This presentation will describe the development of a core outcome set in reduced fetal movement research and will cover what core outcome sets are, why they may be useful, how they are constructed, and the purpose of developing one for use in reduced fetal movement research.
Dexter Hayes is a PhD student at the Tommy’s Research Centre in Manchester, UK. His research focuses on evidence synthesis in reduced fetal movement research and aims to determine if combining the results from published studies can determine whether interventions for reduced fetal movement have a beneficial effect on reducing adverse pregnancy outcomes.
Mr. Hayes has disclosed that he does not have any real or perceived conflicts of interest in making this presentation.
Jeanine Sabatino: Dexter Hayes is a PhD student at the Tommy’s Research Centre in Manchester, UK. His research focuses on evidence synthesis in reduced fetal movement research and aims to determine if combining the results from published studies can determine whether interventions for reduced fetal movement have a beneficial effect on reducing adverse pregnancy outcomes.
Dexter Hayes: Hi. I’m Dexter, and I’m a PhD student at the Tommy’s Stillbirth Research Center in Manchester in England. Today I’ll be talking to you about the Development of a Core Outcome Set for Reduced Fetal Movement Research. This is part of my PhD where I’m looking at different forms of evidence synthesis and studies of reduced fetal movement. I’ll talk a little bit about how this fits into that today, but I’ll be mostly focusing on the development of this core outcome set.
I’m going to describe the development of a core outcome set in reduced fetal movement research, and I’ll cover what core outcome sets are, why they might be useful, how they are constructed, and the purpose of developing one for use in reduced fetal movement research.
A couple of learning objectives. The first one is what are core outcome sets, and why are they useful in research? The second one is how is a core outcome set for use in reduced fetal movement research being developed?
Some background about reduced fetal movement, which I will be referring to as RFM throughout this talk. In the literature you might also see this referred to as DFM or decreased fetal movement or AFM or absent or reduced fetal movement. It’s thought that women present to hospital with concerns about reduced fetal movements in between 5 and 15% of pregnancies, and studies have shown repeated associations between RFM and adverse pregnancy outcomes such as stillbirth and fetal growth restriction.
Because of these associations, studies have aimed to improve pregnancy outcomes with interventions that aim to raise awareness of reduced fetal movement using things such as kick counting or providing information to pregnant women about fetal movements with the aim that this increases the proportion of women with reduced fetal movement who present to hospital. These interventions are often combined with clinical testing, such as ultrasound or measurement of the fetal heart for women who do present to hospital with RFM, and these clinical interventions aim to identify which of these pregnancies are at risk of adverse outcome so that other measures can be taken to prevent these outcomes.
As an example of this sort of intervention and study, this is a recent randomized trial which aimed to examine whether this mindfetalness method, which is a method for raising women’s awareness of fetal movements. Women were told about the importance of fetal movements and that they should contact hospital if they are concerned about their baby’s movements. If they are feeling fewer movements than usual, or if the movements aren’t as strong as usual, for example. This was compared with standard care. Half the women in this trial were assigned to this mindfetalness methods, and half were assigned to standard care.
The main outcome for this trial was Apgar score, and there was not a difference in Apgar scores between these two groups, but there were fewer caesarean sections in the mindfetalness group and also fewer babies who were born small for gestational age.
The first part of my PhD was a systematic review where I combined data from several studies of just that in order to try and see bigger picture and see whether these interventions affect pregnancy outcomes or see whether different interventions, so whether interventions for kick counting, for example, differ in effectiveness from those such as mindfetalness that we’ve just seen. What this review showed was large variation in the outcomes that were reported by studies. This made combining all the data from studies quite difficult.
Just to illustrate this, each row in this table is a different study that I extracted data from for my review, and each column is a different pregnancy outcome that was measured by this study. I’ve just put a little cross in the table to show that our study measured that outcome.
What you can see here is that there were not any studies that measured all of these outcomes. There were not any outcomes that were measured by all studies. Stillbirth, for example, there’s still a couple of cells in this table that don’t have crosses in them for that column. You can see this big gap in the middle of the table as well, which shows that the studies in group B, which were studies of kick counting, quite a lot of them didn’t measure any of these outcomes such as neonatal death, NICU admission, and caesarean section, for example. This meant that combining data to see what the effect of these interventions were on these outcomes was not possible.
What this means is that we still don’t really know what the effects of these interventions for RFM are in terms of whether they affect the portions of stillbirth or other adverse pregnancy outcomes. This review led to us thinking how to try and fix this. One of the ideas was to do a core outcome set as there is currently not a core outcome set for RFM.
What is a core outcome set? A core outcome set describes a standardized set of outcomes that should be measured and reported in all studies in a specific area as a minimum. They are currently used in several health care fields, and also used in maternity care.
As an example of this, here is a core outcome set for preeclampsia, which is a condition that causes high blood pressure in and after pregnancy. This was published last year, I think, and here is a list of outcomes that the study authors recommended should be measured in all future studies of preeclampsia. As you can see, there are quite a few outcomes here. They recommend that this should be used in future trials, and that by using this, it will inform clinical practice and improve care and outcomes of pregnant women and their babies.
I’ll describe how they arrived at that list of outcomes later, but I just also want to talk about this study which was not in pregnancy care, but it was looking at how useful a core outcome set for arthritis has been. It’s only really useful to conduct a core outcome set if it has an impact on future research. This core outcome set for arthritis was one of the earliest core outcome sets, which was published in 1994. This study looked at trials that were carried out a few years after this between 2002 and 2016 and found that the core outcome set was used in 81% of trials that were conducted within this timeframe. After this core outcome set was published, the majority of studies in this area did actually use the core outcome set when deciding which outcome should be measured by that study.
Just to summarize that first part of the presentation, a systematic review shows a lot of variation in the outcomes that were recorded by intervention studies for RFM. This made synthesis of these studies difficult, and this also means that guidelines for the clinical management of RFM are not always informed by high-quality evidence and this can lead to variation in practice.
The aim of this core outcome set is to create one that can be used in research studies that aim to improve identification of reduced fetal movement or studies that aim to combine this with clinical interventions for RFM. This is not for studies that only look at clinical interventions. As we decided, this is a substantially different study population. These studies would be looking at pregnancies with RFM only compared to the studies we’re looking at where it’s all pregnancies, but we might revisit this at a later date perhaps making one of the core outcome sets.
Hopefully this will ensure consistent and relevant outcome measurement, which may also lead to more robust results in future studies. It may also lead to improved wellbeing of pregnant women and better clinical practice.
I’m going to talk a little bit about the methods and how you actually carry out a core outcome set. It’s a three-stage process. It starts with a systematic review. Then that is an online survey called a Delphi survey, and then finally a consensus meeting.
Stage one is the systematic review. This does differ from the other review in my PhD that I talked about as the scope is slightly wider. I couldn’t just use the same studies. We included some extra ones in this, but the idea here is that you identify any relevant study that has been published. Any study that has used an intervention to aim to improve detection of RFM by increasing awareness or kick counting as I talked about, but also other things such as leaflets to disseminate information or things such as mobile phone apps.
Then when you have gathered all of these studies and screened them against your inclusion criteria, you extract all of the outcomes from these studies. So make a big list of anything that was measured by any of these studies, as well as how those outcomes were defined by the studies. What we did then was grouped these into maternal or neonatal outcomes and also combined different definitions of the same outcome into single outcomes.
These are the inclusion criteria that I just mentioned. I’ve spoken about the intervention group before, but we were interested in singleton pregnancies presenting in a maternity care setting after 28 weeks gestation. What’s really important is that we were only looking at studies with a comparator group. We wanted studies that, for example, in the study that I talked about earlier, the mindfetalness trial, they compared the outcomes in the mindfetalness group to a standard care group. What this means is that we could be more confident that any differences in the results can be attributed to the intervention because you have something to compare it to. The study design box just says that really and that we need controlled studies be randomized or non-randomized, and just need a clear inclusion criteria and methods.
The results of this is that we started with 215 outcomes, which were extracted from 38 studies. After combining multiple outcomes and looking at the different definitions that we used by each study, we determined that there were 50 unique outcomes. We split these into maternal and neonatal outcomes and found 24 maternal and 26 neonatal. These are the outcomes that will then be forwarded into stage two of the core outcome set, which is the online survey.
Just to show you the kinds of outcomes that will be included, these aren’t the individual outcomes, but groups that we broke them down into in order to determine which ones were unique or not. This is just the overview. In maternal outcomes, things such as mode of birth, maternal-fetal attachment, and maternal anxiety. Neonatal outcomes are things like birth weight, stillbirth, neonatal admission, and neonatal death.
The next stage is the Delphi survey. This is an electronic international survey which will be conducted including key stakeholders. The aim is to produce a preliminary core outcome set, so to reduce those 50 outcomes into something smaller.
These are our three stakeholder groups. Anyone who is eligible to take part in this survey. The first group is anyone who is, or who has been pregnant as well as their partners. The second group is anyone who works in research, funds research or works in making policies related to RFM. The third group is clinicians. This can be midwives, obstetricians, neonatologists, or GPs. So any clinicians with experience of RFM.
Round one collects demographic data from all of our participants which are used to analyze the data later on. Participants are then presented with a list of the 50 outcomes that we identified and asked to rate how important they feel each of these outcomes are in reduced fetal movement research. This is done using a nine-point Likert scale, which has one of these. For each outcome, participants will rate how important they think each outcome is, and these outcomes will have descriptions in case people are not sure about what they are, and if participants are unsure, they also don’t have to rate every outcome.
Participants are also prompted to add additional outcomes if they think there are any that we have missed. Every outcome from this round is forwarded into the next round. The way these outcomes progress through the study is that we apply standard definitions to the results, though, all of the outcomes from round one go into the second round, but after this, the outcomes are rated again.
To progress to the final round, these outcomes must be scored as between seven and nine, the highest points on the scale that you just saw, by 70% or more of all participants, including at least one from each of these stakeholder groups that I’ve described must also be rated as not important by less than 15% of participants.
In round two, everyone gets the chance to see how their group voted, as well as every other group. Feedback is provided in terms of average scores for each outcome that were voted by age group. Then people are asked to re-score all the outcomes in light of this, as well as any outcomes that were added after round one.
Using the definitions that I described, this gives us a preliminary core outcome set. Some outcomes are removed after the third round, and this gives us a list that is taken into a consensus meeting. This will be an online meeting using Zoom in order to facilitate international attendance. This includes a presentation of the findings of the survey, including the outcomes that were selected and how they were voted for by each group. Then there is a discussion and a vote on each outcome. This is done in a similar way. If outcomes are voted for by at least 70% of participants, including one from each stakeholder group, then these are included in the final core outcome set.
The next steps for us are we have obtained ethical approval from University of Manchester Ethics Committee. We now need to recruit for the Delphi survey, which involves contacting relevant organizations and charities and seeing if they want to forward the details of the survey to their email lists, or if they know people who want to get involved and also recruiting online via social media. We’re encouraging people to forward information about this study to other people who also might be interested.
Just to reiterate who can take part, anyone who is or who has been pregnant or their partners, anyone who works in research related to RFM, or any clinicians with experience of RFM.
Why we think it’s really important for people to take part is that participants’ views will help to shape future research and improve studies of interventions for RFM. We are particularly interested in making sure that the views of people who have been pregnant and the partners are heard and not just those of researchers and clinicians. Obviously, it’s important that the views of researchers and clinicians are taken into account, but fairly often in research, it’s only these voices that are heard, and we want to make sure that people who have been pregnant are included and are treated equally.
This is just the first part of the participant information sheet, which provides a bit more detail in terms of what would be expected from anyone who does take part and things such as distress protocols and what is done with participants data and who to contact during the study.
Thank you for listening. If anyone does want to take part, then please contact me. My email is [email protected] I will add my email into the slides and make it available for anyone who wants it, but I would just like to thank my supervisors at the University of Manchester. Alex, Jo, and Tanya, and also Declan and Valerie who were really helpful in terms of setting up the core outcome set. Thank you.
This presentation was part of the Stillbirth Summit 2021. This individual lecture will be awarded .5 hours of continuing education credit to include viewing the lecture and completing evaluation and post-test. Once received a certificate will be emailed to the address you provide in the post-test. If you did not register for the Summit WITH Continuing Education, you can purchase the continuing education by clicking here. This purchase will provide you access to all Stillbirth Summit 2021 lectures including continuing education credit. There is no charge for viewing the presentation.
To receive continuing education credit for this lecture, the participant must complete the evaluation and post-test.
Please feel free to ask questions of the presenter. We will obtain their answers/comments and provide them here as received.
Hi Dexter. A worthwhile topic for your PhD. Will you recommend specific psychometric measures of attachment and anxiety or will it remain as a broad “they measured these… somehow?”
Mr. Hayes’ Reply:
Thank you for your question. If outcomes with multiple definitions/measures are selected by the Delphi process for the final core outcome set, then we will go back to our list of how these were measured by each of our included studies and see if there is a consensus on the best way of measuring these outcomes. For outcomes such as anxiety or attachment it may be that we report multiple scales. In my meta-analysis I was able to combine data across different anxiety scales so this should not impede synthesis of future studies.