This paper reports the descriptive results of the GLOWING pilot trial relating to midwives reporting of their self-efficacy, outcome expectancies, intentions and routine behaviours in the context of UK guidelines for weight management during pregnancy. The data reported in this study provide some proof of concept for the evidence-based theoretical models that were developed to underpin the GLOWING intervention. Prior to GLOWING, the evidence-base suggested that health professionals’ low self-efficacy was central to the barriers and facilitators to the implementation of the UK guideline recommendations, and that the majority of barriers were related to the support/intervention-related behaviours. This is reflected in the GLOWING data which demonstrated that, pre-intervention, midwives’ self-efficacy was the lowest scoring construct across all behaviours, and the support/intervention-related behaviours tended to score lower than the communication-related behaviours. The GLOWING intervention was developed to address the evidence-based barriers incorporated in the SCT models. There was a particular emphasis on improving midwives’ self-efficacy for both communication- and support/intervention-related behaviours, but with more focus on the support/intervention related barriers to practice. The descriptive data reported by midwives in the GLOWING pilot trial suggests that there was limited change in self-efficacy for the control arm, whereas the self-efficacy scores increased in the intervention arm and were consistently higher than the control arm post-intervention. This was apparent for both communication- and support/intervention-related behaviours, but more so for the support/intervention-related behaviours. The pilot trial was not powered to be able to detect a significant difference between the intervention and control arm, or change in self-efficacy from pre- to post-intervention. However, the data are suggestive that the intervention may be impacting on the target construct of self-efficacy, particularly for the support/intervention-related behaviours.
The data also suggested a potential ceiling effect for midwives reporting of intention and behaviour constructs for communication-related behaviours. Some of these behaviours (e.g. relating to measuring and discussing BMI at the booking appointment) are now embedded into routine care. The BMI measurement influences the further discussions and referrals required relating to clinical management of pregnancy, such as referral for routine screening for gestational diabetes or consultant obstetrician led care (35). While the evidence-base suggests that weight communication might be perceived by midwives as a difficult conversation to have (i.e. impacting on self-efficacy and outcome expectancy constructs), they may feel that this is part of their professional role and routine care, which could explain the potential ceiling effect relating to these constructs and behaviours. However, the support/intervention-related behaviours are not necessarily part of routine clinical care pathways, and midwives may feel less conflict reporting lower intention or behaviour relating to these, as demonstrated in the data, due to it not being an explicit part of their professional role. If not part of routine care, then they may have less experience performing these behaviours, influencing their self-efficacy, and negative experiences when performing these behaviours less frequently may have influenced their outcomes expectancies (e.g. women getting upset).
This paper also reports data on midwives’ beliefs about people living with obesity using the BAOP scale. There was a similarity in pre-intervention BAOP data from midwives in this study (mean score 14.6 SD 5.7) to published data from a general population of UK adults (mean score 14.7 SD 6.7) (36), suggesting that midwives hold similar views about people living with obesity to the wider society. Although the GLOWING pilot study was not powered to detect a statistically significant change in BAOP scores, there was an increase in scores in the intervention arm and no change in the control arm. However, this change was minimal and unlikely to have any meaningful clinical difference on practice. A recent meta-analysis of 17 studies reported that health professionals hold implicit and/or explicit weight-biased attitudes toward people with obesity (37), highlighting the continued need to address this issue with health professionals. The elements of the GLOWING intervention addressing weight bias could be strengthened in light of this descriptive preliminary data and evidence-base of health professionals’ ongoing need for support.
Despite the wealth of evidence of the multiple and complex barriers to practice for health professional guideline implementation relating to maternal obesity and weight management practice (20), an absence of curricula in universities (38), and a call to action for midwifery education and training (39), there is an absence of adequately powered trials of interventions in this field. There have been before and after studies published with no control arm, primarily feasibility studies. In the UK, a feasibility study of a compact midwifery training intervention among 32 practicing midwives resulted in increased knowledge and confidence relating to NICE guideline behaviours (40), and an online programme delivered to 52 final year midwifery students (41) increased students’ subjective norms, perceived behavioural control, and knowledge of BCTs “to discuss lifestyle change with obese patients”, but not their intention or attitudes. In Australia, an online CPD course with before and after questionnaires completed by 36 health professionals identified increased perception of the importance of weight management for pregnancy and confidence to provide advice, but no difference in knowledge (42). Whilst these before and after studies report some positive data, there are limitations relating to the lack of a control arm and small sample sizes. Although the pilot GLOWING showed minimal difference between pre- and post-intervention data in the control arm, there were some small changes (although the study was not powered to know if these were statistically significant). It is possible that the passage of time, and midwives’ exposure to events or resources unrelated to the intervention, could result in changes in midwives’ practice, confidence, knowledge etc. which need to be accounted for using a control arm, or factored into statistical analysis of single arm studies. A definitive trial with adequate power is required to determine the effectiveness and cost-effectiveness of implementation interventions relating to maternal obesity and weight management.
There are strengths and limitations to this research. The intervention was developed following a rigorous approach using evidence-based theoretical models. A recent scoping review of implementation interventions in the maternity context identified that out of 158 published studies, only 14 reported the use of a theory, model and/or framework, and these typically guided data analysis or data collection rather than the design of the study (43), demonstrating the novelty and rigour of the GLOWING study in comparison to others in this field. The questionnaire used to collect data on the SCT constructs had to be developed to be tailored to this study and therefore was not a validated questionnaire; however, it demonstrated good internal consistency. However, this paper reports the descriptive results of a pilot trial, and therefore it is not powered to determine any statistically significant change in midwives reporting of the SCT constructs or BAOP questionnaire. There was loss to follow up, particularly in the intervention arm, which was primarily related to intervention fatigue. However, there were minimal differences in socio-demographics or baseline measurements of SCT constructs between the midwives lost to follow up and those who returned both questionnaires. This paper reports only quantitative data which does not fully reflect the mechanisms. However, we have also conducted a qualitative process evaluation of midwives’ experiences of the intervention, and their views on the impact of the intervention on routine care which will be reported separately. The GLOWING data for pilot trial feasibility will also be reported separately.