Programs that Work
Healthy and Safe Children
Babies born weighing more than 5.5 pounds and improving outcomes for low birth weight babies
Age of Child
Early Childhood (0-8)
Type of Setting
Community-Based Service Provider
Health Care Provider
Type of Service
Health Care Services
Type of Outcome Addressed
Substance Use and Dependence
Evidence Level (What does this mean?)
The Healthy Start Initiative was established in 1991 by the Health Resources and Services Administration (HRSA) of the U.S. Public Health Service. The initiative's primary purpose was to reduce infant mortality by 50 percent and generally improve maternal and infant health in at-risk communities.
Originally, 15 demonstration programs were funded for a five-year period; they included 13 urban and 2 rural programs located around the country (e.g., in the Northeast, South, Midwest, and West). Congress subsequently continued the funding for those programs and other additional programs. As of fiscal year 2002, 96 Healthy Start programs were being funded. The programs are currently administered through the Division of Perinatal Systems and Women's Health, which is part of the Maternal and Child Health Bureau within HRSA.
Each grantee (i.e., participating service area) has a great deal of flexibility in designing its specific program, although several common models have emerged. Certain elements were required of the programs, including a focus on reducing infant mortality, inclusion of the local community in program planning, assessment of local needs, efforts to increase public awareness, implementation of an infant mortality review, development of a package of innovative health and social services for pregnant women and for infants, and evaluation of the initiative.
The grantees are typically city, county, or state health departments. The program is designed to serve pregnant women (particularly women at high risk of poor pregnancy outcomes) and infants, with each service area determining its specific program recipients.
Mathematica Policy Research, Inc. conducted a national evaluation of Healthy Start focusing on the 15 demonstration programs (see Devaney et al., 2000). Two comparison sites were selected for all but one of the demonstration programs -- a single comparison site was identified for the final program. The comparison sites were selected for their similarity in terms of race and ethnic composition, infant mortality rate, and trends in the infant mortality rate between 1984 and 1992 (the years leading up to implementation of Healthy Start). The number of clients served in the demonstration program sites ranged from a low of 1,000 in the Pee Dee region of South Carolina to a high of 7,000 in Cleveland, Ohio. The other sites were in Baltimore, Birmingham, Boston, Chicago, Detroit, the District of Columbia, New Orleans, New York City, Oakland, Philadelphia, Pittsburgh, cities in the Northern Plains states (including 19 American Indian tribal organizations in Iowa, Nebraska, North Dakota and South Dakota), and in several cities in Lake County in Northwest Indiana (specifically East Chicago, Gary, Hammond, and Lake Station).
The evaluation examined both implementation issues and program outcomes. To study implementation, the researchers (Devaney et al.) conducted site visits, telephone interviews, focus groups, document review, observations, and home visits, as follows:
- The site visits involved interviews with program staff, service providers, and consortia members. Researchers used semi-structured protocols to gather comparable information across the programs.
- Telephone interviews were conducted with the grantees as follow-ons to the site visits and included project directors, consortium members, and outreach and case management staff. These interviews addressed any changes in the community affecting their projects, changes they had made to their organizational structures, and what progress they had made in implementing the program. The researchers also solicited thoughts on what lessons the grantees had learned and how to sustain the program past the life of the grant.
- The focus groups were conducted with Healthy Start clients and providers and addressed a range of topics, such as the quality of the clients' lives, the health and well-being of their children, clients' experiences during pregnancy, how they learned about Healthy Start, and why they became involved in Healthy Start and about their experiences with the program.
- The researchers reviewed a range of documents, including documents on case management services received by Healthy Start clients (e.g., number of visits, type of help received, if mother's health was checked, if child's health was checked, ranking of how helpful the services were) and health education materials.
- The researchers observed activities such as consortia meetings and health education classes to assess their content and quality.
- Researchers also accompanied program staff on home visits to observe how staff members interacted with clients.
The outcomes studied included prenatal care utilization, pre-term birth rate, low- and very-low-birth-weight rates, and infant mortality rate. The researchers compared infant birth and death rates for Healthy Start project areas with matched comparison sites from 1984 to 1996. They used several measures to assess various dimensions of prenatal care. One of those instruments is called the "Kotelchuck index," which is used to rate prenatal care depending on when it is initiated, how often it occurs, and other factors. The five categories in the Kotelchuck index are: no care, inadequate care, intermediate care, adequate care, and adequate-plus care. Vital statistics data were collected for the target population in each program area and include information on maternal and paternal characteristics, the timing and extent of prenatal care, pregnancy history, basic demographic characteristics, and birth outcomes. These data helped the researchers gain a better understanding of the Healthy Start target population.
The authors (Devaney et al.) offer two caveats -- the study did not involve random assignment, and the comparison areas likely had interventions similar to those of Healthy Start; therefore, the researchers were probably detecting differences in impact between interventions, not differences in outcomes between an intervention and no intervention.
Key Evaluation Findings
Devaney et al. (2000) found the following:
- Three Healthy Start program sites had significantly lower rates of low-birth-weight babies than their comparison sites. Similarly, three Healthy Start program sites had significantly lower rates of very-low-birth-weight babies.
- The Healthy Start program was associated with a significantly lower pre-term birth rate in four program sites (with decreases ranging from 1.3 percent to 2.9 percent); the differences in the pre-term birth rate in the other (11) sites were not statistically significant.
- The Healthy Start program areas rated significantly higher on the Kotelcheck index than the comparison areas on several dimensions:
- In 8 of the 15 program sites, a higher percentage of women in Healthy Start received adequate or better prenatal care as compared with women in the comparison areas. One Healthy Start program area was higher than the comparison areas on the Kotelcheck index, but not significantly so. The other 6 program sites had less positive outcomes; the comparison sites had higher percentages of women who received adequate or better prenatal care than did those 6 sites.
- In 4 of the 15 program sites, a higher percentage of women in Healthy Start experienced adequate initiation of prenatal care-with adequate initiation defined as the first prenatal care visit occurring by the fourth month of pregnancy. In the remaining programs, the percentage of Healthy Start women experiencing adequate initiation of prenatal care was lower but not significantly lower in 6 programs sites, was the same in two program sites, and was higher but not significantly so in 3 program sites.
- In 9 of the 15 program sites, a larger percentage of women in Healthy Start had an adequate or better number of prenatal care visits than women in the comparison areas. In the other 6 sites, a higher percentage of women in the comparison group had an adequate or higher number of prenatal care visits.
- Implementation generally took longer than expected.
- Programs were generally successful in developing case management programs (case management included initial contact or outreach; intake; assessment, care planning, and referrals; and ongoing contact and tracking).
- Community involvement was considered a key goal of Healthy Start. While all of the programs did encourage community involvement, it was very difficult to secure.
City, county, or state health departments serving at-risk communities
Grants from HRSA for the demonstration programs ranged from $13.8 million (Birmingham) to $30 million (New York). Funding continued, but was reduced, following the demonstration phase.
Although grantees have flexibility in how they structure their programs, there are nine typical intervention models: community-based consortia, outreach and client recruitment, care coordination/case management, family resource centers, enhanced clinical services, risk prevention and reduction, facilitating services, training and education, and adolescent programs.
The researchers documented several characteristics of successful programs:
- Strong program organization and administration, with stable program leadership, is associated with better program implementation and improved outcomes.
- Programs that focus on service coordination, with close links to the existing clinical care system, appear to be more successful than others.
- Community involvement through the employment of community residents is associated with improved outcomes in some but not all programs.
The Healthy Start Initiative described the roles and responsibilities of staff, but not their actual titles or positions. For example, substance abuse counseling and health education can be conducted by staff or contracted out. The case manager positions can be filled by the following types of staff: lay workers (community residents without professional training, but who received on-the-job training or training from Healthy Start), social workers or similarly trained professionals, and public health nurses. The Healthy Start grant money can also be used to hire clinic staff, including obstetrics/gynecology providers, pediatricians, nurses (particularly nurse midwives and nurse practitioners), nutritionists, phlebotomists, nursing assistants, and clerks.
No set or prescribed curriculum.
Issues to Consider
This program received a "promising" rating. While the researchers used a matched comparison design across many sites over several years, the methodology had some limitations and the findings were mixed. Three program sites did experience statistically significant decreases in very-low-birth-weight babies. However, no single program consistently showed improvements across all of the outcome areas, and it is not clear which program or implementation features led to which outcomes.
The lack of understanding about which features lead to which outcomes makes it difficult to replicate the successful features of the demonstration programs. In addition, the comparison sites likely had interventions similar to Healthy Start previously in place that may affect one’s assessment of the influence of Healthy Start compared with doing nothing at all (i.e., the Healthy Start results might have been more positive).
Researchers (Devaney et al.) evaluated the original 15 demonstration programs. Since then, the number of programs has increased to 96. (For more information, see the National Healthy Start Association Web site at www.healthystartassoc.org.)
Director, Division of Perinatal Systems and Women's Health
Health Resources and Services Administration
5600 Fishers Ln., Room 11A-55
Rockville, MD 20857
Harrington, M., B. Foot, and E. Closter, Using Health Education to Reduce Infant Mortality: The Healthy Start Experience, Washington, D.C.: Mathematica Policy Research, Inc., September 1998.
Howell, E. M, B. Devaney, M. McCormick, and T. T. Raykovich, "Back to the Future: Community Involvement in the Healthy Start Program," Journal of Health Politics, Policy and Law, Vol. 23, April 1998, pp. 291-317.