Methods
The study was conducted at a large tertiary referral centre, Mercy Hospital for Women (MHW), in Melbourne, Australia. A cross-sectional survey was used.
All midwives working at the MHW who provide any antenatal care were sent a survey to determine their practices regarding advice given for nausea and vomiting in pregnancy. Midwives from the family birth centre (FBC) and the emergency department were included.
The questionnaire was developed specifically for the project, as no suitable validated questionnaires were identified. Question areas included: demo¬graphic information; what type of general advice midwives give to women experiencing nausea and vomiting in pregnancy; whether midwives recom¬mend any vitamin or herbal supplements; what supplements they recommend; whether they refer women to alternative practitioners; and how they obtain knowledge in this area. Demographic ques¬tions were based on questions used successfully in previous midwifery surveys within the study hospital and other maternity hospitals. The remaining questions were designed to extract as much mean¬ingful information on the primary outcomes as possible, i.e. midwives’ general advice for morning sickness and how herbal supplements are prescribed in this context. Questions were asked specifically about the herbs that are commonly prescribed for nausea and vomiting in pregnancy, to elicit more information about how these herbs are prescribed by midwives, and awareness and knowledge of any harmful effects. Further questions were designed to obtain information on secondary outcomes such as how midwives obtained their knowledge and what influenced the advice that they gave.
The questionnaires were piloted extensively with research colleagues and then with six midwives who were not eligible for inclusion to ensure that all relevant questions on the topic were included, that the questions would elicit as much meaningful information as possible and that they were clear and easily understood. After piloting, the ques¬tionnaires were revised and then repiloted until no further changes were necessary.
A midwifery staff list was sourced from all relevant units and rotating rosters, and the questionnaires were personally delivered to staff mail boxes in each area. Self-addressed envelopes were provided for questionnaires to be returned via internal mail to the research office and return boxes were also provided in all areas. Contact details for the research team were provided so that potential participants could seek any clarification they re¬quired. No identifying information was included on the questionnaires, to maintain anonymity. A reminder letter was therefore sent to all eligible midwives (as the identity of the nonrespondents was not known) 2 weeks later and a final reminder (with a copy of the questionnaire) at 4 weeks.
Data were entered into an Access database (Microsoft Corporation, 2000). Quantitative data, were analysed using Stata (Stata Corporation, 2003). Descriptive statistics were used to summarise the quantitative data, and openended comments were coded into themes.