P1253814504SNmaB

1
Determinants of First Practice Location Choice by
New Physicians
Chiu-Fang Chou1,2, Dr.PH and Anthony T. Lo
Sasso2, Ph.D., Midwest Center for Health
Workforce Studies1, Division of Health Policy and
Administration2 University of Illinois at
Chicago, Chicago, Illinois
Table 1 presents the characteristics of locations
chosen by new physicians trained from NY and CA
during 1998-2003. Malpractice premiums average
nearly 65,000 annually for OB/GYNs and roughly
half that figure for surgeons. PCPs by contrast
faced premiums of roughly 12,000 per year.
Beyond the striking difference in malpractice
premiums, location differences between the three
specialist types were not particularly apparent.
Surgeons appeared somewhat more likely to locate
in states with malpractice damage award caps.
Results
Introduction
Table 4 Physician Fixed Effects Linear
Probability Model Results for Location Choice,
Selected Coefficients for
Race/Ethnicity Interaction Terms

• Table 4 Racial/Ethnicity matching
  Racial/Ethnicity matching plays an important role
  on location choices for new physicians. That is,
  minority doctors appear to prefer to practice in
  areas that have more population of their own race
  and ethnicity. For example, in the Hispanic
  community, the results indicate that 1 increase
  in the proportion of the population that is
  Hispanic makes a Hispanic OB/GYN 13.7 increase
  in the probability of locating in an area. The
  estimate is statistically significant for OB/GYNs
  (plt.05), PCPs (plt.01) and surgeons (plt.1).

OB/GYNs Surgeons PCPs
Proportion of black population -0.000432 -0.000112 0.000154
(0.000433) (0.000269) (0.000089)
Proportion of Asian population -0.001059 -0.000465 -0.000547
(0.000499) (0.000167) (0.000076)
Proportion of Hispanic population 0.000314 0.000124 0.000418
(0.000356) (0.000195) (0.000064)
Proportion of other population 0.001154 0.001628 0.000727
(0.002702) (0.001285) (0.000457)
Proportion of black population Black physicians 0.000264 0.000065 0.000175
(0.000074) (0.000069) (0.000023)
Proportion of Asian population Asian physicians 0.000381 0.000239 0.000134
(0.000122) (0.000080) (0.000030)
Proportion of Hispanic population Hispanic 0.000368 0.000130 0.000127
physicians (0.000143) (0.000076) (0.000018)
Proportion of other population Other race -0.000302 -0.000188 0.000263
physicians (0.000102) (0.000056) (0.000169)
MSA Fixed Effects Yes Yes Yes
Observations 77859 199581 904960
Number of physicians 633 937 3232
R2 0.10 0.04 0.07
Figure 1 displays population weighted averages of
malpractice premiums over time. Premiums grew
quite slowly between 1998 and 2000, but they
increased rapidly after 2000, with OB/GYN
premiums increasing roughly 50 by 2003 and
surgeon and PCP premiums increasing over 70 by
2003. In 2003 OB/GYN premiums were still nearly
1.5 times greater than premiums faced by surgeons
and nearly 5 times greater than premiums faced by
PCPs.
This study is aimed at understanding how new
physicians choose their initial practice
locations. There is considerable disagreement on
the role of malpractice premiums on physicians
practice decisions. New physicians practice
location decisions can have a lasting impact on
the future healthcare workforce because
relocation can be particularly costly for
physicians. The objectives of this study are
threefold (1) The impact of malpractice premiums
and laws affecting premiums (2) Policies aimed
at encouraging physicians to practice in
underserved areas (3) The ethnic and racial
backgrounds of physicians.
Table 1 Summary Statistics of Characteristics of
Location chosen by Physicians
OB/GYNs OB/GYNs Surgeons Surgeons PCPs PCPs
N 633 7 937 10 3230 37
Variable Mean Std Dev Mean Std Dev Mean Std Dev
Personal Characteristics
Age 32.17 3.78 33.39 3.16 32.68 4.96
US citizen 0.96 0.97 0.84
Race and Ethnicity
White 0.56 0.66 0.46
Black 0.12 0.04 0.07
Asian 0.21 0.20 0.33
Hispanic 0.04 0.04 0.08
Other 0.05 0.04 0.04
Gender (Male) 0.30 0.83 0.52
Educational Debt
0 0.18 0.23 0.36
lt100,000 0.42 0.43 0.34
100,00125,000 0.30 0.26 0.23
gt125,000 0.07 0.06 0.06
Location Characteristics
Malpractice premiums 64,685.13 27,138.06 32,734.26 17,501.79 11,999.00 5,799.00
State with malpractice damage cap 0.31 0.46 0.36 0.48 0.33 0.47
Physician hourly wage 68.09 11.60 72.34 11.73 57.79 11.66
Health professional shortage area 0.13 0.10 0.12 0.10 0.14 0.12
Proportion of population by race
White 0.59 0.20 0.63 0.20 0.62 0.21
Black 0.16 0.11 0.14 0.10 0.14 0.11
Asian 0.07 0.06 0.06 0.07 0.07 0.06
Hispanics 0.18 0.13 0.17 0.13 0.18 0.14
Other 0.01 0.002 0.01 0.002 0.01 0.002
Same-specialty physician per 100,000 population 15.08 5.31 20.93 37.74 34.50 23.93
Resident physicians per 100,000 population 58.46 40.18 45.88 39.45 48.21 39.55
Hospital beds per 100,000 population 375.91 117.85 367.68 122.53 369.85 117.87
Hypothesis are the following Hypothesis 1 New
physicians will be more likely to practice in a
place with lower malpractice premium rate.
Hypothesis 2 New physicians will be more
likely to practice in a place with a malpractice
damage award cap. Hypothesis 3 New physicians
with higher educational debt will be more likely
to practice in a health professional shortage
area. Hypothesis 4 New physicians will be more
likely to locate in areas with higher proportion
of the population of their race.
Conclusions

• The results suggest that malpractice premiums
  may be an important factor for some risky
  subspecialties. New surgeons were less likely to
  practice in areas with increasing malpractice
  premiums. OB/GYNs were less sensitive to
  malpractice premiums than were surgeons. PCPs
  appeared to be drawn to areas with higher
  malpractice premiums, suggesting a potential
  substitution effect.
• Symmetric results were observed for the impact of
  state damage award caps. New surgeons and OBs
  were more likely to practice in areas with caps.
  New PCPs were less likely to practice in areas
  with caps.
• The results suggest that health professional
  shortage areas may not be an important factor.
  Only OB/GYNs and PCPs without debt (18 and 36of
  the respective samples) were more likely to
  practice in areas with HPSAs. Suggesting the
  program as structured does not appear to draw
  physicians.
• Racial/ Ethnicity matching plays an important
  role on location choices for new physicians. he
  estimates are relevant for policy given the
  amount of interest in increasing the presence of
  minority physicians in predominantly minority
  communities. Educational debts is also found to
  be a factor influencing location choice of new
  physicians.
• Limitation of this study
• This study used the new physicians who finished
  their residency training in New York and
  California states, which are urban state.
• Malpractice data is limited because it does not
  have detailed information on the number of
  physicians that each company represents in the
  state so it is rough average premium rather than
  a weighted average premium.

Materials and Methods
Statistical analysis involved fixed effects
models to examine the factors affecting the
choice of initial practice location by new
physicians. Data are from a unique survey of
exiting medical residents acquired by the
HRSA-funded New York Workforce Center at SUNY
Albany. These data are matched to data on
malpractice premiums from Medical Liability
Monitor. Additional location information is from
the Area Resource File (ARF). The sample
consists of 9,137 physicians who just finished
their residency training in New York and
California in 1998-2003 and who are beginning
their careers in patient care. The dependent
variable is the choice of location among the 357
metropolitan statistical areas (MSAs) and
non-metropolitan areas within each state in the
United States. Where appropriate, independent
variables have been weighted by area population.
Other local market characteristics include the
number of hospital beds, per capita income, and
the local unemployment rate.
This map displays MSAs chosen by new physicians.
The most frequently chosen option for all three
specialty groups in the sample located in New
York City. The second and third most common
location was Nassau-Suffolk, NY (Long Island) and
Los Angeles-Long Beach, CA.
Table 2 Physician Fixed Effects Linear
Probability Model Selected Results for Location
Choice, with and without Fixed
Location Effects
OB/GYNs OB/GYNs Surgeons Surgeons PCPs PCPs
Malpractice premiums (1000s) 0.000095 0.000011 -0.000003 -0.000052 0.000119 0.000118
(0.000019) (0.000035) (0.000012) (0.000024) (0.000019) (0.000031)

Damage award cap -0.00117 0.001432 -0.000658 0.002796 -0.001000 -0.000119
(0.000671) (0.001857) (0.000310) (0.000957) (0.000116) (0.000303)

Health Professional Shortage 0.01482 0.002408 0.003137 0.001041 0.001813 0.000126
Area (0.00223) (0.003428) (0.001071) (0.001289) (0.000243) (0.000320)

Physician hourly wage 0.00030 0.000049 0.000109 0.000042 0.000061 -0.000042
(0.00004) (0.000087) (0.000019) (0.000047) (0.000007) (0.000018)

Same-specialty physicians per -0.000264 0.000092 0.000014 0.000012 0.000090 0.000055
100,000 population (0.000176) (0.000281) (0.000008) (0.000008) (0.000007) (0.000007)

Resident physicians per 0.000023 0.000109 0.000011 0.000234 0.000004 0.000071
100,000 Population (0.000024) (0.000067) (0.000014) (0.000074) (0.000004) (0.000013)

Hospital beds per 100,000 0.000011 0.000008 0.000008 0.000007 0.000007 0.000006
Population (0.000003) (0.000020) (0.000001) (0.000006) (0.000000) (0.000002)

State Loan Repayment 0.003601 --- 0.001358 --- 0.001102 ---
Program (0.000595) (0.000296) (0.000110)

Constant -0.025456 -0.047313 -0.014833 0.000076 -0.009456 -0.028974
(0.005402) (0.058552) (0.003053) (0.013512) (0.001031) (0.005206)
MSA fixed effects No Yes No Yes No Yes
Observations 77859 77859 199581 199581 904960 904960
Number of physicians 633 633 937 937 3232 3232
R2 0.05 0.10 0.02 0.04 0.03 0.07

• Table 2 shows the regression results.
• Malpractice premiums Increases in malpractice
  premiums have a negative and significant effect
  (plt.05) on the probability of choosing to locate
  in a given area for surgeons. No significant
  effect on the probability of choosing to locate
  in a given area for OB/GYNs, a statistically
  significant positive effect for PCPs. The
  magnitude of the result indicates that a one
  standard deviation increase in malpractice
  premiums for surgeons (approximately 17,500)
  would make a surgeon 0.1percentage point less
  likely to move to a given location, or roughly
  8.
• Malpractice damage caps Statically significant
  (plt.01) for surgeons and no statistically
  significant effect for OBs. The magnitude of the
  result indicates that surgeons are 0.28 of a
  percentage point more likely to locate in an area
  with a cap suggesting a 23.9 (for surgeons)
  increase in the probability of locating in an
  area.

Implications for Policy

• The determinants of first practice location
  choices for physicians vary by specialty.
• Policy makers who create programs regarding the
  physician workforce need to consider the
  different needs of physicians in different
  specialties.
• Increasing malpractice premiums do appear to
  deter some specialists from locating in certain
  area although malpractice damage award caps
  may be an effective tool for state policy makers
  to attract some traditionally high malpractice
  premium specialists.
• Policy makers need to reevaluate the existing
  programs regarding HPSAs and to study the impacts
  of these program on physicians distribution or
  promote these programs to new physicians.

NY/CA Resident Exit Data
22,504
Fellowship/others
Primary Care
12,695 (56.41)
9,809 (43.59)
Not Accepted Jobs/Others
Accepted Jobs
10,080 (79.4)
2,615 (20.60)
Missing Zip/City
Practice Location
ARF
9,137 (90.64)
943 (9.36)
FIPS County Code
MLM
Table 3 Physician Fixed Effects Linear
Probability Model Results for Location Choice,
Selected Coefficients for
HPSA-Physician Debt Interaction Terms
MSA Code
BLS
OB/GYNs Surgeons PCPs
Health professional shortage area 0.006543 -0.000482 0.002874
(0.004020) (0.001917) (0.000519)
Educational debt greater than 0 but less than -0.006532 0.000875 -0.003712
100,000 HPSA (0.003539) (0.002085) (0.000605)
Educational debt between 100,000 and -0.002288 0.004056 -0.004967
125,000 HPSA (0.004587) (0.002690) (0.000618)
Educational debt 125,000 or moreHPSA -0.007823 -0.000760 -0.004656
(0.008715) (0.002382) (0.000660)
MSA Fixed Effects Yes Yes Yes
Observations 77859 199581 904960
Number of physicians 633 937 3232
R2 0.10 0.04 0.07
Acknowledgements
Cap
Location Characteristics
Malpractice premiums
Personal Characteristics
9,137
We gratefully acknowledge the contributions of
Edward K.Mensah, Ph.D., Judith A. Cooksey, MD,
MPH, Surrey M. Walton, PhD, Lorens A. Helmchen,
Ph.D., and Janelle Yi-Ju Lee, Dr. PH. Data for
this study was provided by Gaetano Forte from the
Center for Health Workforce Studies of the State
University of New York at Albany and Robert
Kaestner, Ph.D. Funding for this project was
provided by the National Center for Health
Workforce Analysis of the Bureau of Health
Professions of the Health Resources and Services
Administration, and the University of Illinois at
Chicago.

• Table 3 Health professional shortage areas
  OB/GYNs and surgeons are generally unresponsive
  to shortage areas regardless of debt level.
  However, PCPs with no debt are significantly more
  likely to locate in areas with higher HPSA
  values. The magnitude of the result for PCPs
  without educational debt suggests that they are
  14 more likely to locate in an area that is
  entirely a HPSA.

Contact Information Chiu-Fang Chou, e-mail
address cchou4_at_uic.edu or cchou4_at_gmail.com