicians (ref.) Age (y) < 45 45 to
