The Association between the Distribution of Obstetricians and Birth Outcomes in the US, 19922002

1
The Association between the Distribution of
Obstetricians and Birth Outcomes in the US,
1992-2002
Maternal and Child Health Epidemiology
Conference December 2007 Michael Kogan,
Ph.D. U.S. Department of Health and Human
Services (DHHS) Health Resources and Services
Administration (HRSA) Maternal and Child Health
Bureau (MCHB) Director, Office of Data and
Program Development Gopal K Singh, Ph.D.
(HRSA/MCHB) Timothy R.B. Johnson, M.D. (U of
Michigan) Jennifer Z Huang, M.D., Ph.D.
(Georgetown U) Greg R Alexander, Sc.D. (U of
South Florida) David C Goodman, M.D. (Dartmouth
U)

2
Introduction

• Despite national programs to reduce preterm
  birth (PTB) and low birthweight (LBW), both of
  these pregnancy outcomes have steadily increased
  since the early 1980s, and disparities by
  race/ethnicity remain.

3
Percent Preterm Births in the US, 1990-2006
4
Percent Low Birth Weight Births in the US,
1990-2006
5

• Concomitant with the rise in these rates has
  been an increase in obstetric interventions at
  the time of delivery

6
Rate of Induction of Labor in the US, 1990-2005
7
Rate of Cesarean Delivery in the US, 1990-2006
8
Number of Cell Phones Purchased, 1996-2006
9
Number of Plastic Surgeries, 1996-2006
10
Number of Web Hits for Paris Hilton, 1996-2006
11
Number of Web Hits for Lindsey Lohan, 1996-2006
12
Number of Web Hits for Britney Spears, 1996-2006
13

• There has also been an increase in the number
  of women receiving more intensive prenatal care
  utilization.

14

• Kramer suggested that the increase in preterm
  birth rates may be due to
• An increase in multiple births
• An increased use of ultrasound
• Changes in the limits of fetal viability
• More aggressive management of high-risk
  pregnancies

15
Preterm Birth Rate per 100 Live Twin Births by
Level of PNC Utilization, 1981-1997
16
Intensive PNC Utilization and Twin Preterm Birth,
by Length of Gestation, 1981 and 1997
17
Percentages of Twin Births Delivered Preterm by
First Cesarean Section by Level of PNC
Utilization, 1989-91 to 1995-97
18

• What if the rates of preterm birth and low
  birth weight are also being influenced by the
  distribution of providers?

19

• Low birth weight has been shown to vary by
  region. (Thompson, Goodman, et al. Pediatrics,
  2005)
• There has been a strong correlation between
  within-state declines in fertility and
  within-state increases in cesarean sections,
  implying induced demand. (Gruber, Owings. RAND
  J of Economics, 1996)
• Neonatal intensive care capacity is not located
  in regions of greater newborn need as measured by
  low birth weight rates. (Goodman, Fisher, Little,
  et al. Pediatrics, 2001)

20
Research Question

• After controlling for social, behavioral, and
  obstetric factors, is there an association
  between the distribution of obstetrician/gynecolog
  ists and birth outcomes in 1992 and 2002?

21
Data Sources

• The 1992 and the 2002 National Natality Files
  with county-level data from the National Center
  for Health Statistics
• The Area Resource Files from the Health Services
  and Resources Administration which provides
  county-level data for the number of obstetricians
  per 1000 women ages 15-44 in 1990 and 2001.

22
Dependent Variables

• Preterm birth (
• Very preterm birth (
• Moderate preterm birth (32-34 weeks)
• Mild preterm birth (35-36 weeks)
• Low birth weight (
• Very low birth weight (
• Small for gestational age (uses national
  reference of Alexander et al)

23
Independent Variable and Covariates

• Number of obstetricians/gynecologists by Newborn
  Service Area (N1601) in 1990 and 2001
• Parity
• Maternal education
• Race/ethnicity
• Nativity
• Marital status
• Cigarette and alcohol use
• Amount of weight gained
• Medical risk
• Had labor induced
• Had primary or repeat cesarean section
• Median income by county
• Trimester of prenatal care

24
Analysis

• The distribution of ob/gyns was examined as both
  a continuous variable and divided into quartiles
  in relation to the outcome variables.
• Stratified analyses indicated disparities by
  race/ethnicity.
• Multilevel analyses were conducted using both
  logistic and multilevel modeling methods.

25
(No Transcript)
26
(No Transcript)
27
Birth Outcomes by Number of Obstetrician/Gynecolog
ists per Number of Women 15-44 in Neonatal
Service Areas in the US, 1992
28
Birth Outcomes by Number of Obstetrician/Gynecolog
ists per Number of Women 15-44 in Neonatal
Service Areas in the US, 2002
29
Unadjusted Odds Ratios for Birth Outcomes by
Number of Obstetrician/Gynecologists, 1992
30
Unadjusted Odds Ratios for Birth Outcomes by
Number of Obstetrician/Gynecologists, 2002
31
Adjusted Multilevel Models for Birth Outcomes by
Number of Obstetrician/Gynecologists, 1992
32
Adjusted Multilevel Models for Birth Outcomes by
Number of Obstetrician/Gynecologists, 2002
33
Adjusted Multilevel Models for Birth Outcomes by
Number of Obstetrician/Gynecologists, Non-Whites
Only, 1992
34
Adjusted Multilevel Models for Birth Outcomes by
Number of Obstetrician/Gynecologists, Non-Whites
Only, 2002
35
Summary of Findings

• In both 1992 and 2002, bivariate analysis
  indicated an association between the number of
  obstetrician/gynecologists and birth outcomes,
  with more obstetricians being associated with
  higher levels of the study outcomes, except for
  preterm birth in 2002.
• The multilevel models, adjusting for
  sociodemographic, behavioral, and medical risk
  factors, as well as obstetric procedures
  attentuated the associations.

36
Summary of Findings (cont)

• It appears that the findings differ when
  stratified by race/ethnicity.
• While the overall associations remained similar
  for all women and non-white women in 1992, the
  associations were stronger for non-white women in
  2002 compared to the overall population.

37
Conclusion

• The findings from this study suggest that the
  increasing rates of some birth outcomes may be
  influenced in some populations by the
  distribution of obstetrican/gynecologists.

38
(No Transcript)
39
Contact Information

• Michael Kogan, Ph.D.
• HRSA/MCHB
• Director, Office of Data and Program Development
• 5600 Fishers Lane, Room 18-41
• Rockville, MD 20857
• 301-443-3145
• mkogan_at_hrsa.gov