Outreach and Outcome: Outreach and Outcome: Does the Boston Healthy Start Initiative’s Does the Boston Healthy Start Initiative’s
Interconception Care Model Make a Interconception Care Model Make a Difference?Difference?
Lois McCloskey DrPH, Penny Liu ScD MPH,Lois McCloskey DrPH, Penny Liu ScD MPH,
Xandra Negron BA, Rosie Munoz-Lopez MPH, Xandra Negron BA, Rosie Munoz-Lopez MPH,
Snehal Shah MD, MPH, Barbara Ferrer PhD, MPH, MEdSnehal Shah MD, MPH, Barbara Ferrer PhD, MPH, MEd
BackgroundBackground
Boston Infant Mortality Boston Infant Mortality
Boston Healthy Start InitiativeBoston Healthy Start Initiative
Infant Mortality in BostonInfant Mortality in Boston
Black infants 3-4 times more likely than white infants to die Black infants 3-4 times more likely than white infants to die in first year of life and to be born at low weightin first year of life and to be born at low weight Black IMR 13.2: White IMR 3.6 (2006)Black IMR 13.2: White IMR 3.6 (2006)
Highest proportion of black infant deaths and highest rates in Highest proportion of black infant deaths and highest rates in Project Area in 1990’sProject Area in 1990’s Black IMR in Project Area: 11.5 (2002-06) Black IMR in Project Area: 11.5 (2002-06)
Despite slight improvements in black IMR, disparity WIDENSDespite slight improvements in black IMR, disparity WIDENS
Boston Healthy Start Initiative Boston Healthy Start Initiative (BHSI)(BHSI)
Boston an original HS site from 1992Boston an original HS site from 1992 Housed within Boston Public Health CommissionHoused within Boston Public Health Commission Emphasizes elimination of racial disparity and serves Emphasizes elimination of racial disparity and serves
only black women (self identified) in project areaonly black women (self identified) in project area Includes African American, Haitian, Dominican womenIncludes African American, Haitian, Dominican women Focus on most vulnerable, e.g. homeless, substance usersFocus on most vulnerable, e.g. homeless, substance users
Funds network of 8 health centers and 5 community-Funds network of 8 health centers and 5 community-based agenciesbased agencies
Each provides outreach and case management services Each provides outreach and case management services to women and their infants from pregnancy until child’s to women and their infants from pregnancy until child’s 22ndnd birthday. birthday.
BHSI Model: Outreach and BHSI Model: Outreach and RecruitmentRecruitment
Community Consortium (450 members) Community Consortium (450 members) meets monthly—word of mouthmeets monthly—word of mouth
Partnerships with city-wide programsPartnerships with city-wide programs ““Healthy Baby/Child” (women’s circles)Healthy Baby/Child” (women’s circles) ““Father Friendly”Father Friendly”
1-800 #1-800 #
BHSI ModelBHSI Model
PhilosophyPhilosophy Strength and risk-based careStrength and risk-based care Sustain health in interconception period through:Sustain health in interconception period through:
• Connectivity to health care of all kindsConnectivity to health care of all kinds• ““Surround services” to reduce social isolation and support wellnessSurround services” to reduce social isolation and support wellness
All of this will contribute to improved birth outcomes, esp. among All of this will contribute to improved birth outcomes, esp. among black women at highest riskblack women at highest risk
Assessment Assessment Guided by Guided by Women’s Health QuestionnaireWomen’s Health Questionnaire
• holistic assessment for self and providerholistic assessment for self and provider Risks and strengths: Risks and strengths: depression (Beck’s), smoking, alcohol and depression (Beck’s), smoking, alcohol and
substance use, domestic violence, housing/homelessness, weight, substance use, domestic violence, housing/homelessness, weight, family and social supports and relationsfamily and social supports and relations
BHSI Model: Interconception CareBHSI Model: Interconception Care
Case management Case management assuresassures:: Use of prenatal and postpartum careUse of prenatal and postpartum care Medical home: connection to primary careMedical home: connection to primary care Oral healthOral health Mental healthMental health Family planningFamily planning
Case manager Case manager providesprovides:: Health educationHealth education AdvocacyAdvocacy Social supports and referrals (e.g. housing,domestic Social supports and referrals (e.g. housing,domestic
violence, nutrition)violence, nutrition)
BHSI Model: Case ManagementBHSI Model: Case Management
Structural CharacteristicsStructural Characteristics (M..Issel et al, forthcoming)(M..Issel et al, forthcoming)
Staff mixStaff mix: RNs, SWers, and paraprofessional case : RNs, SWers, and paraprofessional case managers. At each site:managers. At each site:
• .5 RN .5 RN • .2 SW.2 SW• 2 case managers (average case load=20)2 case managers (average case load=20)
Source and timing of referral inSource and timing of referral in: during pregnancy : during pregnancy but how and when varies by sitebut how and when varies by site
Integration with PNCIntegration with PNC: yes in clinical sites, no in : yes in clinical sites, no in community agenciescommunity agencies
SettingSetting: 2 home visits per quarter plus clinic and : 2 home visits per quarter plus clinic and community visitscommunity visits
Our StudyOur StudyDesign and Data SourcesDesign and Data Sources
Analytic MethodsAnalytic Methods
Findings Findings
Evaluation QuestionsEvaluation Questions
Does BHSI improve the likelihood that Does BHSI improve the likelihood that clients receive early and adequate clients receive early and adequate prenatal care and give birth to healthy prenatal care and give birth to healthy babies?babies? How do BHSI clients compare to their How do BHSI clients compare to their
counterparts in the Project Area with respect counterparts in the Project Area with respect to prenatal care use and birth outcomes?to prenatal care use and birth outcomes?
What would you expect to find?What would you expect to find?
Study Design and Data SourcesStudy Design and Data Sources
Retrospective cohort studyRetrospective cohort study First known study to link HSI data to vital recordsFirst known study to link HSI data to vital records Data SourcesData Sources
BHSI program data 2001-2005 (intake only)BHSI program data 2001-2005 (intake only) Massachusetts vitals records 2001-2005Massachusetts vitals records 2001-2005
• Birth fileBirth file• Linked birth-death fileLinked birth-death file• Fetal death fileFetal death file
Comparison group: live births or fetal deaths to black Comparison group: live births or fetal deaths to black women who lived in the BHSI project area and gave birth women who lived in the BHSI project area and gave birth 2001-2005 2001-2005
Data Linking ProcessData Linking Process(Deterministic Matching)(Deterministic Matching)
1. BHSI intake data 2. Vital records in BHSI project area
< 30% match
1. Participant lists from 14 BHSI sites 2. Vital records in BHSI project area
< 45% match
1. Participant lists from 14 BHSI sites 2. Vital records for the Commonwealth
of Massachusetts
93% match
Matching variables: 1. Mother’s last name 2. Mother’ DOB 3. Date of delivery
Study PopulationStudy Population
N=1,443 BHSI client IDs
N=1,336 Matched to MA vitals
records
N=1,282 BHSI clients
(1,277 infants + 5 fetal deaths)
N=107 Miscarriages or moved
out of MA
N=54 Duplicate records
N=9,075 Non-BHSI black infants/fetuses in
BHSI project area (9,021 infants + 54 fetal deaths)
N= 962 BHSI project area
N= 320 Non-BHSI project area
VariablesVariables Exposure:Exposure: BHSI participation statusBHSI participation status
Socio-demographic characteristicsSocio-demographic characteristics: : mother’s mother’s age, country of origin, marital status, educational age, country of origin, marital status, educational attainment, health insurance for prenatal care, plurality, attainment, health insurance for prenatal care, plurality, parity, gravidityparity, gravidity
Clinical characteristicsClinical characteristics:: smoking status during smoking status during
pregnancy, chronic hypertension, pregnancy-induced pregnancy, chronic hypertension, pregnancy-induced hypertension, gestational diabetes, seizure disorder, hypertension, gestational diabetes, seizure disorder, previous preterm birth, small for gestational age, method previous preterm birth, small for gestational age, method of delivery, complications of delivery, baby’s sexof delivery, complications of delivery, baby’s sex
VariablesVariables OutcomesOutcomes
Use of prenatal careUse of prenatal care• Timing: <= 4 months gestationTiming: <= 4 months gestation• Adequacy: Inadequate or Intermediate Adequacy: Inadequate or Intermediate
(Kotelchuck Index)(Kotelchuck Index) Birth outcomes Birth outcomes
• Early (< 32 wks)Early (< 32 wks)• Low weight (< 1500 gms)Low weight (< 1500 gms)• Intrauterine growth restriction (BW < 10Intrauterine growth restriction (BW < 10 thth percentile at GA) percentile at GA)**• Infant or fetal death Infant or fetal death
* (Oken et al 2003)* (Oken et al 2003)
[Also available: breastfeeding at birth, maternal weight gain][Also available: breastfeeding at birth, maternal weight gain]
Analytic MethodsAnalytic Methods Bivariate analyses Bivariate analyses
Association between BHSI status and covariates/ outcomes and Association between BHSI status and covariates/ outcomes and covariates (chi-square statistics)covariates (chi-square statistics)
Association between BHSI status and outcomes Association between BHSI status and outcomes (crude RR’s and 95% CI’s)(crude RR’s and 95% CI’s)
Multivariate analysesMultivariate analyses Association between BHSI status and outcomes with adjustment Association between BHSI status and outcomes with adjustment
for significant covariates (adjusted RR’s and 95% CI’s)for significant covariates (adjusted RR’s and 95% CI’s) Poisson regression model with a robust error variance (Poisson regression model with a robust error variance (McNutt McNutt
et al, 2003)et al, 2003)
Stratified analysis to explore potential confounding and Stratified analysis to explore potential confounding and interactionsinteractions
ResultsResults
Social and clinical riskSocial and clinical riskBHSI v. non-BHSIBHSI v. non-BHSI
BHSI infants were BHSI infants were significantlysignificantly** more more likely likely to be born to mothers to be born to mothers who were:who were:
Teenagers (26% vs. 16%)Teenagers (26% vs. 16%) Foreign born (46% vs. 39%)Foreign born (46% vs. 39%) Unmarried (80% vs. 66%)Unmarried (80% vs. 66%) < High school educated (31% vs. 15%)< High school educated (31% vs. 15%) Publicly insured (84% vs. 63%)Publicly insured (84% vs. 63%) Nulliparous (47% vs. 41%)Nulliparous (47% vs. 41%) Smokers during pregnancy (8% vs. 6%)Smokers during pregnancy (8% vs. 6%)
BHSI infants were BHSI infants were significantly* lesssignificantly* less likely likely to be born to mothers: to be born to mothers: With one or more clinical risk factors for this pregnancy With one or more clinical risk factors for this pregnancy (9% vs. 13%)(9% vs. 13%)
* P < 0.05* P < 0.05
Entry of prenatal care Entry of prenatal care ≤ 4th month of gestation≤ 4th month of gestation
87.885.6
0
20
40
60
80
100
BHSI Non-BHSI
Crude RR= 0.99, 95% CI= 0.96-1.01
%
N= 1,236 N= 8,659
Entry of prenatal care Entry of prenatal care ≤ 4th month of gestation≤ 4th month of gestation (predicted values)(predicted values)
86.389.8
0
20
40
60
80
100
BHSI Non-BHSI
Adjusted RR= 1.04, 95% CI= 1.01-1.07
%
Entry of prenatal care Entry of prenatal care ≤ 4th month of gestation≤ 4th month of gestation Health insurance modifies effect of program on prenatal care entry*Health insurance modifies effect of program on prenatal care entry*
8591
82
95
0
20
40
60
80
100
Public or No Ins. Private Ins.
BHSI Non-BHSI
P < 0.05P < 0.05
BHSINon-BHSI
* Health insurance not associated with prenatal care adequacy in this population
Intermediate or inadequate prenatal Intermediate or inadequate prenatal care usecare use**
22.125.1
0
5
10
15
20
25
30
BHSI Non-BHSI
Crude RR= 1.13, 95% CI= 1.02-1.26
•Based on Kotelchuck Index (Kotelchuck, 1994)
%
N= 1,217 N= 8,569
Inadequate or intermediate adequate PNC Inadequate or intermediate adequate PNC (predicted values)(predicted values)
23.3
20.1
0
5
10
15
20
25
30
BHSI Non-BHSI
Adjusted RR= 0.86, 95% CI= 0.78-0.96
%
Inadequate or intermediate adequate PNCInadequate or intermediate adequate PNC Health insurance modifies effect of program on prenatal care adequacy*Health insurance modifies effect of program on prenatal care adequacy*
* Health insurance not associated with prenatal care adequacy in this population
26
17
30
9
0
5
10
15
20
25
30
Public or No Ins. Private Ins.
BHSI
Non-BHSI
P < 0.05
P < 0.05
BHSI
Non-BHSI
Length of gestation < 32 weeksLength of gestation < 32 weeks
3.5
1.6
0
1
2
3
4
5
BHSI Non-BHSI
Adjusted RR= 0.47, 95% CI= 0.29-0.76
[Crude RR= 0.41, 95% CI= 0.25-0.65]
%
N= 1,255 N= 8,869
Birth weight < 1500 gramsBirth weight < 1500 grams
3.22.3
0
1
2
3
4
5
BHSI Non-BHSI
Adjusted RR= 0.71, 95% CI= 0.47-1.07
[Crude RR= 0.62, 95% CI= 0.42-0.94]
%
N= 1,255 N= 8,852
Intrauterine growth restriction (IUGR)Intrauterine growth restriction (IUGR)*
10.510.0
0
5
10
15
BHSI Non-BHSI
Adjusted RR= 0.95, 95% CI= 0.79-1.13
[Crude RR= 1.01, 95% CI= 0.85-1.20]
%
N= 1,255 N= 8,852
* Having a birth weight <10th percentile at a given gestational age for non-Hispanic black infants (Oken et al., 2003)
LBW, preterm birth, IUGR, or LBW, preterm birth, IUGR, or fetal/infant deathfetal/infant death
17.716.1
0
5
10
15
20
BHSI Non-BHSI
Adjusted RR= 0.91, 95% CI= 0.80-1.03
[Crude RR= 0.93, 95% CI= 0.82-1.07]
%
N= 1,255 N= 8,872
Summary of FindingsSummary of Findings
Recruitment: Mixed bagRecruitment: Mixed bag Low penetration in project area (10-14%)Low penetration in project area (10-14%) Successful recruitment of women at highest social riskSuccessful recruitment of women at highest social risk Lower clinical risk among BHSI clients appropriate in light of shift Lower clinical risk among BHSI clients appropriate in light of shift
to less intensive clinical case managementto less intensive clinical case management
Access to prenatal care: Excellent and more to doAccess to prenatal care: Excellent and more to do BHSI increased early entry into prenatal care and decreased BHSI increased early entry into prenatal care and decreased
less than adequate careless than adequate care Gains seen only for women with public or no insurance, not for Gains seen only for women with public or no insurance, not for
privately insuredprivately insured
Summary of FindingsSummary of Findings
Birth Outcomes: Impact but limitedBirth Outcomes: Impact but limited BHSI participation associated with decreased risk for BHSI participation associated with decreased risk for
very preterm birth (after risk adjustment) and very low very preterm birth (after risk adjustment) and very low weight (before risk adjustment)weight (before risk adjustment)
• Could reflect lower clinical risk of BHSI clients and/orCould reflect lower clinical risk of BHSI clients and/or• Impact of the program on the birth outcome most associated Impact of the program on the birth outcome most associated
with stress related to high social and environmental riskwith stress related to high social and environmental risk • However, timing of entry into program unknownHowever, timing of entry into program unknown
BHSI participation NOT associated with decreased BHSI participation NOT associated with decreased risk for restricted growth or our composite measure of risk for restricted growth or our composite measure of “bad outcome” “bad outcome”
Our Findings In ContextOur Findings In Context
Limitations of StudyLimitations of Study Questionable validity of vital records data on risk factors, esp. Questionable validity of vital records data on risk factors, esp.
clinical risks clinical risks Unable to measure exposure to other interventions in client and Unable to measure exposure to other interventions in client and
non-client group (e.g. city-wide HBHC)non-client group (e.g. city-wide HBHC) Unable to link to BHSI program dataUnable to link to BHSI program data
• ““dose” of servicesdose” of services• content of servicescontent of services• intermediate outcomes, e.g. “medical home”, resolution of mental intermediate outcomes, e.g. “medical home”, resolution of mental
health and social riskhealth and social risk
No information on longer term effects to reflect interconception No information on longer term effects to reflect interconception
care model, e.g. subsequent pregnancies and women’s health care model, e.g. subsequent pregnancies and women’s health over timeover time
Our Findings in ContextOur Findings in Context
Our findings consistent with prior studies on pregnancy-Our findings consistent with prior studies on pregnancy-related case management related case management (M. Issel et al forthcoming)(M. Issel et al forthcoming) Most show significant positive program effect on Most show significant positive program effect on
prenatal care useprenatal care use Some show positive effect on birth weightSome show positive effect on birth weight
• Clearest evidence for Clearest evidence for nursenurse case management case management (Olds D et al 1986, 1988, 1993, 1997)(Olds D et al 1986, 1988, 1993, 1997)
Newest evidence favors cognitive behavioral models Newest evidence favors cognitive behavioral models of PCM to interrupt cycle of risk of PCM to interrupt cycle of risk
(El-Mohandes AAE et al 2008)(El-Mohandes AAE et al 2008)
Take Home Messages So FarTake Home Messages So Far
ProgramProgram You ARE making a difference…..ANDYou ARE making a difference…..AND
Intensify OUTREACH to recruit more women in project areaIntensify OUTREACH to recruit more women in project area• Dig deep into neighborhoods, housing developmentsDig deep into neighborhoods, housing developments
Restore capacity for focus on CLINICAL high risk to maximize ability Restore capacity for focus on CLINICAL high risk to maximize ability to intervene in social stress to intervene in social stress and and clinical pathwayclinical pathway
• Add RN capacity or partner closely with other programs with clinical focusAdd RN capacity or partner closely with other programs with clinical focus
Apply success for PNC to PP care and “medical home”Apply success for PNC to PP care and “medical home” Special initiative to engage women with non-traditional risksSpecial initiative to engage women with non-traditional risks
• Working women, women with private health insurance!Working women, women with private health insurance!
Next Steps for BHSI EvaluationNext Steps for BHSI Evaluation
““Efforts to Outcomes”Efforts to Outcomes” Simplified and systematic real time data system to Simplified and systematic real time data system to
track program activities and outcomestrack program activities and outcomes Will allow us to analyze social risks, referrals and Will allow us to analyze social risks, referrals and
resolutions (domestic violence, housing, nutrition, resolutions (domestic violence, housing, nutrition, family planning)family planning)
Will allow us to track women’s connection to a Will allow us to track women’s connection to a medical home, prenatal care, postpartum care, mental medical home, prenatal care, postpartum care, mental and oral health careand oral health care
BHSI data and PELL data linkedBHSI data and PELL data linked Will allow us to follow women and children (services and Will allow us to follow women and children (services and
outcomes) over time and across pregnancies and birthsoutcomes) over time and across pregnancies and births
ConclusionsConclusions
BHSI is a critical part of Boston’s WHOLE BHSI is a critical part of Boston’s WHOLE strategy to be an MCH organization that works strategy to be an MCH organization that works upstream and downstream for women and upstream and downstream for women and families--families-- Life course perspectiveLife course perspective Systems approachSystems approach Policy changePolicy change
By digging deeper and partnering wider BHSI By digging deeper and partnering wider BHSI can make a bigger difference for Black women can make a bigger difference for Black women and infants in Bostonand infants in Boston