Transcript No Slide Title

Ethnic Disparities in Perinatal
Outcomes in the U.S.
Vijaya K. Hogan, Dr.P.H.
Pregnancy and Infant Health Branch
Division of Reproductive Health
Centers for Disease Control and Prevention
Factors contributing to race/ethnic
health disparities
• Higher exposure to risk
– may be imposed by environmental and social policies
• Higher vulnerability/susceptibility
– may be determined by chronic environmental and social
exposures
• Insufficient resources to protect health
– lack of time, money, etc…for health promotion, health care
access
• Unequal access to care
– “snowball” exposure effect
• Lower quality of care
Outline
•
•
Extent of race/ethnic
disparities in perinatal health
Example of complexity of
prevention: Preterm Delivery
•
Limits of scientific knowledge
•
•
•
Practice limitations
Social context
Proposed approach to
addressing health disparity
Part I
Data: Disparities in Perinatal Outcomes
Maternal Mortality Ratios by Race
United States, 1987-1996
25
19.6
20
15
10
7.7
5.3
5
0
Total US
Danel, et al, 1999
Black
White
Infant Mortality Rates by Maternal Race and
Ethnicity, 1997 U.S. Death Cohort
1600
1400
1368.5
1200
1000
Deaths per
100,000 live 800
births
600
868.5
604.6
595.4
497.7
400
200
0
Black
Am/Ind
White
Hispanic
Race and Ethnicity of Mother
Asian/Pas
Ten Leading Causes of Infant Mortality
United States, 1997
Rate per 100,000 live births
Birth Defects
155.7
95.7
Preterm/LBW
69.4
SIDS
RDS
32.4
Maternal Preg.Comp.
31.9
23.8
Placenta, Cord Comp.
Infections
19.4
Accidents
19.3
11.7
Hypoxia/Birth Asphyxia
10.2
Pneumonia/Influenza
0
20
40
60
80
100
120
140
160
180
Leading Cause-Specific Infant Mortality Rates
By Maternal Race, United States, 1997
Rate per 100,000 live births
300
269.4
Total
White
Black
250
200
182.1
155.7 153.6
150
137.5
95.7
100
65.2
69.4
69.4
57.9
50
32.4
26.1
69.9
31.9
25.9
0
Birth Defects
Preterm/LBW
SIDS
RDS
Maternal Preg.
Comp.
Distribution of Deaths Due to Top Five
Causes of Infant Mortality, 1995
s
t
n
a
f
n
I
k
c
a
l
B
%
0
.
5
%
9
.
9
1
%
4
.
5
%
2
.
3
1
%
8
.
1
1
%
7
.
4
4
s
t
n
a
f
n
I
e
t
i
h
W
%
2
.
4
%
5
.
0
1
% 2
7
.
4
%
1
.
6
%
3
.
1
1
%
2
.
3
4
S
D
R
.
f
e
D
.
h
t
B
r
e
h
t
WO
B
L
/
T
P
S
D
I
S
l
a
n
r
e
t
a
M
.
s
e
p
a
t
a
t
a
d
e
s
u
c
i
l
b
u
p
s
c
i
t
s
i
t
a
t
s
l
a
t
i
v
S
H
C
N
:
e
c
r
u
o
S
All
White
Black
1000
AAP
Campaign
100
Year of Death
*Preliminary Data
00
96
98
90
80
10
70
Deaths per 100,000 live births
Infant Mortality Rates Due to SIDS, United
States by race, 1973-1998*
Infant Deaths due to NTDs
by Race/Ethnicity, United States, 1996
Rate per 100,000 live births
20
18
16
14
12
10
8
6
4
2
0
17.9
16.3
9.9
16.8
16.8
Cuban
Central /
So.
American
9.5
7.9
Black
nonHispanic
White
nonHispanic
Hispanic
Mexican
Source: National Center for Health Statistics, 1996 period linked birth/infant death file
Prepared by March of Dimes Perinatal Data Center, 1999
Puerto
Rican
LBW Among Singletons by Race
US, 1991-1997
Percent
14
12.1
11.8
12
11.8
11.7
11.6
11.4
11.4
10
8
6
4.7
4.7
4.8
4.9
5.0
5.0
5.0
4
2
0
1991
1992
1993
1994
White
1995
Black
Source: National Center for Health Statistics, 1996 period linked birth/infant death file
Prepared by March of Dimes Perinatal Data Center, 1999
1996
1997
Rate per 1,000 live births
Trends in overall* PTD by race-United States, 1989-1996
210
190
170
B lack
150
A ll races
130
 4%
110
90
70
0
19 89
19 90
19 91
19 92
19 93
Year
*Includes multiples and singletons
19 94
1 995
19 96
W h i te
Rate* of Singleton PTD by Maternal
Race/Ethnicity, United States, 1989-1997
Rate per 1000 Live Births
190
170
150
130
110
90
70
50
1989 1990 1991 1992 1993 1994 1995 1996 1997
Black
Nat Am
Hispanic
API
White
Part II:
Complexity of Race/ethnic disparities:
Focus on Preterm Delivery (PTD)
• Affects many infants
– 11% of live births (400,000 infants/year)
• Mortality:
- # 1 cause of infant death among blacks
- # 2 cause of infant death overall
- # 1 contributor to infant mortality disparity
• Morbidity:
– lung disease, vision and hearing impairment,
developmental delays, cerebral palsy
Preterm Delivery
• Good illustration of multiple dimensions to
causal pathway leading to disease and disparity
• If we can successfully address this health
outcome, we will likely have the keys to
addressing all disparities
Limitations of Science
• To eliminate disparities in
preterm delivery, we would need:
– Evidence-based interventions to
prevent preterm delivery
– Evidence-based interventions/activities
to eliminate disparities
Process for Defining
Evidence-Based Strategies
6 Guidelines: Evidence- based Strategies
5 Review and synthesis of existing published
evidence
4 Evaluation studies published in peer-reviewed
literature
3 Evaluation Studies conducted on existing
programs
2 Existing programs refine implementation
1 Programs implemented using conceptually valid
strategies
0 Many potential approaches:
Scientific guidance needed at this stage
to define conceptually valid strategies
Preterm Delivery: Epidemiologic
Risk Factors
•
•
•
•
•
Race/Ethnicity
Incompetent cervix
Marital Status
Substance use
Number of prenatal
care visits
• Smoking
• Multiple Births
• Mother's PrePregnancy
Weight
• Previous Obstetric
History
• Bacterial vaginosis
• Congenital Anomalies
• Stress
• Maternal Age
Patient Outcomes Research Team
(PORT) Study
on Infant Mortality
Goldenberg, et al, 1998
• None of the main OB or behavioral modification
strategies are effective in reducing PTD
• No additional benefit from: Increasing quality
and quantity of PNC, maternal weight gain and
nutritional supplements
• Ineffective against PTD: Bed rest, risk scoring
systems, iron supplementation, tocolytics,
substance use programs, IV hydration, HUAM
Bacterial Vaginosis (BV)
• Lower tract marker of upper reproductive
tract infection
• Can be assessed during pregnancy
Bacterial Vaginosis (BV)
Syndrome resulting from alteration in normal
vaginal flora
 Reduction in hydrogen peroxide-producing
lactobacilli
 Reduction in natural protection against overgrowth
of more harmful bacteria

–

mycoplasmas, gardnerella vaginalis, bacteroides
Microbe concentration increases to level of stool
Bacterial Vaginosis, cont.
• BV is 2 x more prevalent among African
American women
• Higher prevalence not explained by sexual
behaviors or most known risk factors.
• High BV rates in African-American women may
account for up to 30% of excess risk of PTD
Prevalence of BV among Pregnant Black and White
Women by Diagnostic Method
Gram stain
MD Assessment
White
95%CI
27%
18.7%
(19-35)
(11-26)
Black
95%CI
64.4%
40.5 %
(61-69)
(36-45)
Hogan et al, SPEAC Study, Philadelphia
Treatment of BV
Studies demonstrate a reduction of PTD with tx.
• Morales WJ et al. Am J Obstet Gynecol
1994;171:345-349.
• Hauth JC et al. N Engl J Med 1995;333:1732-1736.
• McDonald HM et al. Br J Obstet Gynaecol
1997;104:1391-1397.
Treatment Guidelines
Treatment for BV during Pregnancy:
CDC (November 1997):
screening high risk women in early 2nd trimester
250 mg metronidazole t.i.d. x 7 days

ACOG (February, 1998)
screening high risk women in 2nd trimester
500 mg oral metronidazole b.i.d. x 7 days

Limits of Medical Practice in Reducing
PTD
SPEAC STUDY- Philadelphia Public Health Clinics
• There were 389 true cases of BV
• Of these true cases, 46 were high risk
• Thus, 46 women should have been treated (using
gram stain diagnosis, CDC Rec)
– Actual treated= 24 ; 48% true high risk cases
missed
Limits of Practice.cont
• Clinicians treated 67% of high risk, BV
positive women id’d via clinical assessment
Scenario 1: improved treatment coverage
• If they treated 100% high risk women id’d by their
assessment: 31 women would have been treated (Actual =
24; 7 women missed)
Scenario 2: improved diagnostics
– If they used gram stain to assess BV, and trt’d
67%: 36 women would be treated; actual = 24;
12 women missed
Potential Impact on Preterm Birth
Morales 1994: PTD rate among :
– treated =18%
– untreated=39%
If all 46 high risk women were treated
Expected PTD rate= 18%;
Only 24/46 were trt’d:
{.18 * 24) + (.39 * 22) /46} 100 = 28% PTB rate
(8.2 vs 12.9 PTB’s)
There would be a 36 % lower PTB rate among
high risk women if all were treated.
Stress: Physiologic Effects
• Stress harms health:
– Seeman TE et al “Price of Adaptation: allostatic
load and its consequences” Arch Int Med 157:22592268;1997
– Biondi M et al “ Psychological Stress,
neuroimmunomodulation and susceptibility to
infectious diseases in animals and man: a Review”
Psychotherapy and Psychsomatics 66:3-26; 1997
Stress and Pregnancy
• Stress has negative physiologic effects
• Stress is associated with Preterm Delivery
How Does Stress Affect Health?
– Stress can affect:
• Endocrine system (corticotropin-releasing
hormone (CRH) production
• Immune system response
• Maternal Behaviors
– smoking
– nutrition
– substance use
Stress and the Infection Interaction
Chronic stress can suppress immune response
Acute stress can increase immune response
(slows or halts shut-off of immune response
leading to over -production of cytokines)
(auto-immune response leading to PTD?)
Shulkin, McEwen , Gold; 1993
Chrousos, Gold; 1992
McEwen; 1998
Stress and Vulnerability
Pregnant women who were moderately or highly
stressed were over 2 times more likely to be BV
positive compared to women in the low stress
group
“Maternal Stress is associated with bacterial
vaginosis in human pregnancy”
– Culhane, Rauh, McCollum, Hogan, Agnew and Wadhwa; MCH
Journal 5(2) 2001
Fetal Effects of Stress
Source: New Scientist, 17 July 1999
Practice Limitations:
Timing of Intervention
Analysis of Feto-Infant Mortality, Georgia 1991-1993
ELBW
(<1000)
VLBW
(1000-1499)
IBW
(1500-2499)
NBW
(2500+)
Total
Late
Fetal
Death
(28+wks)
Early
Neonatal
Death
(<7 days)
Late
Neonatal
Death
(7-27 days)
Post
Neonatal
Death
(28+ days)
Maternal
Health
Maternal
Health
Maternal
Health
Maternal
Health
180
1189
119
153
Maternal
Health
Maternal
Health
Maternal
Health
Maternal
Health
138
118
32
63
Maternal
Care
Newborn
Care
Newborn
Care
Infant
Care
217
188
77
185
Maternal
Care
Newborn
Care
Infant
Care
Infant
Care
303
204
133
637
838
1699
361
1038
Total
1641
351
667
1277
3936
How do we determine which strategy has the greatest potential for reducing
maternal, perinatal, and infant mortality?
Age at Death
Birth
Weight
(grams)
Late
Fetal
Death
(28+wks)
ELBW
(<1000)
Early
Neonatal
Death
(<7 days)
Late
Neonatal
Death
(7-27 days)
Post
Neonatal
Death
(28+ days)
Maternal Health
VLBW
(1000-1499)
IBW
(1500-2499)
NBW
(2500+)
Maternal
Care
Newborn Care
Infant Care
Complexity of Causation:
Social Context
Characteristics of the "Ideal" Pregnant
Woman
Based on Epidemiologic Data
•
She is married
•
She planned her pregnancy and her baby is
"wanted"
•
She came early for prenatal care
•
She attends all prenatal visits, on time, asks
appropriate questions
Composite "Ideal" Woman
Based on Epidemiologic Data, continued
•
She has a level of education that allows her to
understand and comply with caregiver
•
She engages in healthy lifestyle and behaviors
(eats healthy, doesn't drink, smoke, use drugs,
etc..)
•
She was actively engaged in healthy, protective
behaviors before pregnancy (e.g. took folic acid)
The Reality:
•
Experiences of Black Women Documented
from Qualitative Research
•
•
•
•
Harlem BirthRight, NYC
Healthy African American Families, Los Angeles
Development of a Chronic Stress Scale for African
American Women, Atlanta
Influence of Social Networks on Access to Prenatal
Care, Chicago
Findings:
•
Complex web of social support sought to replace or
supplement father involvement.
•
Partner support not dependent on marriage. Other
configurations of partnership exist.
•
Women hold traditional views about marriage
•
Social factors which affect males affect options for
partnership
Findings
"I feel we are losing our black men"
87.7%
Jackson, Phillips, Hogue; Atlanta, Ga.
"You have to justify the building of these prisons so that means
you've got to have inmates. What you do is deprive them of an
education … deprive them of employment … So, if they can't get
money legitimately, they get it illegitimately..."
Mullings and Wali, Harlem BirthRight Report, 1997
Public Health Ideal:
"She planned her pregnancy and baby is "wanted"
•
Public Health Questions:
• Why are unplanned pregnancies so prevalent?
•
Implicit Assumptions ??:
• Women are irresponsible about contraception
• Women have adequate knowledge about physiology
• Available contraceptives are effective and acceptable to all
women
Findings:
•
"Unplanned" pregnancies result even with active
efforts to prevent pregnancy
•
Contraceptive failure commonly reported
"Condoms broke", "got pregnant while taking the pill"
•
Lack of availability of acceptable contraceptive
methods
Discomfort with oral contraceptives and diaphragm reported
Findings:
•
Male influences and unprotected intercourse
•
Perceptions of ability to conceive
•
•
"I thought I was infertile"
Perceptions that previous PID or other infections
affected ability to get pregnant
Public Health Ideal:
“She came early for prenatal care”
•
Public Health Questions:
• Why don't women come early for prenatal care?
•
Implicit Assumptions ??:
• Access to care defined by individual choice and availability
of health insurance only, independent of other forces
• Only formal prenatal services provide pregnancy care
The Ideal Road to
Formal Prenatal Care
Social Networks Project - Chicago
Intercourse
Vigilance
Suspicion
Confirmation
Acceptance
Decision
Access to Formal
PNC
The Road to
Formal Prenatal Care
Social Networks Project - Chicago
Intercourse
Vigilance ???
Suspicion ???
Confirmation ???
Acceptance
Decision
Access to Formal
PNC
Findings
•
Initial access to care is preceded by a complex set
of life decisions to determine subsequent course of
pregnancy
•
•
•
impact on partner, social network and employment
assessed
availability of structural supports
assessment of options
Process requires time and high emotional
investment
• Women begin "self care" as soon as pregnancy is
acknowledged
•
Public Health Ideal:
"She attends all prenatal visits on time and is compliant
with caregiver recommendations"
•
Public Health Questions:
• Why do women miss prenatal appointments?
•
Implicit Assumptions ??:
• Women are irresponsible
• Women do not appropriately value prenatal care
• PNC attendance is not dependent on social forces
Findings
•
Social forces influence attendance
•
action depends on women's level of control over these
forces
Findings
•
Ms. "K" missed appointments because her mother
demanded she babysit her younger siblings
•
Ms. "T" missed appointments because she was
juggling a job requiring constant travel
•
..An adolescent living in a tenement missed
appointments because someone stole her school
bus pass, depriving her of transportation
Findings
•
Health Care providers alienate women
•
•
•
•
•
•
Failure to meet needs of women
Demeaning statements and assumptions made
Inadequate explanations provided
Lack of time spent
No inquiry into social circumstances
Health care often perceived as a stressor
“I was cared for by two people. The first one was a midwife. She
examined me and had a lot of time for the patients. The midwife
sits and asks questions and they service you very well. Then I
was cared for by a doctor, and here is where I noticed the
change. The doctor came in and out and it was over. ‘See you
next time’ . . . “
Social Networks Project - Chicago
“Medical personnel frequently consider missing
prenatal care appointments to be an indication
of women’s lack of concern about their
pregnancies . . . A non-black obstetrician told
the ethnographer that she teaches residents
that even if a patient states that she misses her
appointments because of child care or work
difficulties, she is still irresponsible.”
Mullings and Wali, Harlem BirthRight, 1997
Findings
•
Conflicting advice between social network and
providers and between providers
Whose advice do you follow? At what cost?
Public Health Ideal:
“She engages in a healthy lifestyle which begins
before pregnancy"
•
Public Health Question:
• Why do women practice unhealthy behaviors?
• What can we do to promote healthier lifestyles?
•
Implicit Assumptions ??:
• Women have clear lifestyle choices ("good" vs. "bad")
• Women choose unhealthy lifestyles
• Individual behavior independent of social forces
Findings
•
Women are not passive victims
•
Options are limited and constrained by social
environment
•
•
•
•
•
•
housing
material resources
service availability
violence (illegal activity and law enforcement)
race/racism
Living "between a rock and a hard place"
•
Choosing "lesser of two evils"
Findings
•
College education may not confer expected
benefits to African American women in Harlem:
•
equally likely to live in public housing
•
less likely to have any job benefits
•
more likely to report discrimination as a stressor
•
more likely to report experience of most stressors
Findings
Reported stressors are pervasive
partner
• family
• work
• friends
•
children
• housing
• money
• discrimination
•
e.g., all daily activities potentially stressful
Stressors
•
Pervasive experience of multiple and simultaneous
stressors
•
Unique stressors experienced by African Americans
•
effects of race, racism, history and social order
•
Living between a "rock and a hard place" a common
experience
•
Types of stressors don't vary by educational or
income level
Summary:
State of the Science of Eliminating
Disparities in PTD
• Some factors are known to contribute to
disparity, we know how to treat, but we don’t
(BV)
• Some factors contribute to poor outcomes, but
contribution to disparity is unknown, & we do
not know how to treat (social factors, stress)
• Some factors suspected to contribute to
disparity, but no empirical evidence of degree
(stress, health care factors)
Summary, cont
• Little research quantifying extent to which
social factors make some women more
vulnerable in face of risk
• All opportunities for prevention are not utilized
effectively
Part III
Proposed Strategy to Address
Race/Ethnic Disparity in PTD
• Opportunities:
– Yr 2010 Objectives
– Institutional commitment to study and eliminate disparities
• Challenges:
–
–
–
–
limits of scientific knowledge
recognition of state of science in planning action
multiple etiologies, interacting pathways
multiple players needed to address complex web of causality
(e.g. not just medical providers)
Researchers
• Improve the state of the science
• ask the right questions
• account for complexity of problem in study
designs
• use variety of scientific approaches
• develop new methods to model complexity
• include community perspectives in design
and interpretation
Funders
• Recognize the state of the science and develop
RFA’s that support systematic development of
scientific knowledge
• Develop systems for submission and review
that support diversity, innovation, community
collaboration and alternative scientific methods
Public Health Program Planners,
Providers and Policymakers
1 Increase efforts to address social causes
2 Improve collaboration with non-medical partners:
advocates, community leaders, policymakers, law
enforcement
3 Use knowledge we have: Reduce population prevalence
of bacterial vaginosis
4 Develop effective interventions for stress
5 Implement systems for women’s preventive health care
6 Improve quality of health care
7 Intervene on multiple etiologic pathways, multiple
targets,simultaneously