Impact of Physician Assistant Care
on Office Visit Resource Use in the
Perri A. Morgan, Nilay D. Shah, Jay S. Kaufman, and
Mark A. Albanese
a substantive portion of a patient’s office-based visits affects office visit resource use.
Data Source. Medical Expenditure Panel Survey (MEPS) Household Component
data from 1996 to 2004.
Study Design. This retrospective cohort study compares the number of office-based
visitsper yearbetweenadults forwhom PAs provided ? 30percent ofvisitsand adults
cared for by physicians only.
Data Collection/Extraction Methods. The Agency for Healthcare Research and
Quality collects MEPS data using methods designed to produce data representative of
to compare the number of visits per year between persons with and without PA care,
adjusted for demographic, geographic, and socioeconomic factors; insurance status;
health status; and medical conditions.
Principal Findings. After case-mix adjustment, patients for whom PAs provided a
substantive portion of care used about 16 percent fewer office-based visits per year than
patients caredfor byphysiciansonly. This difference intheuse ofoffice-based visitswas
not offset by increased office visit resource use in other settings.
Conclusions. Results indicate that the inclusion of PAs in the U.S. provider mix does
not affect overall office visit resource use.
Key Words. Health workforce, physician assistants, access/demand/utilization of
BACKGROUND AND RATIONALE
The physician assistant (PA) profession has grown dramatically in recent
years,withthenumberofpracticingPAstriplingfromabout 20,000in1991 to
over68,000 in 2008 (American Academy of Physician Assistants2007).There
rHealth Research and Educational Trust
U.S. workforce for every six physicians (National Center for Health Statistics
2005). PAs are providing a growing portion of office visits for medical care in
the United States (Druss et al. 2003), and could buffer predicted physician
shortages. How this change in the provider mix might impact overall use of
medical services, however, is unknown. Does PA participation increase the
effective supply of what have traditionally been physician services, or does it
lead to provision of expanded or redundant services, thereby increasing per
capita office visits per year?
Existing researchsuggeststhat,inthe UnitedStates,participation incare
by PAs and nurse practitioners (NPs) does not increase overall use of medical
services, but this research is from limited settings and patient populations.
Hooker evaluated the effect of provider type (PA or physician) on an episode
of care for four acute uncomplicated problems in a managed area setting and
found that PAs did not require more expenditures or more return visits to
manage the episode of care (Hooker 2002). In a rare randomized trial com-
paring provider types, Mundinger et al. (2000) found similar outcomes and
health resource use among a predominantly female and Hispanic population
The extent to which findings from these studies generalize to other settings or
to care for chronic or serious conditions is unknown. Our literature review
found no national study investigating the effect of PA or NP use on longi-
tudinal health resource use.
Several studies have focused on the production efficiency of PAs and
NPs as measured by the number of office visits provided per unit time or per
unit of labor cost, compared with physicians. These studies have generally
found that PAs provide between 76 and 100 percent as many office visits as
physicians per unit time (Record et al. 1980; Hooker 1993; Larson, Hart, and
and increase efficiency (Medical Group Management Association 2006).
Address correspondence to Perri A. Morgan, Ph.D., P.A.-C., Director of Physician Assistant
Research, Physician Assistant Division, Department of Community and Family Medicine, Duke
University Medical Center, 3848 DUMC, Durham, NC 27710; e-mail: firstname.lastname@example.org.
Nilay D. Shah, Ph.D., Assistant Professor of Health Services Research, is with the Division of
Health Care Policy and Research, Mayo Clinic, Rochester, MN. Jay S. Kaufman, Ph.D., Associate
Mark A. Albanese, Ph.D., is with the Department of Population Health Sciences, Madison, WI.
Impact of Physician Assistant Care 1907
This apparent high productivity might be misleading with respect to
cost-effectiveness if the use of PAs leads to increased per person resource use.
For example, it is possible that PAs may be employed to provide services that
would otherwise not have been provided (complementary services) or that
PAs may schedule more return visits than do physicians, thereby increasing
total office visitsperperson.Although complementaryservicesand additional
follow-up visits could result in higher quality of care, they might not increase
the overall productivity of the workforce, a pressing concern given predicted
physician shortages (Association of American Medical Colleges 2006).
As health services researchers consider the impact of PAs and NPs on
health care provision, it is relevant to ask whether these clinicians replace care
that would otherwise be provided by a physician (substitution of services) or
whether they provide care that would otherwise not have been provided
The substitution model is supported by Hooker’s study of roles of PAs
and NPs in managed care (Hooker 1993), by Gryzbicki’s detailed analysis of
task substitution in a single family practice/general medicine practice in
Pennsylvania (Grzybicki et al. 2002), and by Mundinger’s randomized trial of
NP and physician care (Mundinger et al. 2000) In addition, research on
staffing ratios in health maintenance organizations in the mid-1990s found an
inverse relationship between the numbers of PAs and advanced practice
nurses (APNs) employed and numbers of physicians employed per 100,000
enrolled (Dial et al. 1995). As the use of PAs and APNs increased, the use of
physicians decreased, suggesting that the PAs and APNs were providing ser-
vices that physicians would otherwise provide. Several studies reporting on
the use of physician assistants to replace house staff in academic medical
centers are based on the assumption that PAs replace physician services
(Carzoli et al. 1994; Stoddard, Kindig, and Libby 1994; Schulman, Lucchese,
and Sullivan 1995; Miller et al. 1998).
et al. (2004) found that random assignment of NPs to primary care practices in
not substitutes for doctors but provide a wider range of services than
was available previously.’’ A randomized trial in Britain showed that
NPs scheduled return visits more frequently than physicians (Venning et al.
2000) and a systematic review relying heavily on European experience
indicated that NPs provide longer consultations and make more investigations
than do physicians (Horrocks, Anderson, and Salisbury 2002). Generalizing
these results to PA practice in the United States is problematic because
1908 HSR: Health Services Research 43:5, Part II (October 2008)
practice patterns in the United States may differ from those in Europe, and
because PA practice can vary in meaningful ways from that of NPs (Hooker
and McCaig 2001).
The dichotomization of PA/NP services as either a substitute or
complement for physician care is an oversimplification, and both patterns
are likely to exist in practice. Physicians and PAs/NPs develop diverse prac-
tice arrangements based on personal preferences and practice needs. Some
physicians may choose to hire PAs or NPs to provide preventive and coun-
seling services that the physicians are unable to find the time to provide,
leading to services that are intentionally complementary. Others may work
out substitution practice arrangements in which PAs or NPs see the patients
physicians may be random or may depend on scheduling constraints or
idiosyncratic interests of the providers involved. In many practices, there
will be a mix of substitute and complementary services. Practice patterns may
also evolve over time, as individual physicians, PAs, or NPs develop special
interests or skills.
is an oversimplification, at a macro level it will be useful to assess whether,
on average, addition of PAs or NPs to the national mix of providers has an
effect of substituting for or complementing physician services. Information
about whether increased resource use can be expected as a consequence
of increasing numbers of PAs and NPs, and about the magnitude of
any increased resource use, is necessary for the projection of workforce
This project addresses the research question: Is substantive inclusion of
PAs in patient care associated with increased numbers of office visits per
patient, adjusting for case-mix differences between patients seen by PAs
and physicians? If PAs are functioning as substitutes, we would expect no
increase in office visit resource use per patient when PAs are included in
care. Alternatively, if PAs are providing complementary care, the total
added to the provider mix. Our study expands upon existing research
because it uses a diverse national sample, covers a year of health care
experience for eachperson (rather than a single encounter or episode of care),
and employs a validated means of case-mix adjustment. Because the
data source does not identify NPs as a distinct category of office visit provid-
ers, this study does not include NPs.
Impact of Physician Assistant Care1909
This study uses regression analysis to compare the number of office visits in
oneyear between agroup ofadultsforwhom PAsprovided care foratleast30
percent of visits and a group who reported care only from physicians.
Data are from the Medical Expenditure Panel Survey (MEPS) Household
Component office-based visit files from 1996 to 2004 (Agency for Healthcare
Research and Quality 2006). MEPS is administered by the Agency for
Healthcare Research and Quality to a national probability sample of the
noninstitutionalized civilian population. The complex sampling design em-
ploys stratification, clustering, multiple stages of selection, and overrepresen-
tation of select subpopulations. Sampling weights that account for this
complex design, as well as for nonresponse, are provided for public use by
The MEPS samples households and collects information from a single
household respondent regarding each household member. Each household
each of five rounds over the 2 years. MEPS uses an overlapping panel design,
while the other half are in their second year. The sample for each year is
designed to be nationally representative for purposes of calculating national
estimates at the family and personal levels.
in 1 year. Although MEPS also collects data from events occurring in inpa-
tient, emergency, and hospital outpatient departments, the primary analysis
was limited to office-based encounters because PA data are not included in
some of the other settings.
with increased office visit resource use in other settings or with poor outcomes
requiring emergent medical attention or admission to a hospital, our second-
ary analysis compared total yearly hospital outpatient clinic visits, emergency
department visits, and hospital discharges between the two comparison
1910HSR: Health Services Research 43:5, Part II (October 2008)
A group of patients who reported that a substantive portion of their office-
based visits were attended solely by PAs is compared with a group who re-
ported only physician care. Our primary definition of a substantive portion of
care is when 30 percent or more of visits in 1 year were reported as provided
by a PA. The cutoff point of at least 30 percent of visits is intended to ensure
a minimum exposure to PA care. Because 30 percent is the minimum per-
inthiscategoryreceived morethan 30 percent oftheircare fromPAs. Wealso
evaluate aspectrum ofcut-pointsinordertotestthesensitivity ofourresultsto
The group of 1,762 adults who indicated that PAs were the provider for
at least 30 percent of their visits during the study year (the exposed group) is
compared with 111,184 control persons who reported that a physician
attended all their visits. For convenience, the exposed group will be referred
to as the ‘‘PA1’’ group and the control group will be referred to as the ‘‘phy-
sician-only’’ group. Persons who reported that providers other than PAs or
physicians attended their office visits were excluded. Persons who reported
between 0 and 29 percent of office-based care from PAs were also excluded,
in order to draw a greater distinction between the two groups. Since the
number of annual visits per person is bimodal by age, with children and older
patients having the most visits, averages taken from this bimodal distribution
could produce misleading results. Because of this, persons o21 years of age
For the group with PA1 care, frequency of observations is concentrated
in fractions possible for those with few visits. For example, persons with one
visit in the year can only have 0 or 100 percent of visits provided by a PA,
those with two visits in the year could have 0, 50, or 100 percent, etc. See
Supplementary Appendix S1 for additional graphic description of the com-
parison groups. Because the multimodal distribution of these proportions
makes this variable unsuitable for parametric analysis of the continuous vari-
able, we use the dichotomous independent variable described above.
Case-mix adjustment was used to balance the comparison groups on factors
that could affect the study outcome. We adjusted for variables in each of
the three categories that Anderson proposed to explain health resource use:
Impact of Physician Assistant Care1911
predisposing (age, gender, race, and rural versus urban status), enabling
(health insurance and socioeconomic status as indicated by the percent of
federal poverty level), and need (self-rated health status and medical condi-
tions) factors (Andersen 1995).
Medical conditions for each subject were obtained from the MEPS
medical condition files, which aggregate medical conditions that respondents
report as the cause of any medical encounters in any setting (including hos-
pitals and emergency departments) during the study period. Conditions for
which patients did not seek care will not be included. MEPS professional
coders assign ICD-9 codes to verbal descriptions of conditions reported
by household respondents as reasons for medical encounters in any setting
during the study.
The Ambulatory Diagnostic Groups (ADG) from the Johns Hopkins
Adjusted Clinical Group
a case-mix methodology designed for use with ambulatory and inpatient ad-
system and are used to assign each ICD-9-CM diagnosis into one of 34 unique
diagnostic morbidity clusters based on a variety of factors, including clinical
similarity, likelihood of persistence or recurrence, and expected need for
continued treatment. Because MEPS reports three-digit ICD-9 codes, we used
the modification of the ACG system that is compatible with the three-digit
s(ACG) System are a risk adjustment tool for pre-
Characteristics of persons in the comparison groups are initially evaluated for
group differencesusing Student’st-testfordifferencesinmeansforcontinuous
variables and w2-analysis for categorical variables. To assess the number of
office visits and hospital outpatient visits, a negative binomial model with a
log-link function was used (Byers et al. 2003). Use of the negative binomial
distribution relaxes the strict mean–variance relationship of the Poisson dis-
tribution, allowing the variance to exceed the mean. Coefficient estimates
from the negative binomial regression can be easily transformed to give the
incidence rate ratio. This yields a result that can be interpreted as the per-
centage increase (or decrease) in total number of visits over the 1-year period
for the exposed (PA1) group compared with the control (physician-only)
1912HSR: Health Services Research 43:5, Part II (October 2008)
Comparison of total emergency department visits and hospital dis-
charges between the two groups was performed by using the zero-inflated
Poisson model (ZIP) (Lambert 1992; Greene 1994). When modeling health
care resource use variables that contain a significant portion of zeros, ZIP has
the advantage (relative to the traditional two-part model) (Duan et al. 1983)
to use the service, yet did not use any during the recall period of the study.
Because emergency department visits and hospitalizations are more rare
compared with outpatient visits, accounting for zero-probability of service use
is very important. The ZIP models also produce incidence rate ratios, with
interpretation analogous to our other analyses.
All analyses were adjusted for the complex sample design and for non-
response using MEPS weight, strata, and cluster variables. These variables are
designed to produce national estimates representative of the U.S. noninstitu-
tionalized civilian population. All p-values are two-sided.
Table1 compares the PA1 and physician-only groupsregardingthevariables
used for risk adjustment. Persons in the PA1 group were younger and more
likely to be female, white, rural, of higher socioeconomic status, privately
insured, and with better self-rated health. These group differences are con-
sistent with other research comparing patients seen by PAs with those seen by
physicians (Hooker and McCaig 2001; Morgan 2007) and underscore the
need for risk adjustment when comparing use and outcomes between these
For the ADGs, the PA1 and physician-only groups differed by 44
to see persons with ‘‘Time limited-minor-primary infections’’ (48 percent of
persons in the PA1 group versus 37 percent of persons in the physician-only
and 17 percent for physician-only), ‘‘Time limited-minor’’ (25 percent for
PA1 and 19 percent for physician-only), and ‘‘Allergy’’ (18 percent for PA1
and 14 percent for physician-only) problems. The physician-only care group
was more likely to see persons in the ‘‘Chronic medical: stable’’ (40 percent
PA1, 44 percent physician-only) and ‘‘Chronic medical: unstable’’ (15 per-
cent PA1, 20 percent physician-only) groups. These differences are generally
Impact of Physician Assistant Care1913
Table1: Characteristics of PA1 and Physician-Only Groups
Age (mean, years)
Metropolitan statistical areaw
Near poor (%)
Low income (%)
Mid income (%)
High income (%)
Private insurance (%)
Public insurance (%)
Self-rated physical health
Very good (%)
Very poor (%)
No. of visits per year (mean)
Ambulatory Diagnostic Group (ADG) categories§
1. Time limited-minor (%)
2. Time limited-minor-primary infections (%)
3. Time limited-major (%)
4. Time limited-major-primary infections (%)
5. Allergies (%)
6. Asthma (%)
7. Likely to recur-discrete (%)
8. Likely to recur-discrete-primary infections (%)
9. Likely to recur-progressive (%)
10. Chronic medical: stable (%)
11. Chronic medical: unstable (%)
12. Chronic specialty: stable-orthopedic (%)
13. Chronic specialty: stable-ear, nose, and throat (%)
14. Chronic specialty: stable-ophthalmology (%)
1914HSR: Health Services Research 43:5, Part II (October 2008)
in the direction of physician-only care predominating among persons with
more complex problems, and again demonstrate the necessity of risk adjust-
Table 2 shows the incidence rate ratios indicating the proportionate
contrast in expected number of office visits over the 1-year period for the
exposed (PA1) group compared with the control (physician-only) group. For
risk ratio of .84 indicatesthat the numberofvisits per yearis reduced by about
16 percent for persons in the PA1 group compared with persons in the phy-
sician-only group, adjusted for demographic, geographic, socioeconomic,
insurance status, and health factors. When the cut-point was lowered to 10
15. Chronic specialty: unstable-orthopedics (%)
17. Chronic specialty: unstable-ear, nose, and throat (%)
18. Chronic specialty: unstable-ophthalmology (%)
20. Dermatologic (%)
21. Injuries/adverse events: minor (%)
22. Injuries/adverse events: major (%)
23. Psychosocial, time-limited: minor (%)
24. Psychosocial, recurrent or chronic, stable (%)
25. Psychosocial, recurrent or persistent (%)
26. Signs/symptoms: minor (%)
27. Signs/symptoms: uncertain (%)
28. Signs/symptoms: major (%)
29. Discretionary (%)
30. See and reassure (%)
31. Prevention/administrative (%)
32. Malignancy (%)
33. Pregnancy (%)
34. Dental (%)
nThe p-values refer to t-value for test for difference in means for continuous variables and to w2for
difference in proportions for categorical values.
wMetropolitan statistical area using US Census definition.
zPoverty categories are based on Current Population Survey poverty line. Poor denotes below
poverty line, near-poor denotes 100–125 percent of poverty level, low income denotes 125–199
percent poverty level, middle income denotes 200–399 percent of poverty level, and high income
denotes over 400 percent of poverty level.
§Ambulatory diagnostic groups are obtained from the Johns Hopkins Adjusted Clinical Group
System. ADG groups 16 and 19 have been discontinued.
For the PA1 group, N51,762. For physician-only group, N511,184.
Impact of Physician Assistant Care 1915
percent, there was no difference seen between the comparison groups. Above
the level of at least 25 percent of visits provided by a PA, substantive PA
proportion of PA visits used to define the PA1 group increases, the number
of office visits per year consistently decreases in a dose–response pattern
Table 3 shows the results of the secondary analysis examining whether
reduced office visit use in the PA1 group was offset by increased number of
visits in other settings. Persons in the PA1 group had about 25 percent fewer
emergency department visits (po.05). The results for hospital outpatient and
inpatient settings were not statistically significant.
In this national sample from the United States, adults who reported receiving
substantive care from PAs (30 percent or more of yearly office-based visits)
were younger, healthier, and less medically complex than those who reported
seeing physicians only. After controlling for demographics, socioeconomic
status, insurance status, health status, and medical conditions, these persons
had about 16 percent fewer office-based visits per year than those who
reported receiving only physician care.
Comparing the PA1 Group and the Physician Group at Different Cut-Points
for Percentage of Visits to PAsn
Incidence Rate Ratios for Yearly Office-Based Visit Resource Use
Percentage of Visits
to PAs only N (PA1 Group)
nNegative binomial model adjusted for age, gender, race, rural–urban status, insurance status,
poverty category, self-rated physical health, and 34 ambulatory diagnostic group indicator vari-
wPopulation-adjusted incidence risk ratio showing proportionate contrast in expected number of
office visits over the 1-year period for the exposed (PA1) group compared with the control
1916HSR: Health Services Research 43:5, Part II (October 2008)
The decreased number of office visits that we found with substantive PA
care may be due to differences between the comparison groups that were not
Physician Group for Other Health Care Settingsn
Incidence Rate Ratios for the PA1 Group Compared with the
Clinical SettingIncidence Rate Ratioz
Hospital-based outpatient clinicsn
nNegative binomial model adjusted for age, gender, race, rural–urban status, insurance status,
poverty category, self-rated physical health, and 34 ambulatory diagnostic group indicator vari-
wZero-inflated Poisson model adjusted for age, gender, race, rural–urban status, insurance status,
poverty category, self-rated physical health, and 34 ambulatory diagnostic group indicator variables.
zPopulation-adjusted incidence risk ratio showing proportionate contrast in the expected number
ofepisodesofcareover the1-yearperiodfor the exposed(PA1)groupcomparedwiththe control
provided by a PA.
% of yearly visits provided by PAs
Incidence rate ratio
Physician-Only Group at Different Cut-Points for Percent of Visits to PAsn
Incidence Rate Ratios for the PA1 group Compared with the
nPopulation-adjusted incidence risk ratio showing proportionate contrast in ex-
pected number of office visits over the 1-year period for the exposed (PA1) group
compared with the control (physician-only) group. Incidence risk ratio was not statis-
tically significant at the 10 percent cut-point.
Impact of Physician Assistant Care 1917
practices or health systems, PAs are assigned patients who are expected to
behind thesedecisions are relatively intangible and not reflected by our control
variables. Although the Adjusted Clinical Group
adjustment has been validated to predict health resource use (Weiner et al.
1992), risk adjustment can never completely eliminate selection bias. Our
results indicating that the group with substantive PA care also had fewer emer-
gency department visits suggest that this group may have been more healthy
than the physician-only group. Thus, residual selection bias may explain our
Another potential explanation for our finding is that, as persons have
more health care visits, they are more and more likely, by chance, to en-
counter a PA, but less likely to stay above the threshold cut-points (10, 25, 30
percent, etc.) used in our analysis. If this is true, our primary independent
variable (substantive PA care) could be endogenous with our outcome vari-
able (number of visits per year). For a discussion of this problem, and of our
efforts to address it, see Supplementary Appendix B.
When interpreting these findings, it is important to account for
the strengths and limitations of MEPS data for research on PAs. Because
household respondents may not accurately report the type of provider
who saw a patient, the ‘‘physician-only’’ group likely contains some persons
toa PAwhen aphysicianwasnotalsoseen,some persons inthecontrolgroup
were likely exposed to PAs on visits when they also saw the physician.
The magnitude of this contamination is difficult to quantify. Although MEPS
likely underrepresents the extent of PA participation in care (Morgan et al.
2007), MEPS still provides a relatively large national sample that is diverse
with regard to patient demographics, geography, socioeconomic status,
type of health plan, and health status. MEPS data provide one of the few
national representative longitudinal sources of information about national
patterns of health care use. The longitudinal design is well-suited to our
research question because it supports the analysis of a person’scare over time,
pattern of surveying respondents every 4–5 months minimizes recall
bias more than surveys that query respondents about events over a full year.
In sum, while MEPS has weaknesses for this research, it likely provides
one of the best data sources available to investigate the effect of PAs on
sSystem that we used for risk
1918HSR: Health Services Research 43:5, Part II (October 2008)
Overall, these results indicate that under the practice conditions and
relative prevalence of PAs and physicians in the health workforce between
1996 and 2004, PAs tended to serve as substitutes, rather than complements,
for physician services. This suggests that an increase in the number of PAs in
the provider mix should not be expected to increase per person office visit
resource use. It is not possible to predict whether there is some point at which
the impact of adding PAs to the workforce might change. It is possible that,
when the ratio of PAs to physicians reaches a certain point, addition of more
PAs could begin to change the type and amount of services provided.
Our results indicate that, after adjustment for a variety of indicators of patient
complexity, use of PAs as the sole provider for a substantive portion of office-
Our study found that a group of adults with 30 percent or more of yearly visits
attended by PAs alone had, on average, about 16 percent fewer visits per year.
will not increase per person office visit resource use. In this respect, our findings
indicatethatPAsservemoretoextend physician servicestopatientsthan toplay
a complementary role that leads to increased health care resource use.
2006)materialize,and iftherapid growthofthePA profession continues(Hooker
2002), PAs will provide a larger share of patient care in the United States in the
study suggest that the use of PAs may increase efficiency in health care delivery.
Joint Acknowledgment/Disclosure Statement: This research was funded in part by
an unrestricted grant from the Physician Assistant Education Association and
the American Academy of Physician Assistants.
The authors would like to thank Melissa Gregg for her help with data
management and S. Philip Morgan for his comments on the manuscript.
Impact of Physician Assistant Care1919
Agency for Healthcare Research and Quality. 2006. ‘‘Medical Expenditure Panel
Survey: Data and Statistics’’ [accessed on May 30, 2006]. Available at: http://
American Academy of Physician Assistants. 2007. ‘‘2006 AAPA Physician Assistant
Census Report’’ [accessed on April 9, 2008]. Available at http://www.aapa.org/
Andersen, R. M. 1995. ‘‘Revisiting the Behavioral Model and Access to Medical Care:
Does It Matter?’’ Journal of Health and Social Behavior 36 (1): 1–10.
Association of American Medical Colleges. 2006. AAMC Statement on the Physician
Workforce, pp 9. Washington, DC: Association of American Medical Colleges.
Byers, A. L., H. Allore, T. M. Gill, and P. N. Peduzzi. 2003. ‘‘Application of Negative
Binomial Modeling for Discrete Outcomes: A Case Study in Aging Research.’’
Journal of Clinical Epidemiology 56 (6): 559–64.
Carzoli, R. P., M. Martinez-Cruz, L. L. Cuevas, S. Murphy, and T. Chiu. 1994. ‘‘Com-
parison of Neonatal Nurse Practitioners, Physician Assistants, and Residents in
the Neonatal Intensive Care Unit.’’ Archives of Pediatrics and Adolescent Medicine
148 (12): 1271–6.
Dial, T. H., S. E. Palsbo, C. Bergsten, J. R. Gabel, and J. Weiner. 1995. ‘‘Clinical
Staffing in Staff- and Group-Model HMOs.’’ Health Affairs 14 (2): 168–80.
Care by Nonphysician Clinicians in the United States.’’ New England Journal of
Medicine 348 (2): 130–7.
Duan, N., W. Manning, C. Morris, and J. Newhouse. 1983. ‘‘A Comparison of
Alternative Models for the Demand for Medical Care.’’ Journal of Business and
Economic Statistics 1 (2): 115–26.
Greene, W. H. 1994. ‘‘Accounting for Excess Zeroes and Sample Selection
in Poisson and Negative Binomial Regression Models.’’ Working Paper No.
EC-94-10. Stern School of Business, New York University, New York,
Grzybicki, D. M., P. J. Sullivan, J. M. Oppy, A. M. Bethke, and S. S. Raab. 2002.
‘‘The Economic Benefit for Family/General Medicine Practices Employing
Physician Assistants.’’ American Journal of Managed Care 8 (7): 613–20.
Hooker, R. S. 1993. ‘‘The Roles of Physician Assistants and Nurse Practitioners
in a Managed Care Organization.’’ In The Role of Physician Assistants and Nurse
Practitioners in Primary Care, edited by D. K. Clawson and M. Osterweis, pp.
51–68. Washington, DC: Association of Academic Health Centers.
—— —— ——. 2002. ‘‘A Cost Analysis of Physician Assistants in Primary Care.’’ Journal of the
American Academy of Physician Assistants 15 (11): 39–42, 45, 48 passim, 2002 Nov.
Hooker, R. S., and L. E. Berlin. 2002. ‘‘Trends in the Supply of Physician Assistants
and Nurse Practitioners in the United States.’’ Health Affairs 21 (5): 174–81.
Hooker, R. S., and L. F. McCaig. 2001. ‘‘Use of Physician Assistants and Nurse Prac-
titioners in Primary Care.’’ Health Affairs 20 (4): 231–8.
1920HSR: Health Services Research 43:5, Part II (October 2008)
Horrocks, S., E. Anderson, and C. Salisbury. 2002. ‘‘Systematic Review of Whether
Nurse Practitioners Working in Primary Care Can Provide Equivalent
Care to Doctors.’’ British Medical Journal 324 (7341): 819–23.
Lambert, D. 1992. ‘‘Zero-Inflated Poisson Regression, with an Application to Defects
in Manufacturing.’’ Technometrics 34 (1): 1–14.
Larson, E. H., L. G. Hart, and R. Ballweg. 2001. ‘‘National Estimates of Physician
Assistant Productivity.’’ Journal of Allied Health 30 (3): 146–52.
Laurant, M. G., R. P. Hermens, J. C. Braspenning, B. Sibbald, and R. P. Grol. 2004.
‘‘Impact of Nurse Practitioners on Workload of General Practitioners: Ran-
domised Controlled Trial.’’ British Medical Journal 328 (7445): 17.
Medical Group Management Association (MGMA). 2006. Physician Compensation and
Production Survey: 2006 Report Based on 2005. Englewood, CO: Medical Group
Miller, W., E. Riehl, M. Napier, K. Barber, and H. Dabideen. 1998. ‘‘Use of Physician
Assistants as Surgery/Trauma House Staff at an American College of Surgeons-
Verified Level II Trauma Center.’’ Journal of Trauma Injury Infection and Critical
Care 44 (2): 372–6.
Morgan, P. 2007. ‘‘Physician Assistants in Office-Based Medical Care.’’ Doctoral Dis-
sertation in Population Health Sciences. Madison, WI: University of Wisconsin-
Morgan, P., J. Strand, T. Ostbye, and M. Albanese. 2007. ‘‘Missing in Action: Care by
Services Research 42 (5): 2022–37.
Mundinger, M. O., R. L. Kane, E. R. Lenz, A. M. Totten, W. Y. Tsai, P. D. Cleary,
W. T. Friedewald, A. L. Siu, and M. L. Shelanski. 2000. ‘‘Primary Care Out-
comes inPatients Treatedby
A Randomized Trial.’’
283 (1): 59–68.
Trends in the Health of Americans. Hyattsville, MD: National Center for Health
Record, J. C., M. McCally, S. O. Schweitzer, R. M. Blomquist, and B. D. Berger.
1980. ‘‘New Health Professions after a Decade and a Half: Delegation, Produc-
tivity and Costs in Primary Care.’’ Journal of Health Politics, Policy and Law 5:
Roblin, D. W., D. H. Howard, E. R. Becker, E. Kathleen Adams, and M. H. Roberts.
2004. ‘‘Use of Midlevel Practitioners to Achieve Labor Cost Savings in the
Primary Care Practice of an MCO.’’ Health Services Research 39 (3): 607–26.
Schulman, M., K. R. Lucchese, and A. C. Sullivan. 1995. ‘‘Transition from
Housestaff to Nonphysicians as Neonatal Intensive Care Providers: Cost, Im-
pact on Revenue, and Quality of Care.’’ American Journal of Perinatology 12 (6):
Stoddard, J. J., D. A. Kindig, and D. Libby. 1994. ‘‘Graduate Medical Education
Reform.’’ Journal of the American Medical Association 272 (1): 53–8.
Venning, P., A. Durie, M. Roland, C. Roberts, and B. Leese. 2000. ‘‘Randomised
Controlled Trial Comparing Cost Effectiveness of General Practitioners
Impact of Physician Assistant Care 1921
and Download full-text
320 (7241): 1048–53.
Weiner, J. P., B. H. Starfield, R. N. Lieberman, J. P. Weiner, B. H. Starfield, and R. N.
Lieberman. 1992. ‘‘Johns Hopkins Ambulatory Care Groups (ACGs).’’ HMO
Practice 6 (1): 13–9.
The following supplementary material for this article is available online:
Appendix SA1: Author Matrix.
Appendix A: Additional Graphic Description of Data.
Figure A1: Frequency Distribution of Number of Visits per Year to PAs.
Figure A2: Frequency Distribution of Number of Visits per Year to
Physician-Only Group and to PA1 Group.
Figure A3: Distribution of Portion of Visits Provided by PAs Among
Persons Who Saw PAs.
Table A1: Distribution of Number of Visits per Year, Total Sample.
Appendix B: Endogeneity Bias and Alternative Analyses.
Table B1: Quantile Regression Analysis Results.
Table B2: Propensity Score Analysis with Stratified Matching.
This material is available as part of the online article from http://
(this link will take you to the article abstract).
Please note: Blackwell Publishing is not responsible for the content or
functionality of any supplementary materials supplied by the authors. Any
queries (other than missing material) should be directed to the corresponding
author for the article.
1922 HSR: Health Services Research 43:5, Part II (October 2008)