Evaluating midwifery units (EMU): Lessons from the pilot study

Abstract

Objectivethis paper describes the pilot study that was undertaken to test the feasibility of the recruitment plan designed to recruit women who booked to give birth in two freestanding midwifery units in NSW, Australia. The pilot preceded the full prospective cohort study, Evaluating Midwifery Units (EMU), which aimed to examine the antenatal, birth and postnatal outcomes of women planning to give birth in freestanding midwifery units compared to those booked to give birth in tertiary level maternity units in Australia and New Zealand. Designa prospective cohort study with two mutually-exclusive cohorts. Settingtwo freestanding midwifery units in NSW and their corresponding tertiary referral hospitals. Participantsa total of 146 women with few identified risk factors recruited between 13 September 2009 and 31 March 2010 whose planned place of birth was either a freestanding midwifery unit or tertiary maternity unit. Measurements and findingsthe pilot study identified the feasibility of relying on the booking midwife to recruit eligible women from several antenatal booking clinics to the study. Low rates of eligible women were invited resulting in a lower than expected consent rate. In addition, although mostly only low-risk women were invited to participate, some women requiring medical consultation at the time of booking were inadvertently recruited into the study. The results of this pilot study led us to revise the study protocol to find ways of including the outcomes of all women without identified risk factors who booked at either the freestanding midwifery units or the tertiary referral hospital in that area. This paper describes the revisions that were made to the study plan. Key conclusionsfive lessons were learned from the pilot study. We found that recruitment protocols employed for the cohort study were too complicated and required simplification to maximise the potential of the study. The study protocol needed to be changed for the main study to ensure a larger sample size and to ensure the risk profile of each cohort was as similar as possible. Sources of data needed to be expanded to produce a complete data set. Implications for practicepilot studies are extremely useful tools in testing methods to inform research protocols. We found that the first months spent undertaking a pilot study ensured a stronger design with the potential to show more meaningful results.