Effects of Competition on Infertility Treatment Outcomes

1
Effects of Competition on Infertility Treatment
Outcomes

• Melinda Henne
• Research In Progress Seminar
• Stanford University
• June 1, 2005

2
Research In Progress

• Results are preliminary
• Please do not cite or reference

3
Acknowledgements

• Kate Bundorf
• Laurence Baker
• Jia Chan
• Shannon McConnville

4
Outline

• Multiple Births and ART overview
• Literature on Competition
• Model
• Definition of Market
• Measure of Competition
• Equations
• Results
• Conclusions

5
Outline

• Multiple Births and ART overview
• Literature on Competition
• Model
• Definition of Market
• Measure of Competition
• Equations
• Results
• Conclusions

6
Trends in multiple birth rates
7
Increasing utilization of ART in the U.S.
8
Trends in ART Birth Rates
9
ART Cycle
Embryo Transfer
Pregnancy Test
Egg Retrieval
3 weeks Birth Control Pills
3-5 days
2-3 weeks Daily shots frequent office visits
2 weeks Progesterone Treatment (shots, vaginal)
10,000-15,000
10
Number of ART Clinics in United States
11
Fertility Inc. Clinics Race to Lure Clients

• Infertility has become a big, fiercely
  competitive business, with a billion dollars in
  revenues and with more and more doctors fighting
  for a limited number of patients. The growth of
  the field has been fueled by rising success rates
  and increased demand from patients . . .
• critics say, they may implant too many embryos .
  . .
• Since most patients pay with their own money . .
  .fertility specialists say that if they want to
  survive, they have to get the attention of both
  patients and referring doctors

New York Times, January 1, 2002
12
Outline

• Multiple Births and ART overview
• Literature on Competition
• Model
• Definition of Market
• Measure of Competition
• Equations
• Results
• Conclusions

13
Competition in Healthcare

• Hospital Based
• Emergent or Urgent Care
• Mortality after AMI
• Mortality from Pneumonia
• Effects
• Lower costs, increased quality, decreased
  mortality, improved efficiency
• Excess capacity, higher costs, worse outcomes

14
Steiner

• Cross sectional analysis, 2000
• Competition Area
• Metropolitan Statistical Area or County
• Competition (low 1-2, intermediate 2-7,
  highgt7)
• Demand (low lt130,000, intermediate
  130,000-600,000, high gt600,000)
• Conclusion
• As competition increased, the percentage of
  high-order multiple pregnancies decreased

2002
15
Hamilton and McManus

• Time-series analysis, 1995-2000
• Competition areaCombined or Metropolitan
  Statistical Area
• Competition
• HHI
• Clinics (dichotomousmonopoly or not)
• Demographic controls
• Conclusion
• Clinics in areas with more than one clinic
  transfer fewer embryos and have lower multiple
  birth rates

2004, 2005
16
Outline

• Multiple Births and ART overview
• Literature on Competition
• Model
• Definition of Market
• Measure of Competition
• Equations
• Results
• Conclusions

17
Questions??

• What happens to clinic volume with more clinics
  in an area?
• Are practices different when there is more
  competition?
• How do clinics try to improve birth rates?
• Do clinics transfer more embryos?
• Do clinics cherry pick patients?
• How do clinics try to reduce multiple birth
  rates?
• Do clinics sacrifice birth rates for this goal?
• Do clinics transfer fewer embryos?
• Do clinics have to take all patients, regardless
  of prognosis?

18
Competition Area
Fixed Radius

• Geographic Area

C
C
D
D
A
B
E
A
B
E

1. Number of Clinics

1. Number of clinics
2. HHI

19
Location of ART clinics in U.S in 2002
SART, 2002
20
Outline

• Multiple Births and ART overview
• Literature on Competition
• Model
• Definition of Market
• Measure of Competition
• Equations
• Results
• Conclusions

21
Measure of Competition

• Number of clinics within defined area
• Continuous
• Positive skew (log transformed)
• Categorical definitions (monopoly, low
  competition2-3, low-medium competition 4-5,
  high-medium competition 6-10, high competition
  11-15, very high competition 16)
• Herfindahl-Hirschman Index (HHI)
• Sum of market shares
• Captures both the size and number of competitors

22
Outline

• Multiple Births and ART overview
• Literature on Competition
• Model
• Measure of Competition
• Definition of Market
• Equations
• Results
• Conclusions

23
Dependent Variables

• Cycles performedpositive skew (log transformed)
• Births per CycleNormal
• Multiples per CyclesNormal
• Multiples per BirthNormal
• Number of Embryos TransferredNormal
• Proportion of young patientsNormal

24
Summary Statistics
Mean S.D. Min Max
Cycles (lCycles) 206.71 (4.55) 276.05 (1.07) 0 (0) 3328 (8.11)
Births/Cycle 0.21 0.09 0 1
Multiples/Cycle 0.08 0.06 0 1
Multiples/Birth 0.34 0.17 0 1
Embryos Transferred 3.63 0.77 1 7.59
Embryos Transferred (lt35) 3.47 0.80 1 9.7
Embryos Transferred (35) 3.64 0.75 1 7.59
Proportion of Patients lt35 0.46 0.12 0 1
25
Independent Variables

• Time varying market characteristics
• Population
• Minority Rate
• Income
• Unemployment Rate
• Education
• Instruments
• Non-federal MDs
• Medical Schools
• Number of Hospitals

• Measure of Competition
• Year, Market Fixed Effects
• Clinic Characteristics
• Size
• Years in Business

26
Summary statistics
Mean S.D. Min. Max.
Women age 25-44 (1000) 291.30 474.24 10.30 3465.80
Proportion age 25-29 () 23.62 1.97 17.07 36.13
Proportion age 30-34 () 25.33 1.56 21.53 28.94
Proportion age 35-39 () 26.37 1.05 20.10 29.94
Minority Rate () 22.91 14.45 1.60 79.73
Per Capita Income (1000) 27.38 3.76 18.41 48.48
Unemployment Rate () 5.27 3.98 1.32 43.52
Proportion with College Degree 24.02 5.61 12.26 42.15

Number of non-Federal MDs 4293.16 9262.10 120 85,773
Presence of Medical School 0.68 0.47 0 1
Number of Hospitals 31.48 41.42 1 294
27
Expanded market and individual clinic volume
Cycles/1000m,t a b1Clinicsm,t b2Xm,t
b3FE e Cyclesc,m,t a g1Clinicsm,t
g2Xm,t g3 FE e

• X vector of market-time characteristics
• FE market and year fixed effects
• Instruments for clinics
• Non-federal MDs
• Medical Schools
• Number of Hospitals

• Possibilities
• b1, g1 gt0
• b1 gt0, g1 lt0

28
Does competition affect treatment and outcomes?
Yc,m,t a b1Cc,m,t b2Xm,t b3 FE e

• Y outcome
• C measure of competition
• X vector of market-year characteristics
• F market and year fixed effects

29
Outline

• Multiple Births and ART overview
• Literature on Competition
• Model
• Measure of Competition
• Definition of Market
• Equations
• Results
• Conclusions

30
Number of Cycles Performed(log transformed)
ln(Cycles)/1000 ln(Cycles)/Clinic
ln (Clinics) 1.286 (0.100) -1.005 0.185

2-3 Clinics 0.992 0.102 -0.810 0.175
4-5 Clinics 1.485 0.143 -1.260 0.221
6-10 Clinics 1.996 0.168 -1.506 0.271
11-15 Clinics 2.213 0.199 -1.687 0.319
gt15 Clinics 2.581 0.266 -2.000 0.339
31
Clinic volume compared to market volume
Mean ln(Cycles)/1000 1.04
ln(Number of Cycles)
Mean ln(Cycles) 4.55
Number of Clinics
32
Births Per Cycle
CSA Radius 20 Radius 50
2-3 Clinics -0.004 0.008 0.006 0.006 0.009 0.008
4-5 Clinics -0.014 0.011 -0.010 0.008 0.006 0.011
6-10 Clinics -0.001 0.013 -0.010 0.010 0.006 0.013
11-15 Clinics -0.010 0.016 -0.014 0.010 -0.010 0.015
gt15 Clinics 0.011 0.018 -0.033 0.011 -0.015 0.017
F-test 0.97 (0.436) 3.79 (P 0.002) 1.47 (P 0.195)
33
Multiples per Birth
CSA Radius 20 Radius 50
2-3 Clinics 0.046 0.023 0.006 0.006 -0.009 0.018
4-5 Clinics 0.063 0.033 0.010 0.0197 0.020 0.024
6-10 Clinics 0.018 0.040 0.001 0.019 -0.024 0.028
11-15 Clinics 0.014 0.049 -0.027 0.022 -0.036 0.034
gt15 Clinics 0.016 0.054 -0.055 0.024 -0.057 0.037
F-test 1.47 (0.195) 2.37 (P 0.037) 1.21 (P 0.300)
34
Mandatory Reporting
Multiples Per Birth
Births Per Cycle
Before 1995 After 1995
2-3 Clinics 0.009 0.034 0.007 0.018
4-5 Clinics -0.042 0.039 0.038 0.021
6-10 Clinics -0.032 0.041 0.015 0.022
11-15 Clinics -0.031 0.045 -0.027 0.027
gt15 Clinics -0.124 0.083 -0.041 0.026
Before 1995 After 1995
2-3 Clinics -0.003 0.013 0.006 0.009
4-5 Clinics -0.018 0.015 -0.006 0.011
6-10 Clinics -0.022 0.016 -0.009 0.011
11-15 Clinics -0.022 0.017 -0.017 0.014
gt15 Clinics -0.088 0.032 -0.030 0.013
35
Embryos Transferred
Overall Under 35 Over 35
2-3 Clinics -0.087 0.072 -0.066 0.070 -0.094 0.070
4-5 Clinics -0.082 0.085 -0.005 0.082 -0.065 0.082
6-10 Clinics -0.063 0.089 0.044 0.086 -0.068 0.086
11-15 Clinics -0.085 0.107 -0.029 0.104 -0.098 0.104
gt15 Clinics -0.108 0.104 0.128 0.101 -0.136 0.102
F-Test 0.39 (P 0.856) 1.64 (P 0.146) 0.56 (P 0.733)
36
Patient Selection
Proportion of Patients lt35
2-3 Clinics -0.007 0.011
4-5 Clinics -0.030 0.013
6-10 Clinics -0.038 0.014
11-15 Clinics -0.073 0.016
gt15 Clinics -0.069 0.016
37
Age of Clinic
Multiples Per Birth
Births Per Cycle
New (1-2 Years) Old
2-3 Clinics 0.068 0.025 -0.19 0.011
4-5 Clinics 0.061 0.027 -0.033 0.013
6-10 Clinics 0.056 0.028 -0.030 0.013
11-15 Clinics 0.033 0.033 -0.036 0.016
gt15 Clinics 0.009 0.032 -0.050 0.016
New (1-2 Years) Old
2-3 Clinics 0.290 0.081 -0.016 0.028
4-5 Clinics 0.286 0.089 0.034 0.033
6-10 Clinics 0.290 0.094 0.014 0.035
11-15 Clinics 0.168 0.110 -0.051 0.042
gt15 Clinics 0.181 0.105 0.017 0.040
38
Age of Clinic
Proportion of young patients
Embryos Transferred
New (1-2 Years) Old
2-3 Clinics 0.225 0.178 -0.166 0.079
4-5 Clinics 0.373 0.194 -0.181 0.094
6-10 Clinics 0.540 0.202 -0.205 0.098
11-15 Clinics 0.552 0.239 -0.251 0.118
gt15 Clinics 0.610 0.234 -0.311 0.114
New (1-2 Years) Old
2-3 Clinics 0.013 0.030 -0.011 0.011
4-5 Clinics 0.006 0.033 -0.048 0.013
6-10 Clinics -0.019 0.034 -0.044 0.014
11-15 Clinics -0.110 0.041 -0.064 0.017
gt15 Clinics -0.107 0.040 -0.058 0.016
39
Size of Clinic
Multiples per Birth
Births per Cycle
Small (lt120) Medium (120-500) Large (gt500)
2-3 Clinics 0.019 0.025 -0.021 0.016 0.028 0.022
4-5 Clinics 0.038 0.033 -0.027 0.018 0.024 0.026
6-10 Clinics 0.008 0.038 -0.010 0.019 0.016 0.024
11-15 Clinics -0.024 0.046 -0.049 0.021 0.052 0.026
gt15 Clinics -0.079 0.050 -0.049 0.024 0.03 0.027
Small (lt120) Medium (120-500) Large (gt500)
2-3 Clinics 0 0.010 0.021 0.009 -0.018 0.018
4-5 Clinics -0.019 0.0120 0.018 0.010 -0.016 0.021
6-10 Clinics -0.025 0.014 0.025 0.011 0 0.020
11-15 Clinics -0.026 0.018 0.019 0.012 0.004 0.021
gt15 Clinics -0.042 0.019 -0.007 0.014 -0.023 0.022
40
Size of Clinic
Proportion of young patients
Embryos Transferred
Small (lt120) Medium (120-500) Large (gt500)
2-3 Clinics -0.008 0.018 0.011 0.016 -0.021 0.021
4-5 Clinics -0.044 0.021 0.015 0.018 -0.056 0.026
6-10 Clinics -0.063 0.023 0.017 0.018 -0.044 -/-23
11-15 Clinics -0.123 0.029 -0.003 0.021 -0.048 0.025
gt15 Clinics -0.097 0.028 -0.018 0.021 -0.100 0.025
Small (lt120) Medium (120-500) Large (gt500)
2-3 Clinics -0.162 0.109 -0.054 0.120 -0.160 0.170
4-5 Clinics 0.007 0.126 -0.207 0.140 0.004 0.210
6-10 Clinics -0.068 0.139 -0.081 0.141 -0.185 0.184
11-15 Clinics -0.002 0.175 -0.060 0.158 -0.486 0.202
gt15 Clinics -0.038 0.165 -0.190 0.158 -0.297 0.200
41
Results Summary

• As more clinics enter an area, the market is
  expanded, but each individual clinic performs
  fewer cycles
• Outcome differences are observed in a narrow
  radius
• Embryo transfer practices do not appear to be
  affected by the number of neighboring clinics
• Clinics with many neighbors are less able to
  select for patients with higher liklihood of
  success

42
Results Summary-2

• Young clinics
• Have higher birth rates and multiple birth rates
• Have a smaller proportion of young patients
• Transfer more embryos per cycle
• Small clinics and large clinics
• Have a smaller proportion of young patients
• Have lower birth rates, but higher multiple birth
  rates
• Small clinics transfer more embryos per cycle
• Large clinics transfer fewer embryos per cycle

43
Conclusions

• The definition of the market is vital to
  determine the impact of competition among IVF
  clinics
• Patient selection probably plays a more
  significant impact on outcomes than differences
  in practice behavior
• Mandatory reporting of success rates may
  influence practice patterns and patient selection
• Clinic characteristics influence outcomes and may
  confound results of previous competition in IVF
  studies