Date of Award


Document Type

Open Access Dissertation


Health Promotion, Education and Behavior

First Advisor

Edward A. Frongillo


Labor and birth related health outcomes remain suboptimal in the US. New initiatives and improvement efforts include peer and social support. Doula support is built on the concept of social support, including emotional support, comfort measures, information, and advocacy. Benefits of doula support include increases in women’s satisfaction with their childbirth experiences, postpartum interaction with infant, breastfeeding initiation, and APGAR scores, and reductions in cesarean deliveries, length of labor, use of analgesia, and healthcare costs to healthcare systems. Doula services based within the hospital are a relatively new phenomenon. How and why hospitals are incorporating doula services has not been studied. This dissertation identified factors and decisions involved in the adoption of hospital-based doula programs across the United States, and investigated the scope and services of U.S. hospital-based doula programs including doula training, contractual relationships of doulas with the hospital, cost to women, how doulas are connected with women, and the scope of doula commitment.

This was a qualitative study. We targeted all identifiable hospital-based doula programs in the US for inclusion in this study. Through an internet search and snowball sampling, 50 programs were identified. Of these, 40 met inclusion criteria: 1) currently operational, and 2) hospital-based. We conducted semi-structured phone interviews vi with doula program managers at 32 U.S. hospitals, for a response rate of 80%. We reviewed two additional program websites whose existence was verified, but did not participate in interviews. This information was included in Manuscript 1. All interviews were recorded via CallTrunk for iphone and transcribed for analysis. We used emergent coding techniques. Adoption of innovations and Shiffman’s social constructionist frameworks guided data collection and analysis.

The majority of hospital-based doula programs were initiated by individuals of various positions within the hospital (n=28). Only four programs were started by independent doulas that approached hospital administrators. Individuals involved with program initiation viewed doula support as beneficial in several respects: for the health of women, as a tool to attract women to the hospital, and as an important component of initiatives to improve maternity care. These actors also viewed doula support as a mechanism to advance outcomes for women (e.g., satisfaction, reduction of interventions, improved outcomes for vulnerable groups of women) and influence practices (e.g., introduce doulas into medical model of care, ensure equitable access to doula support, provide options) within the institution.

Doula support was instituted within hospitals as either a stand-alone program, that complemented general service, or introduced as a component of a larger initiative within maternity care. It was continued within institutions because of its importance for business via patient satisfaction, cost savings, and marketplace edge, and through persistent advocacy from individuals within the hospital.

Variation across doula service delivery was common. Most programs required (n=14) and/or provided on-site training (n=14); six required certified doulas. Most programs were offered free of charge to women (n=27), although others cost as much as $750. We identified three contractual relationships of doulas with the hospital: volunteer, staff, and contract. Doula assignment models included prenatal assignment (n=7) and on-call shifts (n=22); some offered both (n=6). The scope of doula commitment across the labor and birth trajectory varied; most programs encouraged doulas to stay with the woman until the baby was born.

The ideas through which program adopters understood and portrayed the importance of doulas were important for garnering support at program initiation. Perspectives of the individuals involved with program initiation and development sometimes differed from the larger institutions’ understanding and rationale for formal program adoption and continued support of the program. Doula service delivery varied across hospital settings. Among the hospitals that participated in the study, the majority were volunteer programs. The average program was 10 years old. It remains to be known whether and how the variations in doula service delivery across programs influence outcomes. Whether and how these variations influence labor and birth outcomes require further research.