Effectiveness and cost effectiveness of a 12 month automated text message intervention for weight management in postpartum women with overweight or obesity: protocol for the Supporting MumS (SMS) multi-site, parallel-group, randomised controlled trial

Background [6a]

This trial addresses the development of overweight and obesity across the childbearing years. Entering pregnancy with a high body mass index (BMI) increases many health risks for these mothers and their babies, including gestational diabetes, hypertensive disorders, caesarean section and birth complications, compared with recommended weight women.(3, 4, 5, 6) Excessive gestational weight gain (GWG) and postpartum weight retention (PPWR) are common (7, 8) and many women are at risk of further weight gain across the extended postpartum period.(9) By 18 months postpartum, about one in five women have moved into a higher BMI category (8) which increases the risk of complications in subsequent pregnancies and contributes to long-term overweight and obesity.(9)

Effective and appropriate weight management interventions in women during the postpartum period, which account for the well-recognised barriers for new mums such as time constraints and childcare, are lacking.(9, 10, 11, 12, 13, 14) Previous intervention studies are characterised by poor recruitment and high rates of attrition (1, 15, 16, 17, 18) and have not adequately considered the difficulties in reaching this population and specific barriers to lifestyle behaviour change that come with having a baby.(19, 20, 21, 22, 23) In-person and structured weight management approaches, such as those delivered in community and group settings, are unlikely to be successful with this population due to the commitment required, and such approaches may exacerbate health inequalities.(24, 25, 26, 27) More appropriate ways of recruiting and engaging with postpartum women to achieve sustained behaviour change are required (1, 9) and a need to evidence cost-effectiveness to inform implementation.(12)

Recent evidence, mostly from feasibility or pilot studies, indicates that technology usage shows promise in postpartum weight management, perhaps because this form of delivery fits more seamlessly with women’s lives at this stage.(17, 28, 29, 30) Mobile technologies can offer a more flexible and individualised ‘any time, any place’ approach to behavioural weight management interventions.(9) Text message interventions have high reach potential and allow for flexible scheduling and interactivity,(31) potentially making them convenient for mums with limited time and helpful for overcoming health inequalities. Systematic reviews support the use of text messaging in behavioural weight management interventions but indicate further evaluation is required.(32, 33, 34, 35, 36). They are cost-effective and readily scalable for changing other health behaviours.(37, 38)

The Supporting MumS (SMS) intervention, consisting of a library of text messages, was developed with personal and public involvement (PPI) to support weight management in the postpartum period.(28) Messages focus on diet and physical activity with embedded behaviour change techniques (BCTs) informed by behaviour change theory and evidence. In a previous pilot evaluation, one hundred women with overweight or obesity from Northern Ireland (NI), who had a baby within the previous 24 months, were recruited and randomised to receive text messages about weight loss (intervention group) or child health and development (active control) for 12 months. The study demonstrated the feasibility and acceptability of the intervention and an independent Trial Steering Committee (TSC) judged that the pre-specified progression criteria to proceed to a full randomised controlled trial (RCT) (i.e., successful recruitment, high retention and no differential attrition, high acceptability of the intervention, and evidence of positive indicative effects) were met.(28)

Rationale [6a]

This RCT addresses the following research question: Is an automated 12 month text message intervention, designed to support weight loss and weight loss maintenance for postpartum women with overweight or obesity, effective and cost-effective for weight loss at 12 months, compared with an active control group receiving text messages on child health and development?

SMS is a parallel, 2-arm multi-site RCT using automated and bidirectional text messages to support weight management over a 12 month period in postpartum women with overweight or obesity who have had a baby in the last 24 months, comparing weight change at 12 months from baseline between the intervention and active control groups. This RCT protocol is based on the pilot RCT methodology, but changes were to remove: 1) use of a sealed pedometer to assess physical activity as this was not acceptable to women and, 2) a discussion forum intervention component as it was not used by pilot RCT participants. We added considerations for data collection in the event of COVID-19 restrictions. In moving from a single site pilot study to a multi-site RCT recruiting an ethnically and socioeconomically representative sample, we conducted a six-month period of intervention and recruitment method adaptation, with PPI, ensuring the broad acceptability and cultural relevance of the messages and enhancing the generalisability of the findings to different groups of women across the United Kingdom (UK), thus informing implementation.

The increased prevalence of obesity, its associated complications and weight management services, is costly to the National Health Service (NHS). NHS costs associated with caring for a pregnant woman with a BMI of ≥25 kg/m² are up to 37% higher than caring for a woman with a healthy BMI.(39) Both National Institute for Health and Care Excellence (NICE) guidance (13) and an associated review of the cost-effectiveness of weight management interventions following childbirth,(40) outline the potential of postpartum interventions to be a cost-effective way of reducing the long-term risks of obesity, heart disease, cancer and diabetes, and recommend health professionals should advise women within two years of having a baby to eat a nutritious diet and keep active to encourage postpartum weight reduction.(13, 41)

Postpartum weight management interventions combining dietary and activity behaviour change have moderate positive influence on maternal weight (12, 15) and those including self-regulatory BCTs are likely to be more successful.(10, 12) The postpartum period could be an ideal time to intervene to shape new health behaviours as women may be motivated and receptive to health information.(42) However, women highlight a need for additional weight management support during the postpartum period as little is currently provided.(43)

The present trial allows women to take part up to 24 months postpartum. Qualitative evidence suggests that every woman’s postpartum journey is different (19, 20, 21, 22, 23) and the optimal time to engage women in postpartum weight management is currently unknown.(9, 11) Allowing this extended inclusion window accommodates different maternal weight trajectories across this time.(44, 45) Furthermore, women may prepare for another pregnancy during this period, so there is potential to support better health for subsequent pregnancies.(9)

Mobile phone usage and ownership are widespread (95% of UK households) amongst all sectors of society, irrespective of socioeconomic status.(31) Text messaging is a simple communication mode that does not require smartphone or tablet technology, unlike web-based interventions or apps. This supports health equity through high reach potential at a low cost, making it conducive with scale-up and implementation, unlike some weight loss interventions which can be resource intensive and expensive (46) and can exacerbate health inequalities.(47, 48) A text messaging intervention can be delivered flexibly and on a sustained basis, can be reactive and proactive, and does not necessarily rely on participant initiation.(31, 49) Using the real-time advantages offered by mobile technologies to deliver behavioural weight management support to postpartum women has the potential to encourage behaviours that may improve maternal health in both the short and long-term.(50) Trial results are relevant for other population groups given the possible advantages of this weight management approach. The NHS aims to embrace the potential of digital strategies for improving UK health (51) and text message interventions could complement any weight management support currently delivered across the NHS, to help ease the current burden on an over-stretched service.

The SMS trial adopts a fully automated approach to intervention delivery whilst still offering feedback to participants. The trial makes a novel and important contribution to the field of behavioural text message interventions in a number of ways: there are currently few such interventions that are fully automated, where text messaging is the main mode of delivery and two-way messaging is used to encourage engagement and delivery of specific BCTs; there are also few which consider both weight loss and weight loss maintenance, last 12 months or more and use an active control to minimise disappointment bias.(28)

The National Institute of Health and Care Research (NIHR) awarded further funding to explore postnatal mental health outcomes and needs to inform application of the priorities in the NHS Long Term Plan via existing studies. We added data collection methods related to mental health (postnatal depression, depression and anxiety, mother and child relationships, service and treatment experiences) in February 2022, allowing an opportunity to examine poor postnatal mental health prevalence rates, trajectories, predictors, inequalities and service needs, in an ethnically and socioeconomically diverse sample from all four UK countries. The trial also allows further validity and reliability testing of the Me and My Baby (MaMB)/Me and My Child (MaMC) screening tool which was co-produced by health visitors and the community to assess the mother and child relationship. Funding for this workstream was contracted in March 2022.

Explanation for the choice of comparators [6b]

An active control comparator is used as pilot study findings indicate that this design aspect supports good participant retention, acceptability and satisfaction, where previous weight loss trials with postpartum women have reported high levels of attrition and disappointment bias in those allocated to no treatment or usual care control groups.(1) Active control participants are assessed for the same outcome measures as the intervention group at 6, 12 and 24 months. We advertise the purpose of the trial as testing two new text message services to minimise women’s preference for one group over another. The participant information sheet (PIS) clearly describes that women have a 50:50 chance of being assigned to receive messages about child health and development or messages about weight management, thus preserving participant autonomy.

Objectives [7]

Trial Design [8]

A pragmatic, multi-site, parallel, 2-arm, assessor-blind, 1:1 superiority RCT comparing weight change at 12 months for the SMS intervention and active control groups, with mixed methods process evaluation, cost-effectiveness modelling, consent for future data linkage to longer-term health outcomes for mothers and babies, 24 month participant follow-up (i.e. 12 months post-intervention) and implementation consultations with stakeholders.

Trial setting [9]

A multi-site trial with five recruitment sites in the UK: 1) Belfast, Northern Ireland (trial co-ordinating site); 2) Stirling, Scotland; 3) Cardiff, Wales; 4) London, England; and 5) Bradford, England. Recruitment sites were selected to target geographical areas of ethnic and socioeconomic diversity for the purposes of participant recruitment via community and NHS pathways.

Eligibility criteria [10]

Women (in accordance with the NICE Postnatal care guideline NG1941, the term ’woman’ used in this trial includes people who do not identify as women but who are pregnant or have given birth) are eligible to participate in the trial if they are aged 18 years or over, have a BMI of ≥25 kg/m², have had a baby within the last 24 months and own a mobile phone to allow them to receive personal text messages.

Women are excluded if their baby is less than 6 weeks old or if they have insufficient English language to understand short written messages, if they have had or plan to have any type of weight loss surgery, if they report having ever received a diagnosis of anorexia nervosa or bulimia from a doctor, or, if they are on a specialist diet and receiving dietetic care. Women are excluded if they are participating in any other weight management research study/programme currently or in the previous three months. Women are not eligible if they are pregnant and any consented participant becoming pregnant during the trial is excluded, as the intervention has not been designed for use in pregnancy.

Interventions

Active control [11a]

The active control messages were also developed by the trial team with PPI input (28), for mothers with babies aged six weeks up to 36 months (based on participant inclusion criterion). The messages relate to general childcare and development, with content consistent with evidence-based information provided by the NHS Start4life service (59) including weekly play ideas/activities and information on milestones, home safety, separation anxiety and similar topics. Active control messages do not address the intervention target or active ingredients (i.e. diet, physical activity and BCT content related to weight loss) nor do they prompt bidirectional functionality. Women receive three messages each week over 12 months with messages corresponding to the age of their baby, i.e., if their baby is six months old at randomisation, the participant starts to receive messages corresponding to this age.

Women allocated to the control group are offered to receive a booklet summarising the intervention content on completion of the trial.

Criteria for discontinuing or modifying allocated interventions [11b]

Participants in both groups have the ability to pause, stop and restart the messages throughout by sending an instruction to the text system e.g., ‘STOP’, ‘START’. Researchers ask women choosing to discontinue the messages to provide any reason for doing so, to inform the process evaluation. Participants discontinuing messages remain in the trial unless they formally withdraw from the trial assessments.

Strategies to improve adherence to interventions [11c]

Text message delivery is fully automated and centrally delivered via the LSHTM system which is intended to ensure treatment fidelity. Regular monitoring of the text message system is completed by the system developer and trial team members (CI and clerical officer) to check the content of sent and received messages (e.g., ensure web links are working, review spontaneous participant responses). Intervention engagement is assessed via text messages received from participants in response to the weekly weight prompt and the replies to bidirectional messages as described in [11a], self-report items in questionnaires at 6, 12 and 24 months and qualitative interviews at 6 and 12 months.

Relevant concomitant care permitted or prohibited during the trial [11d]

Women who become pregnant during the trial will be excluded, as text message content has not been designed for pregnancy and primary outcome data (weight) can only be collected from women who are not pregnant. This is a self-directed intervention and women are permitted to access any other weight-loss services via NHS, voluntary or private sector pathways. Details of engagement with other services are captured at each data collection timepoint. Any weight loss surgery undergone by participants during the trial, and prescribed or over-the-counter medications that may affect weight, are documented at 6, 12 and 24 months to assess the impact on trial outcomes.

Outcomes [12]

Trial outcomes, measures and the assessment schedule at 0, 6, 12 and 24 months are outlined in Table 2.

• ➢ Primary outcome: between-group difference in mean weight change (kg) from baseline to 12 months.

• ➢ Secondary outcomes: between-groups difference in mean weight change (kg) from baseline to 24 months and between-groups differences in mean waist circumference (cm), mean BMI (kg/m²), in the proportions of women gaining a substantial amount of weight (>5kg), dietary intake (Fat and Fibre Barometer plus questions on sugar intake), physical activity (IPAQ-SF) and infant feeding practices (Infant Feeding Survey), by the 12 and 24 month follow-up timepoints respectively.

• ➢ Health economic outcomes: NHS resource use and personal expenditure (GP usage, hospitalisations, prescribed and over-the-counter medications, lifestyle products and services, expenditure on food and drink, alcohol and smoking products); health-related quality of life (EQ-5D-5L with visual analogue scale) and Capability (ICECAP-A).

• ➢ Process Evaluation: to explore participants’ experiences and interactions with the intervention by site, ethnicity and socioeconomic status, including evaluating dose (number of women who choose to receive all text messages and those who pause or stop the messages), engagement with the intervention (text message responses and qualitative data), reach (trial records of recruitment, loss to follow-up and withdrawal) and acceptability (participants’ satisfaction ratings and qualitative data).

• ➢ Exploratory outcomes: moderator and mediator analysis (Table 2)

• ➢ Qualitative sub-study: telephone interviews with a purposive sample of participants at 6 and 12 months to understand the experiences of women receiving the intervention and active control text messages and the contextual factors related to engagement including postnatal mental health, to inform trial and process evaluation outcomes.

• ➢ Stakeholder consultation: if the intervention is shown to be effective at 12 months, stakeholder interviews will be conducted to explore factors relevant to intervention scalability and implementation, including any further work needed to develop a coherent implementation model.

Participant timeline [13]

The trial flowchart includes the participant timeline for the RCT (Figure 1).

• Download figure
• Open in new tab

Sample size [14]

The target sample size is 888 women over 12 months, with site-specific targets across the five geographical areas (London and Bradford, n=189, respectively; Belfast, Stirling and Cardiff, n=170, respectively). Pregnancy-related withdrawals are not replaced unless the loss to follow-up (12%) or pregnancy rates (15%) observed in the pilot RCT (28) are exceeded by the 6 month timepoint, which is monitored during Project Management Team (PMT) meetings.

Recruitment [15]

Based on successful pilot study recruitment processes,(28) trial information is disseminated via community settings likely to be attended by mothers of young children (e.g., parent and toddler groups, libraries, breastfeeding support groups), social media and through contacts with healthcare professionals and GP practices, including both routine appointments and contacting potentially eligible women identified using patient database searches. A range of recruitment pathways are used to target ethnically and socioeconomically diverse populations.

In Bradford, principal trial recruitment is through BiBBS, an existing birth cohort study, with trial information disseminated to participants who previously consented to be informed of other relevant research studies.

Screening [10]

All trial information materials disseminated via described recruitment pathways, direct women who are potentially interested in taking part to contact the research team via telephone, email or by accessing the trial website/scanning a QR code to submit a secure pro-forma with their contact details.

Site researchers contact women expressing an interest in their area to provide a PIS, privacy notice and copy of the consent form (supplemental appendix 2) via email or post. Audio versions of the PIS and privacy notice are available on the trial website to increase accessibility. To comply with the Welsh Language Act 1993, Welsh language versions of the PIS and Consent Forms are available, on request. Women are given at least 48 hours to consider the information before being contacted again by telephone by the research team to discuss the trial procedures and any queries then invited to take part. Up to three contact attempts are made to invite all women expressing interest for screening. BMI eligibility is initially determined using women’s self-reported height and weight but later measured and verified by the researcher prior to collecting informed consent. Women who are deemed ineligible due to their baby being less than six weeks old will be invited for rescreening when they reach six weeks postpartum.

Informed consent and withdrawal [26a]

For women meeting the eligibility criteria and willing to proceed with trial participation, a baseline researcher visit is arranged. Home visits are the primary approach to data collection, with women also offered the option of attending another venue if preferred (University building, community venue or their workplace).

Research staff, trained in Good Clinical Practice (GCP) guidelines, obtain written informed consent prior to collecting baseline data. At the baseline visit, women are offered the opportunity to ask further questions and confidentiality of data is explained. Informed consent is then taken. Women are reminded that they have the freedom to choose whether to take part or not. If willing to proceed, women are asked to initial and sign the trial consent form, with optional consent to be invited for telephone interview participation at 6 and 12 months. During the 6 and 12 month trial visits, women providing this optional consent at baseline are asked if they are still happy to be contacted to take part in a telephone interview. Participants completing telephone interviews are asked to provide their verbal consent for audio recording and for anonymised quotations to be used in trial publications before commencing interviews.

Once consented, participants are advised that they can withdraw from the trial at any time, without giving a reason, should they wish. Withdrawals are documented in the CRF along with any reason given for withdrawing. Participants excluded due to pregnancy are given the option to continue receiving the text messages. Anyone choosing this option needs to complete and return a pregnancy form to acknowledge that the messages have not been designed for pregnancy but may be useful after pregnancy. Data collected up to the point of withdrawal is retained for analysis unless the participant specifies otherwise.

Additional consent provisions for collection and use of participant data and biological specimens [26b]

Optional aspects of trial consent include permission to be informed of future follow-up to this trial or other relevant research studies and permission for future data linkage to routinely collected health records for long term health outcomes (mother and youngest child) which is not within the current trial timeline. No biological specimens are collected.

Assignment of interventions: allocation

Blinding [17a]

Researchers responsible for recruitment and collecting outcome data are blinded to group allocation. It is not possible to blind participants, but they are requested not to discuss the text messages they are receiving with researchers at each visit. This worked well in the pilot study with blinding maintained.(28) All statistical analyses are conducted blinded to group allocation until analyses of the 12 month data point are complete at which point the randomisation sequence is provided by the independent statistician who prepared it. We record any circumstances in which unblinding has occurred. The trial clerical officer, Chief Investigator and LSHTM text message system developer are not blind; these individuals see what group individuals are assigned to on the text message system for the purposes of monitoring the intervention delivery and the clerical officer is responsible for preparing the group-specific 12 month questionnaires. Blinded researchers do not see these questionnaires until after primary outcome data is collected. Interviews at 6 and 12 months are carried out by a researcher who has not had any previous contact with the participant.

Group allocation information for that participant is provided to the researcher before the interview so the correct interview guide can be used.

Emergency unblinding [17b]

Given the nature of the intervention, emergency unblinding is not required.

Data collection, management and analysis

Data protection and confidentiality [27]

All data is collected, stored and handled in accordance with the QUB Research Management Policy (2015), GDPR and any future relevant data protection legislation and all trial team members must be compliant with these regulations. Detailed description of the trial data processing is documented in a Data Privacy Impact Assessment and in the privacy notice given to women expressing an interest in the trial.

Participants are allocated a unique 5-digit participant identifier at randomisation which is used on all data collection forms/files to allow identification of data for each participant, with the key for unlocking pseudonymised data held by site teams in password-protected files. Hard copies of forms are securely stored in locked filing cabinets in locked offices in buildings requiring keypad access and which are alarmed outside normal working hours.

Consent and contact details forms, containing identifiable information, are stored separately to other data (CRFs and questionnaires) and will not be kept for longer than is required. All electronic forms or data files are held on password-protected systems that are routinely backed up and are only accessible using username, password and/or multi-factor authentication. Third-party transcription providers sign a confidentiality agreement detailing data handling requirements. Any file transfer between QUB and other parties (sites/transcription provider) is done, subject to relevant data sharing agreements, via secure and encrypted file transfer systems (e.g., QUB Dropoff). Participants will not be identifiable from any published trial report.

Data collected from women expressing an interest in trial participation but who are not recruited will be confidentially shredded/deleted.

Access to data [29/31c]

Access to trial records and source data can be granted to authorised Sponsor (or delegate), REC, host institution and regulatory authority representatives for trial-related monitoring, audits and inspections, as per participant consent.

Only employed trial team members have access to personal/sensitive data of participants and data is shared with those responsible for the healthcare of the participants only when health concerns are raised e.g., EPDS score of nine or greater, see [30].

The Trial Manager, Statistician and Health Economist will have access to the anonymised dataset at the trial end, to permit analysis.

The CI will manage access rights to the trial data set, in collaboration with the sponsor, and formal requests for data access need to be made in writing to the CI. The trial will comply with the good practice principles for sharing individual participant data from publicly funded trials and data sharing will be undertaken in accordance with the required regulatory requirements. Any transfer of data to other institutions will be governed by a Data Access Agreement and will be in the form of an anonymised dataset. In the event of publications arising from further data analyses, those responsible will need to provide the CI with a copy of any intended manuscript for approval prior to submission.

Statistics and analysis

Aligned with best practice and following appropriate reporting guidelines, a Statistical Analysis Plan,(79) Health Economics Analysis Plan (80, 81) and Process Evaluation Analysis Plan are written and approved by the PMT and TSC prior to analysis. Trial results will be reported in accordance with the CONSORT guidelines, with the flow of participants summarised in a CONSORT flow diagram (http://www.consort-statement.org/) and tables of summary baseline characteristics presented, by total recruited sample and 12 month completers.

Project Management Team (PMT) [5d]

The PMT is the key decision-making group consisting of the co-applicants, Trial Manager and PPI representative. The PMT meet online every 2-4 months to oversee the trial design, management and conduct, including monitoring participant safety, see [22].

Adverse event reporting and harms [22]

Serious adverse events (SAEs) are recorded by researchers from the time of the participant’s consent into the trial until one month after the end of the 12 month intervention period and reported in line with GCP guidelines and the Sponsor’s adverse event reporting procedures. Site PIs review SAEs within 24 hours to determine severity, relatedness and expectedness. The CI (or delegate) is informed of any event evaluated as being severe, related and unexpected, to conduct a second review. Onward reporting of any SAEs categorised as related and expected by both the PI and CI is actioned by the CI or Trial Manager to the Sponsor and REC within 24 hours and 15 days respectively. All SAEs are summarised and included in annual reports to the REC and reported during meetings of the PMT and TSC.

Ancillary and post-trial care [30]

This is deemed a low-risk public health trial and no evidence of harm has been reported in text message studies for other health behaviours or in our pilot RCT, so no post-trial care is delivered. All women are given a signposting ‘Useful Contacts’ leaflet during trial visits advising them on relevant agencies and organisations that they might contact for support on a variety of issues e.g., bereavement. We monitor any emerging safety issues or unintended consequences and any participants experiencing such issues will be referred to their GP for appropriate treatment. This includes the onward reporting of EPDS scores of nine or above (Table 2). The EPDS is a screening questionnaire used to detect postnatal depressive symptoms and the scale guidance indicates that women with scores meeting this threshold would benefit from further clinical assessment. We calculate scores within one week of receiving the completed EPDS at each follow-up timepoint and letters are sent to the GP and the participant to inform of scores ≥9, as per PIS and consent processes.

Ethical and Regulatory considerations [24]

The trial is conducted in accordance with the recommendations for physicians involved in research on human participants adopted by the 18^th World Medical Assembly, Helsinki 1964, and later revisions, the UK Policy Framework for Health and Social Care, relevant UK data protection legislation and the principles of GCP. QUB acts as sponsor for the trial and has in place comprehensive SOPs for the approval and monitoring of research. This protocol and other relevant trial documents were reviewed and approved by the West of Scotland Research Ethics Service REC 4 22/WS/0003. Correspondence with the REC is retained in the Trial Master File. An annual progress report is submitted to REC until the trial end. The CI will report to REC when the trial ends (or is terminated prematurely) and a final report with trial results submitted within one year of this.

Appropriate approvals from participating organisations e.g., NHS R&D approvals are obtained before recruiting via applicable routes.

Dissemination [31a]

Dissemination plans informed by PPI and participant and stakeholder engagement will be developed and reviewed annually. A range of outputs including NIHR final report, peer-reviewed open access publications, conference presentations, policy briefings, non-academic research summaries and a draft scaling-up plan, are anticipated to be disseminated relevant for different audiences, including trial participants, postpartum women throughout the UK, healthcare professionals, service commissioners, academics, the Government and public health bodies.