The Effect of a Community Health Worker Home Visiting Randomized Trial on the Health of Newborn Chil

1
The Effect of a Community Health Worker Home
Visiting Randomized Trial on the Health of
Newborn Children
Cristian Meghea, PhD Lee Anne Roman,
PhD (Cristian.Meghea_at_hc.msu.edu) Institute for
Health Care Studies Michigan State University,
East Lansing, MI, USA iHEA 6th World Congress on
Health Economics Copenhagen, Denmark, July 2007
Pre-publication information. Please do not cite
2
1. Introduction

• Good infant health - key predictor of many
  subsequent outcomes
• Children in low-income families are in worse
  health and lag behind other children in many
  other areas
• Challenging to provide services that improve
  child health for vulnerable families
• Contain health care costs

3
2. Introduction

• Home visiting programs have been used to engage
  low-income families in health services
• Home visitors are usually nurses, contributing
  their in-depth health knowledge
• Community health workers (CHW) also provide home
  visiting
• Similar characteristics to those they serve
• Bridge cultural differences and promote health
  among groups with barriers to care
• Better able to relate to, and act as role models
  
• Nurse-CHW health community knowledge

4
3. This study

• Will test whether a Nurse-CHW home visiting
  program for low-income pregnant women
• Improved infant health and
• Contained health expenditures
• (compared to a nurse home-visiting program)

5
4. Rationale

• Some nurse home visiting programs had limited
  success
• Many nurse visiting programs stop their efforts
  if women are repeatedly not at home, break
  appointments, or refuse services
• Nurse-CHW teams, building on the strengths of
  both nurses and CHWs, may be more successful
  than nurse home visiting
• The health expenditures are increasing

6
5. The team home visits

• The nurse-CHW team home visiting program was
  implemented in a population of pregnant
  low-income women (Medicaid)
• A longitudinal randomized trial pilot
  intervention (nurse-CHW-team vs. nurse)
• Enrolled lt24 weeks, 530 pregnant women,
  followed-up until the infant was one year of age

7
6. Data

• Infant health outcomes at 6 months and at 12
  months of age (mother-assessed child health
  health conditions ER and hospital visits)
• Info on the mother also at the enrollment
  interview (lt24 weeks), and at 34-36 weeks
• Medical claims (health expenditures)
• Nurse-CHW visits intervention group Nurse
  visits control group

8
7. Methods

• Compare the child health outcomes and child
  related health expenditures in the intervention
  and control groups
• Treatment-control multiple regression analysis,
  accounting for child and mother characteristics
• Outcomes health conditions, immunization, ER,
  hospital stays, health expenditures

9
8. Similar treat/control groups
10
9. Infant health improved?
11
10. Infant health improved?

• At least 1 health problem virtually identical
  percentages in treatment and control groups
• Asthma smaller incidence in treatment group
• Immunization up-to-date no difference
• ER visits difference not significant
• with hospital stays no difference

12
11. Lower health expenditures?

• Treatment group savings of 750 in child
  related health expenditures
• Mainly (500) in inpatient hospitalizations
• Less severe hospitalizations?
• Even larger savings for low psychosocial
  resources (1220 baby related expenditures)
• Longer-term effects?

13
12. Long term effects

• Longer-term effects?
• The prevented cases of asthma (possibly 1/3)
• Not having to live with asthma as child and adult
• In 1996, the total economic impact was 2 bil.
• Direct medical expenses
• Parents loss of productivity (school absences)
• Even preventing less than 1/3, substantial
  improvement

14
13. Next steps

• Early for conclusions
• Use the longitudinal nature of data
• Control for mother and infant characteristics
• Use medical claims to check the mother-reported
  measures on infant health