Table 1.

Description of included quantitatative studies

Study	Study context/objective	Methods, sample characteristics, response rate and sample size	Study period	Choices compared
Donaldson (1998) [19]	This study was conducted in Aberdeen (Scotland), an area with an OU and an AMU in the same hospital, to assess the feasibility of the use of ‘willingness to pay’ as a measure of women’s strengths of preference for intrapartum care (OU vs. AMU).	Methods: Willingness to pay study designed to evaluate ‘low risk’ women’s preference for type of intrapartum care (OU vs. AMU) at around the time of the booking visit. Questionnaires were mailed to ‘low risk’ women before booking. Sample characteristics: Women at ‘low obstetric risk’. No details reported. Response rate: 75 %, n = 113 (only 102 questionnaires (69 %) were used for analysis for various reasons).	May 1994	Hypothetical attributes of OU vs. AMU.
Emslie (1999) [21]	This study was conducted to explore women’s preferences and experiences following the opening of an FMU in the study area (Peterhead near Aberdeen in Scotland). Women in this area had four choices: home birth, FMU and both OU and AMU available approximately 35 miles away (in Aberdeen). A DOMINO (Domiciliary in and out) delivery service was also available to women registered with the FMU. The FMU was based in the Peterhead Community Hospital. The largest general practice is located in Peterhead with two rural practices in the surrounding area.	Methods: Questionnaire survey mailed to women in the FMU’s catchment area at around 14 weeks gestation, at 36 weeks gestation and 6 weeks postnatally. This survey was one component of a mixed methods study. Sample characteristics: Over half (59 %) of respondents (n = 254) were registered with the main GP practice in the FMU catchment area; 41 % of women were nulliparous; 70 % were aged under 29 years and 28 % were under 24 years of age. Response rate: 77 % for 14 week survey, n = 254. Of these 83 % responded to 36 week survey, n = 210.	January to December 1995	Study focuses on FMU vs. hospital (OU/AMU) choices made by women in the catchment area of a newly opened FMU.
Hundley (2001) [16]	Pilot study to explore feasibility of using discrete choice experiment to assess women’s preferences for aspects of intrapartum care. The study was conducted in three areas in Grampian, Scotland where different models of care were available. Linked study: Hundley (2004).	Methods: Discrete choice experiment. Data were collected by postal questionnaire from women recruited at booking. Sample characteristics: Of the 301 ‘low risk’ respondents, the mean age was 28; 55 % were nulliparous; the vast majority (91 %) were married or cohabiting. The women were more socioeconomically advantaged than the national population. Response rate: Estimated response rate was 40 %, n = 301.	January to November 1999	Study evaluates preferences for different service attributes.
Hundley (2004) [15]	This study was conducted to investigate the effect of service provision on consumer preferences, in particular, whether women who have access to systems of care which offer particular attributes value these attributes more highly than women for whom the attributes are not a realistic option. Three groups of ‘low risk’ women participated from areas with different services available (OU/AMU, FMU and OU/AMU without an epidural service). The areas also differed in the degree of continuity of carer offered. For primary report see Hundley (2001).	Methods: Discrete choice experiment. Data were collected by postal questionnaire from women recruited at booking. Sample characteristics: See Hundley (2001) for characteristics of the overall sample. ‘Low risk’ women in the three study groups were similar, but there were more nulliparous women in the Aberdeen (OU/AMU) group and women in the Elgin (OU/AMU without epidural service) group were less deprived. The Peterhead and Elgin groups were relatively small (n = 48 and n = 60) compared to the Aberdeen group (n = 193). Response rate: Estimated response rate overall was 40 %. Response rate varied by area (33 %–44 %), n = 301 (193 from the Aberdeen group, 48 from the Peterhead Group and 60 from the Elgin group).	January to November 1999	Preferences for particular service attributes in women with access to: OU/AMU vs. FMU ~30 miles from OU/AMU vs. OU/AMU without an epidural service.
Lavender (2005) [22]	This project was commissioned by the Department of Health (UK) to inform the Children’s National Service Framework. The aim was to identify models of maternity care which provide a safe, equitable and sustainable service that meets the needs of the current and future population and offers choice to women.	Methods: Questionnaire survey of pregnant women in a purposive sample of 12 maternity units in England. Units were included that offered different birth settings (home, FMU, AMU and OU) and varied in size (50 births to 6000 births). This survey was one component of a mixed methods study. Sample characteristics: Half (51 %) of the 2071 questionnaires returned were from district general hospitals (presumed to be OUs), 38 % were from university hospitals incorporating midwife-led units (presumed to be OU/AMUs) and 11 % were from FMUs. The mean age of participants was 29 and the mean gestational age was 29 weeks. Just over half (54 %) were multigravid with most having given birth to one child previously (46 %); 84 % were ‘white-European’ and 90 % had English as a first language; approximately 15 % (n = 303) were classified as being from ‘ethnic minority groups’. Response rate: Overall 71 %, with unit response rates varying from 59 to 85 %. n = 2071.	January to March 2002	Preferences for a range of service attributes.
Pitchforth (2008) [20]	A discrete choice experiment to evaluate preferences for key attributes of intrapartum care in women living in remote rural areas in Scotland served by FMUs and small consultant units without neonatal facilities.	Methods: Discrete choice experiment. Sample characteristics: The mean age of respondents was 30 years, 43 % women had delivered their first baby. Response rate: 62 %, n = 877 (including 22 of whom returned blank questionnaires).	April 2004 to January 2005	Preference for hypothetical attributes of midwifery-led vs. consultant care
Rennie (1998) [23]	A pilot study to identify women’s preferences for aspects of intrapartum care and to evaluate whether they differ in the postnatal period compared with late pregnancy.	Methods: A questionnaire survey of pregnant women at around 34 weeks gestation, with a follow-up questionnaire 10 days after the birth. Sample characteristics: Despite stratified sampling there was a preponderance of nulliparous women (65 %); 81 % of participants were married and two thirds (66 %) were owner occupiers. Most (70 %) were planning to attend antenatal education. The mean age of respondents was 27. Response rate: 96 % for the 34 week survey (n = 207); 86 % of respondents also completed the postnatal questionnaire (n = 185).	February to March 1996	Study focuses on service attributes preferred antenatally vs. postnatally.
Rogers (2011) [24]	This study was conducted to evaluate the viability of converting an AMU in outer London to an FMU following the planned closure of the OU in the hospital. The study focused on whether users of the existing AMU would choose the new FMU or would look for an alternative.	Methods: A questionnaire survey conducted amongst a cross-sectional sample of ‘AMU users’: women who were either booked, considering booking or who had given birth at the AMU situated in a hospital where a relocation of the OU was planned. Sample characteristics: The majority of study participants were pregnant (89 %) and the remaining 11 % had just had a baby. Sixty percent of participants were nulliparous. Response rate: 53 %, n = 121.	October 2009	AMU vs. FMU

Note that the for some studies, the calculation of response rates varied between reports. In these instances we directly quote the response rate reported by the authors