Study protocol: quantitative fibronectin to help decision-making in women with symptoms of preterm labour (QUIDS) part 2, UK Prospective Cohort Study

Introduction The aim of the QUIDS study is to develop a decision support tool for the management of women with symptoms and signs of preterm labour, based on a validated prognostic model using quantitative fetal fibronectin (fFN) concentration, in combination with clinical risk factors. Methods and analysis The study will evaluate the Rapid fFN 10Q System (Hologic, Marlborough, Massachusetts, USA) which quantifies fFN in a vaginal swab. In QUIDS part 2, we will perform a prospective cohort study in at least eight UK consultant-led maternity units, in women with symptoms of preterm labour at 22+0 to 34+6 weeks gestation to externally validate a prognostic model developed in QUIDS part 1. The effects of quantitative fFN on anxiety will be assessed, and acceptability of the test and prognostic model will be evaluated in a subgroup of women and clinicians (n=30). The sample size is 1600 women (with estimated 96–192 events of preterm delivery within 7 days of testing). Clinicians will be informed of the qualitative fFN result (positive/negative) but be blinded to quantitative fFN result. Research midwives will collect outcome data from the maternal and neonatal clinical records. The final validated prognostic model will be presented as a mobile or web-based application. Ethics and dissemination The study is funded by the National Institute of Healthcare Research Health Technology Assessment (HTA 14/32/01). It has been approved by the West of Scotland Research Ethics Committee (16/WS/0068). Version Protocol V.2, Date 1 November 2016. Trial registration number ISRCTN41598423 and CPMS: 31277.

The overall aim of the QUIDS study is to develop a decision support tool for the 2 management of women with symptoms and signs of preterm labour, based on a 3 validated prognostic model using quantitative fFN testing. The study has been 4 conceptually divided into two parts. In this, the protocol for QUIDS Part Two, we 5 detail the protocol for a prospective cohort study. This will externally validate a 6 prognostic model developed in QUIDS Part One. [1] More detailed background about 7 the diagnosis of preterm labour and background to the study is provided in the 8 introduction of QUIDS Protocol Part One.[1] 9 10 Fetal Fibronectin (fFN) is a biochemical test of preterm labour which has potential to 11 help improve diagnosis of impending preterm delivery. [2] Much of the evidence about 12 fFN to date relates to the qualitative fFN test, which provides a positive or negative 13 result on the basis of a single threshold of 50ng/ml. [2,3] This test has been largely 14 replaced with the Rapid fFN 10Q System, which provides a concentration of fFN 15 (quantitative fFN) and may be a more useful predictor of preterm delivery. fFN is now 16 only available with a quantitative analyser in the UK, but there is no consensus as to 17 which women to use the test in, or how to interpret the results. 18 19 The QUIDS study will address this evidence gap by providing evidence about the 20 potential value of the quantitative fFN test, along with guidance about how to 21 interpret results. Here we detail the protocol for external validation of a prognostic 22 model developed in QUIDS Part One. [ The aim of the QUIDS study is to develop a decision support tool for the 1 management of women with symptoms and signs of preterm labour, based on a 2 validated prognostic model using quantitative fFN testing. 3 4 The study protocol has been divided into two parts (see flow chart Figure 1). The 5 protocols for Parts One and Two are reported in separate manuscripts. 6 7 In QUIDS Protocol Part One we have described how we will perform (i) an Individual 8 Participant Data (IPD) meta-analysis, and (ii) and Economic Analysis. The protocol 9 details how we will develop and internally validate a prognostic model using 10 quantitative fFN and other risk (prognostic) factors and to evaluate the added value 11 of quantitative fFN toward this prognostic model performance. We will also provide 12 an economic rationale for the prognostic model and analyze its cost-effectiveness 13 from the perspective of the NHS. 14 15 In this, the QUIDS Protocol Part Two, we will detail the prospective cohort study to 16 externally validate and, if necessary, refine the prognostic model. This will be 17 performed in at least eight UK hospitals with different settings (rural/urban) and 18 different levels of neonatal care facilities. In addition, acceptability of quantitative fFN 19 testing, and effects on maternal anxiety will be performed. We will assess the 20 potential cost-effectiveness of the final prognostic model/decision support tool. This 21 additional analysis will allow us to model the full costs and effect impacts of the 22 different prognostic model and compare these in a cost-effectiveness analysis to 23 The result is invalid if the test does not meet internal quality controls that are 12 performed automatically with every test. In the event of an invalid result, the test can 13 be repeated with any remaining clinical specimen. A quality control can be performed 14 by a reusable Rapid fFN 10Q QCette® QC Device, which verifies that the analyser 15 performance is within specification. The following exclusion criteria will apply: 2 • Contraindication to vaginal examination (e.g. placenta praevia). 3 • Higher order multiple pregnancy (triplets or more). 4 • Moderate or severe vaginal bleeding. 5 • Cervical dilatation greater than 3cm. 6 • Confirmed rupture of membranes. 7 • Sexual intercourse, vaginal examination or transvaginal ultrasound in the 8 preceding 24 hours factors may invalidate results. These women will be 9 initially excluded from the study, but can be included if still symptomatic after 10 The study will not include any community maternity units (staffed by midwives, with 2 or without involvement of non-obstetric medical staff), which cover a small proportion 3 of women, mainly in remote and rural areas. In the Perinatal Collaborative Transport 4 Study (CoTS study) of perinatal transfers in Scotland, [7] which involved 52,727 5 births, only 69 (0.13%) women were transferred to a consultant-led obstetric unit 6 from community maternity units, and only a proportion of these were for suspected 7 preterm labour. The small number of women cared for in community maternity units 8 means their inclusion would not be an efficient use of study resources. 9 10 Given that management of women with symptoms of preterm labour and inter-11 hospital transfer patterns are likely to vary depending on level of available neonatal 12 care and distance to transfer, we will include a mixture of hospitals with different 13 levels of neonatal care facilities in both rural and urban settings. We will include units 14 with Special Care Units (providing special care for their own local population), Local 15 Neonatal Units (providing special care and high dependency care and a restricted 16 The original consent form will be stored in the Investigator Site File (ISF) file, a copy 2 is given to the woman, a copy added to the medical notes and a copy sent to the 3 Trial Office. 4 5 After providing consent, the participant will be asked to complete a short State Trait 6 Anxiety Inventory (STAI) questionnaire and complete a contact details form. They will 7 also be issued with a letter thanking them for taking part in the trial and giving details 8 of the second questionnaire to be completed. 9 10

Sample Collection 11
Samples for analysis will be taken with a fFN specimen collection kit, which consists 12 of a sterile polyester tipped swab and a specimen transport tube containing 1 ml 13 extraction buffer (an aqueous solution containing protease inhibitors and protein 14 preservatives including aprotinin, bovine serum albumin, and sodium azide). During 15 speculum examination the sterile swab will be lightly rotated across the posterior 16 fornix of the vagina for ten seconds to absorb vaginal secretions. Samples should be 17 taken before any other swabs (e.g. for microbiology) or cervical manipulation and the 18 speculum lubricated with normal saline as other lubricants may interfere with the 19 antibody-antigen reaction of the test. Following specimen collection the swab should 20 be removed, immersed in extraction buffer, the shaft of the swab snapped off, and 21 the transport tube sealed. 22 23 Before analysis samples are gently mixed and as much liquid as possible expressed 24 from the swab by rolling the tip against the inside of the tube.  The Hologic Rapid fFN 10Q analyzer has integrated quality control measures, and 2 we will keep records of these as well as any additional staff training that occurs after 3 the study starts. It is recommended that a daily pre-calibrated reusable quality control 4 cassette be inserted and analysed every 24 hours to verify that the analyser 5 performance is within specification. A daily quality control (QC) should be performed 6 if one has not been done in the preceding 24 hours before a patient test is to be 7 done. Logs of results are stored on the machine and can be downloaded, and we will 8 also ask the participating sites to keep a monthly paper log of QC tests done. Each 9 patient test has an internal quality control, with a procedural control line that verifies 10 the threshold level of signal by the instrument. Sample flow detection ensures the 11 sample travels across the cassette properly, and confirms absence of conjugate 12 aggregation. We believe that these measures will help ensure the validity of results. 13 However, to provide further evidence of integrity and comparability of results from 14 each site we will request that all participating sites enrol in the Wales External Quality 15 Assurance Scheme (WEQAS) Point of Care Quality Assurance Scheme. WEQAS will 16 provide a sample for analysis to each site bimonthly, and provide reports on analyser 17 performance and variability. [8] 18 19 Data Collection 20
However, accurate prediction of preterm birth is challenging, even when the clinical symptoms are suggestive of preterm labour. In randomised trials approximately 80% of women diagnosed with preterm labour remained pregnant after 7 days. 6,7 Interventions in preterm labour and preparations for preterm birth may include administration of corticosteroids to accelerate fetal lung maturation 8,9 and magnesium sulphate for fetal neuroprotection, 10 in utero transfer to a facility with appropriate maternity and neonatal services, and tocolysis to optimise time before birth to enable these. 11 Whilst such interventions can improve outcomes for mothers and babies who do experience preterm birth, they are not necessarily benign, especially for those in whom preterm birth does not occur.
The maximal beneficial impact of corticosteroids occurs with administration between 48 hours and seven days before birth, thus timing is especially important in optimising benefit for the neonate. For women who remain at risk of preterm birth after seven days of the initial dose, repeated doses reduce respiratory distress in the neonate 9 but have been found to be associated with a dosedependent reduction in birthweight. 12,13 A five-year follow-up study of women who received repeated doses of antenatal corticosteroids due to risk of preterm birth found an increased risk of neurodevelopment impairment in infants born at term. 14 Therefore developing a strategy to establish the optimal time to give steroids is a research priority.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Magnesium sulphate administration immediately prior to birth has been shown to reduce cerebral palsy, 10 but there is a risk of magnesium toxicity leading to respiratory depression in the mother and, theoretically, the neonate. 15 Whilst there is no clear beneficial effect of tocolytics on the incidence or outcome of preterm birth, 16 their use is recommended if the days gained prior to preterm birth can be used appropriately, for example transfer to a suitable maternity unit or the administration of drugs to protect the neonate. 11 Tocolysis is linked with various maternal and neonatal complications, 17 hence the need for therapy targeted only for those at risk of preterm birth and close monitoring of the mother and fetus throughout.
Often, inpatient admission is recommended if preterm labour is suspected. Previous literature has highlighted the social isolation and support needs that women with high-risk pregnancies who are hospitalised experience. 18 In some cases, in-utero transfer is indicated to ensure that birth takes place in a specialist unit with appropriate neonatal care facilities. This policy has been shown to reduce mortality 19,20 and morbidity 21 in preterm neonates, especially those born very premature.
Qualitative research has indicated that women generally acknowledge the potential benefit of in utero transfer to their baby and, hence, are willing to endure the inconvenience and upheaval that it entails. 22,23 However, the experience is associated with an emotional, social and financial burden on women and their families, especially for the substantial proportion of women who do not deliver prematurely following in utero transfer. When describing their experiences of in utero transfer, women expressed shock at the prospect of the transfer, feeling socially isolated, and having no control over the situation, in addition to the practical difficulties experienced particularly by women who already had children. 22,24,25 In a large survey of women who had experienced in utero transfer, over a quarter lamented the financial cost 24 particularly with respect to their partner's outlay for travel, food, accommodation, and phone bills, exacerbated with requiring time off work. 22 Furthermore, in utero transfer is costly to maternity services. Securing a maternal and neonatal bed  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  in another unit is a time-consuming task that often falls to delivery suite midwives to arrange, whilst also continuing to provide care to the woman. 26 In a large observational study of all in utero transfers that took place in Scotland in a six-month period, nearly one third of all transfers were due to threatened preterm labour. 27 Under half of the women transferred from one consultant-led unit to another gave birth within 48 hours. 27 Such unnecessary transfers are costly to women, their families and maternity services. Qualitative research into women's experiences of preterm labour have highlighted the need for caregivers to create an environment where women are enabled to discuss their fears 28 and exert control over how they manage their preterm labour care. 25 Accurate prediction of preterm birth could reduce the burdens and risks associated with unnecessary interventions, and enable women and their clinicians to make informed decisions regarding their care. Numerous diagnostic tests have been used in preterm labour, including biochemical tests of vaginal secretions and cervical length. 29 One such test is fetal fibronectin, a near-bedside test that provides a positive or negative result and has excellent negative predictive value. 30 Thus fetal fibronectin can identify which women will not benefit and may be put at risk by the interventions described previously, and reduce costs to maternity services. 31 Developments in fetal fibronectin testing have led to a quantitative test that provides a concentration of fetal fibronectin in vaginal secretions, giving women and clinicians more information on which to base their management decisions. 32 Qualitative evidence has indicated that women feel a sense of increased responsibility to their babies and themselves during a high risk pregnancy, such as threatened preterm labour. 33 Women want to be involved in decision making about their care to different degrees and feel most satisfied when their caregiver supports them to make decisions in the way they felt most comfortable. 33
Funding has been received from the National Institute for Health Research Health Technology Assessment Programme for a large, multicentre trial to develop a mobile application decision support tool for the management of women with symptoms and signs of preterm labour, based on a validated model using quantitative fetal fibronectin testing. This study is the precursor to that trial, with the aim of determining the decisional needs of pregnant women with the symptoms and signs of preterm labour, their families and caregivers, using a qualitative framework approach. The outcomes of this qualitative study will inform the development of the mobile application decision support tool, using the findings from an individual patient data meta-analysis. The tool will then be externally validated and refined in the multi-centre trial, QUIDS.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Methods A qualitative framework approach will be used, based on data collected from focus groups and semistructured telephone interviews.

Setting
Focus groups will take place in three maternity units: Liverpool Women's NHS Foundation Trust, Birmingham Women's NHS Foundation Trust and Royal Edinburgh Hospital, NHS Lothian. There will be focus groups for women and a separate focus group for partners. Clinicians who care for women with threatened preterm birth will be interviewed by telephone.

Sample
A purposive sample of women and partners will be recruited to cover a variety of experiences of preterm labour and birth. Women will be stratified by their prior experience and relevant characteristics, including ethnicity, previous obstetric history, living in an urban or rural setting and proximity to a tertiary neonatal referral centre. Two focus groups of 4-8 women will be conducted at each site; one for pregnant women who are at high risk of preterm birth, and one for postnatal women who have recently experienced preterm birth. One partners' focus group will be conducted at one of the sites. If women or partners are unable to attend a focus group but still wish to participate, a semi-structured telephone interview will be offered.

Principal inclusion criteria for women's antenatal focus groups
Women who are currently pregnant who: • Have previously experienced preterm birth following preterm labour, • Have experienced threatened preterm labour in this pregnancy, • Are at high risk of preterm birth for another clinical reason, such as prior cervical surgery.

Principal inclusion criteria for women's postnatal focus groups
Women who have experienced preterm birth following preterm labour at <34 weeks whose babies are stable and well and are receiving care on the special care baby unit or neonatal intensive care unit.

Principal inclusion criteria for partners' focus groups
Partners of women who fit the eligibility criteria for either focus group.

Principal exclusion criteria for the focus groups
Non-English speaking individuals.

Women and partners
Eligible women will be identified by clinicians in the preterm birth clinic and other antenatal clinics, and antenatal, triage or labour wards (for the antenatal focus groups) and the special care baby unit or postnatal clinics (for the postnatal focus groups) at each site. Eligible partners will be identified by the same method. Clinicians who are aware of and understand the research aims will approach women and partners to request consent for a researcher to contact them. Importantly, only postnatal parents whose babies are being cared for on the SCBU who are considered stable and well by the clinicians will be approached. With consent the researcher will make contact to talk to the women and/or their partners about the research, either face-to-face or over the telephone. Potential participants will be given the participant information sheet (PIS) (appendix _) that is relevant to them and given verbal information about the study. Each participant will be given time to read the information and the opportunity to have any questions answered. Willing participants will be asked to provide their written consent prior to the focus groups.

Clinicians
Eligible clinicians will be approached by the researchers, via email or face-to-face. Clinicians will be given the clinician PIS (appendix _) and the opportunity to read the information and have any questions answered. Willing clinicians will be asked to provide their written consent prior to the interviews.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   V 1.3  21/10/15 All participants (women, partners and clinicians) will be reassured that they are not compelled to participate, that they can withdraw from the study at any time, and that non-participation will not affect their care or employment in any way.

Data collection
The primary aim of this research is to determine the decisional requirements of women, their partners and clinicians for the management of preterm labour. Qualitative semi-structured interviews, in a focus-group setting or individual telephone interviews, provide a means of collecting rich, in-depth data with a specific focus. 37 Hence, structured topic guides will be used to initiate and concentrate the discussion (appendices 7-10).
Focus groups are the preferred format for eliciting the view of women and women's partners.
Encouraging discussion among a homogenous group with a shared interest is likely to provide rich insight and understanding into the group's experiences, beliefs and norms as a result of their social interaction. 38 Conversely, interviewing clinicians individually avoids the potential pitfall of professional embarrassment stifling ideas in a group setting. Interviewing individual clinicians with a range of professional experience should ensure that the decisional requirements of clinicians at all levels of experience are understood.
Demographic details and baseline characteristics will be collected prior to the interviews, either as a self-completion questionnaire, or questions asked by the researcher over the telephone. All interviews will be audio recorded, with the participants' consent, and field notes taken. The focus groups will be facilitated by at least two researchers. This is to ensure that all pre-specified areas of interest are covered and that non-verbal communication and group interactions are documented within the field-notes, which will provide context for the data analysis. Recapping will be used to  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59

Analysis plan
A framework approach to data analysis will be used. This approach was developed to manage and interpret large volumes of data collected to inform health policy, meaning they had focussed aims and objectives. 37 Likewise, this research has clear aims, as described previously, in addition to the methodological aim of collecting rich data about the experiences and beliefs of women, their partners and clinicians in relation to managing preterm labour.
Framework analysis follows specific, clearly documented stages of analysis that are transparent so that others can review the interpretation processes and understand how the findings were reached. 39 Transparency is particularly important in this study as the findings will inform the development of an application to aid management decisions in clinical practice. Following verbatim transcription of the interview recordings, the researchers will become familiar with the data by reading the transcripts and field-notes several times. The next stage is to develop a theoretical framework by re-reading the transcripts and making notes as recurring characteristics are  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   V 1.3  21/10/15 recognised. The characteristics will then be collated into themes, which are based on the text itself, supported by the field-notes. The resulting thematic framework will be applied back to the transcripts and field-notes to check that it reflects the context of the original data. The transcripts will be coded, so that portions of text are linked to a discrete theme. A sample of transcripts will be independently coded by two people. The data will be charted and indexed to identify the preterm labour or professional experience of the participant, thus enabling the attribution of themes to a particular group. Finally, the content of the charts will be interpreted and mapped against each other to devise themes and sub-themes categories. Once again, this will involve review of the original data. Explanatory accounts will be developed to clarify the data and quotable sections of data will be identified. The final categories will be discussed between the researchers until consensus is met. The researchers will maintain reflexive journals throughout the data collection and analysis stages, recognising and ameliorating, as far as possible, the fact that their presence and assumptions impact on the data and the findings. 40 This method of data analysis creates a clear audit trail thus ensuring rigour. Each stage of analysis refers back to the original data so that context and meaning is not lost in the final framework of themes and subthemes. The data analysis process will be managed using NVivo software, a qualitative data analysis tool.

Safety
The physical safety of participants will be ensured through adhering to the health and safety policies of the host units where the focus groups take place.
The emotional wellbeing of the participants will be safeguarded by following the Distress Policy (see appendix 11). The Supervisors of Midwives (SOM) team in each unit will be informed of the study and women and their partners will be given the SOM team contact details, should they become distressed or upset as a result of talking about their experiences. Participants will also be given the contact details for accessing local counselling services.

Informed consent
All participants will be fully informed about the study and the subsequent QUIDS trial via verbal and written communication. All eligible individuals will be given the participant information sheet (appendix __) and provided with an opportunity to have any questions answered. Written consent will then be gained prior to the commencement of the focus groups/interviews.

Confidentiality
Demographic information will be collected from participants to attribute themes from the data to particular groups within the analysis and dissemination of findings. Demographic information, which will contain potentially identifiable information, will be kept in a secure lockable cabinet. Audio recordings will be stored on an encryptable audio device only until they are transcribed. Once transcribed the audio recordings will be deleted. Transcription services are provided by '1 st Class Secretarial', who subscribe to the Data Protection Act and have also signed the Code of Practice on Data Handling. Hard copies of audio transcripts and field-notes will be kept in a separate secure  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   V 1.3  21/10/15 lockable cabinet to the demographic information. The transcripts and field-notes will be coded to identify which participant provided that data; the codes will only be known by the researchers.
Participant's data will not be used for any purpose other than this study and the subsequent QUIDS trial.

Data Protection
Participants will be informed that publications from this study will contain direct quotes from the focus groups/interviews and categorisation of their experience of preterm labour (e.g. experienced preterm birth), which could enable personal identification.
All researchers involved in this study must comply with the requirements of the Data Protection Act 1998 with regard to the collection, storage, processing and disclosure of personal information and uphold the Act's core principles. All computers used for processing data are password protected and subject to the strict data protection policies of the researcher's institution.

Good clinical practice training
All researchers involved in this study must hold evidence of recent Good Clinical Practice training.

Safety of researchers
An individualised risk assessment will be conducted to identify any risks to researchers or participants involved in this study. The lone working policy of the institution will be adhered to at all times. The only anticipated lone working will be during travel to and from the interview sites.
The lone working policy of the researcher's institutions mandates that researchers wear a GPS tracking and audio transmitting device during all lone-working, off-site research activity with participants. Participants will be informed if this device is being used.

Insurance / Indemnity
The researcher's institution holds public liability insurance and professional indemnity insurance (appendices 12, 13 and 14).

Aims and Methodologies
The aim of the QUIDS study is to develop a decision support tool for the management of women with symptoms and signs of preterm labour, based on a validated prognostic model using quantitative fFN testing.
The study protocol has been divided into two parts (see flow chart Figure 1). The protocols for Parts One and Two are reported in separate manuscripts.

Endpoints
The primary endpoint is spontaneous preterm delivery within seven days of qfFN test, in women tested at less than 36 weeks gestation. This is both an important endpoint for women and caregivers (determined in QUIDS Qualitative study -a preceding qualitative study to identify the decisional needs of women, their partners and clinicians; Supplementary Material) as well as a clinically important endpoint.
Antenatal steroids (which significantly reduce morbidity and mortality in preterm babies [17]) are most effective if delivery occurs within seven days of administration.
As repeated doses of antenatal steroids may be harmful, it is crucial to ensure steroids are timed correctly. This analysis will be performed if feasible to do so within the constraints of the data available for model development.

Health technologies being assessed
The study will evaluate the Rapid fFN 10Q System (Hologic), which provides a

Target population
The target population is pregnant women attending hospital with signs and symptoms of preterm labour.

Inclusion Criteria
We prespecified inclusion of prospective cohort studies or RCTs of women with signs and symptom of preterm labour (as defined by investigators) that include quantitative fFN results determined by 10Q rapid fFN analyzer system and pregnancy outcome data; and the Principal Investigator of which has agreed to collaborate and provide data.

Exclusion criteria
We will exclude studies where fFN concentration was measured by ELISA and studies where IPD is not available for meta-analysis

Search Strategy
When applying for funding for this study (April 2014)  https://www.google.co.uk) and systematic reviews. We also consulted preterm birth Study manuscripts and/or protocols were screened by two researchers. We identified a total of 10 studies of quantitative fFN that were potentially eligible. Four early datasets (in three manuscripts) used ELISA to determine the concentration of fFN and were excluded as the different method of analysis and earlier period of study would increase heterogeneity. [5,19,20] Therefore, six studies fulfilled the eligibility criteria (see Table 1).

Establishment of the quantitative fFN IPD Collaboration
We contacted the principal investigators (PIs) of the six eligible studies of qfFN invited them to participate (see Table 1). Five of these agreed to provide their IPD as evidenced by their involvement as co-applicants on the funding application and/or coauthorship of this protocol (Mol, van Baaren, Khalil, Shennan, David). The PI of the 6 th study (Elovitz) indicated IPD may be available after publication of her study.
The five included studies (Table 1)  and Europe (two studies). All women in the included trials provided informed consent for participation in clinical trials, and for their IPD to be used in subsequent analyses.  Edinburgh. PIs will be asked to provide de-identified data, and consider all recorded 3 variables (even if not reported publications). We will assess study quality according 4 to QUADAS-2 [25] QUIPS [26] and CHARMS [27] guidelines. 5 6 Sample Size Considerations 7 The size of the IPD meta-analysis is limited by the number of studies with data 8 available (Table 1). In model development the number of covariates that can be 9 considered is limited by the number of events, with guidance suggesting at least ten 10 events required for each covariate. [28,29] In our IPD meta-analysis data we have 11 139 events (preterm labour within 7 days of testing) and therefore deemed that it was 12 sensible to evaluate quantitative fFN and up to 13 other factors (covariates) for 13 potential inclusion in our model. 14 15

Data Items 16
The following factors which are thought to influence risk of spontaneous preterm 17 birth, will be requested and considered for inclusion as covariates in the prognostic 18 Prior to analysis data will be checked for outliers and missing data will be identified. 1 Descriptive statistics will be performed to summarise data. Problems identified will be 2 discussed with the PI of the original study, and amended as indicated by consensus 3 discussion. 4 5

Data Analysis and Prognostic Model Development 6
Multivariable logistic regression modelling will be the primary method of analysis. The 7 primary endpoint for the prognostic model will be delivery within seven days. Another 8 endpoint found to be important in focus group consultations performed in QUIDS 9 Qualitative (Supplementary Material) included delivery within 48 hours, and we will 10 use this as a secondary endpoint if feasible (i.e. if sufficient number of cases with 11 delivery within 48 hours). We will develop an initial model with quantitative fFN 12 concentration, and then consider a model with other predefined clinical predictor 13 variables (see Data Items, above). 14 15 Tocolysis (which may delay onset of labour, although likely not beyond 48 hours) will 16 be included as a categorical variable (administered/not administered). We will 17 explore treatment effect by sensitivity analysis with and without the assumption that 18 tocolysis could delay delivery within 48 hours by a maximum odds ratio of 5.39, 95% 19 credible interval 2.14 to 12.34, based on data in Haas et al. [30]. 20 21 As the outcome is binary, a logistic regression modelling framework will be used to 22 develop the model. A multi-level structure will be used to account for clustering of 23 patients within studies, and heterogeneity of the effects of included factors (hereafter 24 called 'predictors') will be accounted for using random-effects, with between-study 25 heterogeneity quantified using the estimated variance ('tau-squared') and the I 2 26 statistic. A separate intercept term per study will be included in the model, to account 27 for the clustering and also guage how predictions may require tailoring to different 28 populations. Predictors with large heterogeneity in the prognostic effect across 29 In the primary analysis, we will use data from the first recorded attendance with signs 4 and symptoms of preterm labour to determine the relationship between that individual 5 episode and outcome. Data from subsequent attendances will be analysed 6 subsequently, and may be included in an appropriate model. As a parsiminous model 7 is sought, to reduce the factors included in the model that may otherwise delay its 8 use, we will use backward stepwise selection based on an information criterion (e.g. 9 Akaike's information criterion p<0.15) to identify a parsimonious set of factors to be 10 included in the model; hereafter these are referred to as included 'predictors'. 11 Further, an approach of adding specialist tests, such as cervical length, only after 12 considering simpler clinical assessment will be used, to maximise the utility of the 13 model by ensuring that extra tests with their additional costs are only be included if 14 they add to the predictive power. 15 16 Linearity between continuous variables and outcome will be assessed using cubic 17 spline plots and data will transformed where appropriate before inclusion in 18 multivariable analysis (e.g. using fractional polynomial methods). Missing data will be 19 assessed to determine whether missing at random is appropriate, and if so, multiple 20 imputation of observed participant characteristics will be used, with missing data 21 imputed within each original study separately, before the meta-analysis. The results 22 of these analyses will be compared with a complete case analysis. 23 24 Assessing Apparent Model Performance 25 The apparent performance of the model will be assessed by its overall fit, and the 26 observed discrimination and calibration in the IPD used to develop the modle. Overall 27 fit of the models will be expressed with Nagelkerke R 2 . The ability of the models to 28 discriminate between women with and without spontaneous preterm birth will be 29  16 determined by the area under the receiver operating characteristics curve (AUC), 1 also known as the C statistic. Agreement between predicted and observed 2 proportions of women with spontaneous preterm birth will be visualized using a 3 calibration plot, and measured using calibration slope and calibration-in-the-large. 4

5
Internal validation: assessing Optimism In Model Performance 6 Apparent performance is likely to be optimistic, as it is examined in the same data 7 used for model development. Therefore internal validation will also be undertaken 8 using a non-parametric bootstrap re-sampling technique in which each modelling 9 step is repeated in each bootstrap sample, to obtain a new model in each bootstrap 10 sample, and then its apparent performance (AUC and calibration slope) in the 11 bootstrap sample is compared to its performance in the original dataset. The 12 'optimism' is the mean difference (across all bootstrap samples) between the 13 apparent value in the bootstrap sample and the observed value in the original 14 dataset. This optimism estimate is then subtracted from the original model's apparent 15 performance, to give an optimism-adjusted estimate of each measure of performance 16 for the original model (e.g. R 2 , C statistic, Calibration slope). 17 18

Production Of Final Model From IPD Meta-Analysis Via Uniform Shrinkage 19
The optimism-adjusted calibration slope will be used as a uniform shrinkage factor, to 20 adjust the parameter estimates (log odds ratios) of the original model. The beta 21 coefficients in the original model will be multiplied by the shrinkage factor, and the 22 study intercept terms re-estimated to ensure perfect overall calibration is maintained 23 (across all studies and, ideally, in each study separately). This will thereby produce a 24 final model containing the updated intercepts and the shrunken beta coefficients. [ 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  The added value of quantitative fFN will be examined throughout the whole model 5 process, in particular its improvement on discrimination, calibration and other 6 meaningful factors (such as clinical decisions) using appropriate techniques (such as 7 net reclassification improvement and decision analysis methods). 8 9

Subgroup analyses 10
Subgroup analysis will be performed for multiple pregnancy, women with a previous 11 preterm labour, gestation and those with criteria that are suggested to indicate 12 preterm labour (number of uterine contractions in a set time period and/or cervical 13 change). This will allow us to do a subgroup-analysis in which we assess whether the 14 predictive capacity of quantitative fFN is similar in all subgroups. 15 16 Health Economic Analysis 17 An early stage decision-analytic model will be built using evidence from current 18 literature and from the IPD meta-analysis to explore the potential cost-effectiveness 19 of different prognostic models including quantitative fFN. 20 A literature review will be undertaken to inform model design and identify additional 21 model parameters with searches of Medline, Embase, Cochrane Library and the 22 Paediatric Economic Database Evaluation for economic analyses including the use of 23 fFN testing in woman with threatened preterm labour. Any evidence on resource use 24 (test administration, treatments for preterm labour, hospital stay, hospital transfers, 25 etc), quality of life and diagnostic outcome data from the IPD meta-analysis will be 26 synthesized with the wider evidence based on current practice for women attending 27 hospital with signs and symptoms of preterm labour. The economic analysis will be 28 undertaken from the perspective of the UK NHS adhering to good practice guidelines 29  [32] A decision tree will be developed to model the 1 clinical pathway. The model will be used to explore potential cost effectiveness of 2 the prognostic model at different thresholds on the Receiver Operator Curve, 3 providing an economic rationale for the chosen prognostic model. 4

ETHICS AND DISSEMINATION 6
Trial Management And Oversight Arrangements 7 Project Management Group 8 The trial will be coordinated by a Project Management Group (PMG), consisting of 9 the grant holders (Chief Investigator and Co-applicants), the trial manager, 10 representatives from the Study Office and CHaRT (the supporting CTU), plus service 11 user representatives (PAG). The PMG will meet approximately every four months by 12 teleconference or face to face. The study will be conducted in accordance with the principles of Good Clinical 22 Practice (GCP). A favorable ethical opinion has been obtained from the appropriate 23 REC (reference 16/WS/0068) and local R&D approval will be obtained prior to 24 commencement of the study at each site. 25 26

Dissemination 27
On completion of the study, the study data will be analysed and tabulated, and a 28 clinical study report will be prepared. Results will be communicated to the academic 29  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  presentations. The TRIPOD reporting guidelines will be adhered to. [18] Summaries 2 of results will also be made available to investigators for dissemination within clinics. 3 Social media will be used to signpost publications and conference presentations and 4 highlight important findings. Twitter and Facebook will be used to disseminate 5 findings to professional organizations, charities, stakeholders and the public. 6 Communication to the general public will further be facilitated by our close links with 7 charities such as Tommy's [33].    1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59      Methods and analysis: The study will evaluate the Rapid fFN 10Q System (Hologic, 6 Malborough, MA) which quantifies fFN in a vaginal swab. In QUIDS Part 2 we will 7 perform a prospective cohort study in at least eight UK consultant-led maternity units, 8 in women with symptoms of preterm labour at 22+0 to 34+6 weeks gestation to 9 externally validate a prognostic model developed in QUIDS Part 1. The effects of 10 quantitative fFN on anxiety will be assessed, and acceptability of the test and 11 prognostic model will be evaluated in a subgroup of women and clinicians (n=30). 12 The sample size is 1600 women (with estimated 96-192 events of preterm delivery 13 within 7 days of testing). Clinicians will be informed of the qualitative fFN result 14 (positive/negative) but be blinded to quantitative fFN result. Research midwives will 15 collect outcome data from the maternal and neonatal clinical records. The final 16 validated prognostic model will be presented as a mobile or web-based application. 17 Ethics and dissemination: The study is funded by the National Institute of 18 The overall aim of the QUIDS study is to develop a decision support tool for the 2 management of women with symptoms and signs of preterm labour, based on a 3 validated prognostic model using quantitative fFN testing. The study has been 4 conceptually divided into two parts. In this, the protocol for QUIDS Part Two, we 5 detail the protocol for a prospective cohort study. This will externally validate a 6 prognostic model developed in QUIDS Part One. [1] More detailed background about 7 the diagnosis of preterm labour and background to the study is provided in the 8 introduction of QUIDS Protocol Part One. [ The aim of the QUIDS study is to develop a decision support tool for the 1 management of women with symptoms and signs of preterm labour, based on a 2 validated prognostic model using quantitative fFN testing. 3 4 The study protocol has been divided into two parts (see flow chart Figure 1). The 5 protocols for Parts One and Two are reported in separate manuscripts. 6 7 In QUIDS Protocol Part One we have described how we will perform (i) an Individual 8 Participant Data (IPD) meta-analysis, and (ii) and Economic Analysis. The protocol 9 details how we will develop and internally validate a prognostic model using 10 quantitative fFN (as a continuous variable) and other risk (prognostic) factors and to 11 evaluate the added value of quantitative fFN toward this prognostic model 12 performance. We will also provide an economic rationale for the prognostic model 13 and analyze its cost-effectiveness from the perspective of the NHS. 14 15 In this, the QUIDS Protocol Part Two, we will detail the prospective cohort study to 16 externally validate and, if necessary, refine the prognostic model. This will be 17 performed in at least eight UK hospitals with different settings (rural/urban) and 18 different levels of neonatal care facilities. In addition, acceptability of quantitative fFN 19 testing, and effects on maternal anxiety will be performed. We will assess the 20 potential cost-effectiveness of the final prognostic model/decision support tool. This  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  (antenatal steroids, tocolysis or magnesium sulphate) is being considered due 10 to signs of pre-term labour. 11 • Women aged 16 years or above. 12 The broad inclusion criteria reflect current clinical practice and enable the 13 generalisability of the results of the trial for routine clinical care. We will include 14 women who re-attend seven days or more after initial recruitment with signs and 15 symptoms of preterm labour and also women who remain symptomatic but 16 undelivered seven days later in whom repeat testing by the clinician is deemed to be 17 appropriate. This will be in line with manufacturer's recommendation for fFN testing. 18 19 The following inclusion criteria will apply on speculum examination: 20 • Cervical dilation ≤ 3cm 21 • Intact membranes 22 • No significant vaginal bleeding, as judged by the clinician. 23 • Once it has been established that the women meets the above criteria, on 24 speculum examination, the fFN swab can be taken. 25 Participants that sign the consent but are not eligible upon examination to have an 26 fFN swab taken will still be enrolled and have outcome data collected. 27  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  The following exclusion criteria will apply: 2 • Contraindication to vaginal examination (e.g. placenta praevia). 3 • Higher order multiple pregnancy (triplets or more). 4 • Moderate or severe vaginal bleeding. 5 • Cervical dilatation greater than 3cm. 6 • Confirmed rupture of membranes. 7 • Sexual intercourse, vaginal examination or transvaginal ultrasound in the 8 preceding 24 hours factors may invalidate results. These women will be 9 initially excluded from the study, but can be included if still symptomatic after 10 24 hours, when fFN accuracy will be restored. 11 12

Co-Enrolment 13
This trial involves validating a decision support tool relating to a test that is currently 14 commonly used in clinical practice. As such, there are no additional interventions. 15 Co-enrolment in other non-interventional trials will be allowed. Co-enrolment in trials 16 of tocolytic treatments or other management strategies that may influence timing of 17 delivery as a primary outcome will not be allowed. Participation in QUIDs would not 18 preclude babies being subsequently involved in interventional trials. Co-enrolment 19 will be recorded in the electronic case report form (eCRF). 20

Setting 22
The prospective cohort study will take place in at least eight consultant-led obstetric 23 units in the UK. More than 93% of pregnant women in the UK deliver in consultant-24 led units. [5,6] The vast majority of women with symptoms of preterm labour will 25 present to a consultant-led unit for assessment, either directly or following advice 26 from their community midwife or General Practitioner. 27  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  The study will not include any community maternity units (staffed by midwives, with 2 or without involvement of non-obstetric medical staff), which cover a small proportion 3 of women, mainly in remote and rural areas. In the Perinatal Collaborative Transport 4 Study (CoTS study) of perinatal transfers in Scotland, [7] which involved 52,727 5 births, only 69 (0.13%) women were transferred to a consultant-led obstetric unit 6 from community maternity units, and only a proportion of these were for suspected 7 preterm labour. The small number of women cared for in community maternity units 8 means their inclusion would not be an efficient use of study resources.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Women with signs and symptoms of preterm labour will be identified on presentation 2 to obstetric services. A member of clinical staff, usually the doctor or midwife 3 assessing the woman, will identify potentially eligible participants, provide a 4 participant information leaflet and invite consent. A suitably trained member of clinical 5 staff (doctor or midwife) or research team will consent participants. 6 7 Posters and leaflets will be situated in antenatal areas of participating hospitals to 8 alert women that the study is taking place, and women will be allowed as much time 9 as possible to consider participation without unduly delaying further clinical 10 assessment. Participants will receive adequate oral and written information and 11 appropriate participant information and informed consent forms will be provided.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   13 Certain exclusion criteria can only be assessed at speculum examination (for 1 example vaginal bleeding or evidence of ruptured membranes), so a proportion of 2 women will not be eligible for fFN testing after consent is given. These women will 3 still be enrolled and delivery outcomes collected. The decision whether to use this 4 data for analysis will be the decision of the Chief Investigator and Statisticians. 5 6 Withdrawal Of Study Participants 7 Women will be able to withdraw consent for us of their data at any time until the end 8 of the study. 9 10 Study Assessments (See Table 1) 11

Eligibility Assessment (Screening And Recruitment) 12
Women presenting with signs and symptoms of pre-term labour will be identified on 13 presentation to obstetric services. The doctor or midwife assessing the woman will 14 identify potentially eligible participants and provide an invitation letter and short 15 information leaflet. 16 17 After the woman has had the opportunity to consider whether she would like to 18 participate, she will be asked to complete the Screening and Consent Form. The 19 clinician will then decide whether the fFN test can be carried out. If the test can be 20 carried out (according to manufacturer's guidelines), then the participant will be fully 21 enrolled and that their delivery outcomes will still be collected. 22 23 If the woman declines to participate and she is willing to provide a reason for this, the 24 reason given will be entered on to an anonymous log. Baseline demographics will be 25 collected on consenting women, together with height and weight, information on 26 medical history, obstetric history, estimated date of delivery and presenting signs and 27 symptoms. 28  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  The original consent form will be stored in the Investigator Site File (ISF) file, a copy 2 is given to the woman, a copy added to the medical notes and a copy sent to the 3 Trial Office. 4

5
After providing consent, the participant will be asked to complete a short State Trait 6 Anxiety Inventory (STAI) questionnaire and complete a contact details form. They will 7 also be issued with a letter thanking them for taking part in the trial and giving details 8 of the second questionnaire to be completed. 9 10

Sample Collection 11
Samples for analysis will be taken with a fFN specimen collection kit, which consists 12 of a sterile polyester tipped swab and a specimen transport tube containing 1 ml  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  The sample taken will be run at a near bedside Hologic Rapid fFN 10Q analyser, 1 specially adapted for the QUIDS study. As fFN (or other similar biochemical tests of 2 preterm labour) are part of standard care, it would be unethical to blind clinicians 3 from the qualitative fFN result. The analyser will thus reveal a qualitative fFN result 4 (positive/negative/invalid) for clinicians to base clinical decision-making on, according 5 to local protocols. The quantitative fFN result however, will be stored as a three-letter 6 code, blinding caregivers from the result. Samples will be run as per manufacturers 7 instructions (described above in the section "Health technologies being assessed"). 8 9

Repeat fFN Tests 10
If there is clinical indication for further fFN tests (eg because of ongoing symptoms of 11 preterm labour after seven days), the results will also be recorded. 12 13

Labour/Delivery/ Neonatal Assessments 14
Admission for delivery will not be a formal study visit but data will be collected using 15 information recorded in the participant's notes. Delivery data will be collected on the 16 maternal outcomes of delivery, including method of delivery, indication for delivery 17 method, onset of labour, date and gestation of delivery and blood loss. 18

19
Questionnaires 20 All participants who are eligible to participate will be asked to complete a STAI 21 questionnaire before the speculum examination. The same questionnaire will be 22 repeated 24-48 hours post examination. The second questionnaire will be provided 23 on paper with a pre-paid envelope to be returned by post to the Trial Office.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  The Hologic Rapid fFN 10Q analyzer has integrated quality control measures, and 2 we will keep records of these as well as any additional staff training that occurs after 3 the study starts. It is recommended that a daily pre-calibrated reusable quality control 4 cassette be inserted and analysed every 24 hours to verify that the analyser 5 performance is within specification. A daily quality control (QC) should be performed 6 if one has not been done in the preceding 24 hours before a patient test is to be 7 done. Logs of results are stored on the machine and can be downloaded, and we will 8 also ask the participating sites to keep a monthly paper log of QC tests done. Each 9 patient test has an internal quality control, with a procedural control line that verifies 10 the threshold level of signal by the instrument. Sample flow detection ensures the 11 sample travels across the cassette properly, and confirms absence of conjugate 12 aggregation. We believe that these measures will help ensure the validity of results. 13 However, to provide further evidence of integrity and comparability of results from 14 each site we will request that all participating sites enrol in the Wales External Quality 15 Assurance Scheme (WEQAS) Point of Care Quality Assurance Scheme. WEQAS will 16 provide a sample for analysis to each site bimonthly, and provide reports on analyser 17 performance and variability. [ Screening data and data about quantitative fFN testing will be collected on paper 8 based CRFs and research midwives will input these into the web based electronic 9 database. Clinical outcome data will be collected from the medical records. 10 11

Maternal Acceptability and Anxiety 12
Maternal anxiety will be measured pre and post-test (24-48h) using the validated 13 State Trait Anxiety Inventory (STAI) questionnaire. Acceptability of fFN testing and 14 the decision support will be assessed using follow up interviews (face to face or 15 telephone, according to maternal preference) which will be conducted with a sub-16 group of participants (n=30) purposively sampled and stratified according to 17 geographical location, outcome (preterm labour or not) and anxiety scores. 18 Acceptability will also be assessed in a cohort of clinicians (n=30). 19 20

Statistics and Sample Size Calculation 21
Guidance for external validation suggests at least ten events (preterm delivery within 22 seven days of test) are required for each covariate included in a prognostic 23 model. [9,10] Data from the cohorts included in our IPD meta-analysis suggests an 24 event rate of between 6 and 12%. [1] Based on these estimates a sample size of 25 1,600 will provide 96 and 192 events (preterm delivery within 7 days). 26 27 A UK study has shown that 8.9% of pregnant women present with symptoms of 28 preterm labour and are eligible for quantitative fFN [11] and we anticipate 50% 29 recruitment rate is achievable, thus overall 4.5% of maternities could be recruited. 1 We will initially include eight units in the cohort study with a combined delivery rate of 2 approximately 36,000 per annum. We anticipate that we will achieve target 3 recruitment within 12 months (1 year * 36,000 * 0.089 * 0.5 = 1,602). If however, the 4 recruitment rate or event rate is lower than predicted, we will increase the number of 5 sites included in the study and/or the recruitment period, to ensure that a minimum of 6 60 events (preterm delivery within 7 days of test) are achieved, allowing for external 7 validation of at least six covariates in our model. 8 9 It is possible that the IPD meta-analysis will find there is potential added value of 10 combining quantitative fFN testing with cervical length measurement. [12,13] As 11 cervical length measurement has significant resource requirement (estimated NHS 12 cost £68.16 per test) and lack of out of hours provision further limits availability in 13 many NHS hospitals, we think it is very unlikely that cervical length scanning will 14 improve performance of the prognostic model to such a degree as to make it cost 15 effective. We will assess the incremental costs and effects of cervical length 16 measurement in the proposed health economic model performed in parallel with the 17 IPD meta-analysis, and will feed into design considerations during the first iteration of 18 the prognostic model. 19 20 If inclusion of cervical length ultrasound is found to be potentially cost-effective, we 21 will assess the feasibility of including it in the prospective cohort study. We anticipate 22 that including cervical length measurement in the prospective cohort study would be 23 extremely difficult in the current NHS setting as the majority of units do not have 24 24 hour availability of transvaginal ultrasound and/or trained personnel to perform scans.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59

Validation Of Prognostic Model 2
The prognostic model developed in the IPD will be externally validated using data 3 collected in the prospective cohort data, using the measures of discrimination and 4 calibration described in QUIDS Protocol Part One, [1] including R 2 , C statistic, 5 calibration slope, calibration-in-the-large, and calibration plots of observed versus 6 predicted risks across deciles (with Loess smoother). The average performance of 7 the model will be summarised across the centers in the cohort study. Between-center 8 heterogeneity in performance will also be summarised, and reduced (if necessary) by 9 recalibration techniques regarding the strategy for the choice of baseline risk 10 (intercept). That is, the predictor effects will not be modified from the IPD meta-11 analysis model, but the intercept may need to be tailored to improve validation in UK 12 centers (e.g. for rural settings). Based on the findings, a final model and its 13 implementation strategy will then be recommended for use. 14 15

Economic Analysis 16
The economic model will be refined, integrated and updated with data from the 17 prospective study cohort, so as the most up to date and validated evidence is used to 18 inform a cost-effectiveness decision. Such an iterative approach to economic 19 evaluation is now well established. [14,15] The care pathway following diagnosis will 20 be included in the economic analysis, using data from the cohort study such as the 21 diagnostic test accuracy data, resource use data (i.e. steroid use, other medications, 22 time in hospital, hospital transfer) and secondary outcome data (i.e., treatment of 23 side-effects, morbidity, mortality) so as to capture the full costs and effect impacts 24 (quality of life, morbidity and mortality) for both the mother and baby. Resource use 25 data will be combined with unit cost information from the British National 26 Formulary [16] and NHS reference costs. [17,18] Outcomes will be reported as the 27 incremental cost per correct diagnosis, and incremental cost per Quality Adjusted 28 Life Year (QALY) gained of the qfFN prognostic model compared to current practice 29 (no qualitative fFN model). The analysis will adhere to the NICE reference case and 1 the recommended guidelines for decision modeling and reporting of economic 2 analyses. [18] Probabilistic sensitivity analysis will be undertaken to explore how 3 uncertainty in the model inputs impact on the cost-effectiveness outcome. [19] 4

Acceptability of fFN Testing and Effects on Anxiety 5
Maternal anxiety will be measured before and after quantitative fFN testing using the 6 validated STAI. The STAI Form Y is a widely used tool for measuring both temporary 7 "state anxiety" and the more general, long-standing "trait anxiety". The STAI is 8 designed for the self-reported assessment of the intensity of feelings of 9 apprehension, tension, nervousness, and worry. STAI-Anxiety scores increase in 10 response to physical danger and psychological stress, making it highly appropriate 11 for this study. The use of STAI in pregnancy studies is discussed by Hundley, et al 12 and we will interpret the results accordingly. [20] 13 14 The questionnaire will be administered prior to fFN testing (baseline) and 24-48 15 hours after the test, to assess early reactions to the test and any acute anxiety 16 prompted by the result of the test. We will also be able to assess any differences in 17 those presented with a high risk or low risk result. Although it might be interesting to 18 assess anxiety again in the latter stages of pregnancy, it is likely that, in this 19 population, many pregnancies will not reach full term. Thus we believe our strategy of 20 repeat questionnaire administration will allow measurement of longer term anxiety 21 induced or alleviated by the test, whilst minimising bias due to preterm or term 22 delivery itself or loss to follow up. 23 24 Follow up interviews will be performed with a sub-group of participants (n=30) to 25 enable deeper exploration of women's views regarding fFN testing, to gain insight 26 into the rationale for responses given in the questionnaires. Interviews will be 27 conducted following confirmation of pregnancy status. Acceptability of the prognostic 28 model will also be assessed with women and a group of clinicians. All interviews will 1 be audio recorded with consent, and field notes taken to ensure an audit trail. prognostic model developed as part of the IPD-meta-analysis, will be tested with 10 women and clinicians, as part of the acceptability studies described above. A final 11 version will be updated with the validated (and, if necessary revised) prognostic 12 model generated from the prospective cohort study. The multidisciplinary trial 13 steering committee will oversee the development process, and decide how material 14 is selected for inclusion. 15

Trial Management And Oversight Arrangements 18
Project Management Group 19 The trial will be coordinated by a Project Management Group (PMG), consisting of 20 the grant holders (Chief Investigator and Co-applicants), the trial manager, 21 representatives from the Study Office and CHaRT (the supporting CTU), plus service 22 user representatives (PAG). The PMG will meet approximately every four months by 23 teleconference or face to face. 24 25 The Trial Manager based in Edinburgh will oversee the study and will be accountable 26 to the Chief Investigator. The Trial Manager supported by the trial administrator(s) 27 will take responsibility for the day-to-day transaction of study activities. They will be 28 supported by the CTU at CHaRT to provide expertise and guidance. The Trial 29 Manager will be responsible for checking the CRFs for completeness, plausibility and 1 consistency. Any queries will be resolved by the Investigator or delegated member 2 of the trial team. On completion of the study, the study data will be analysed and tabulated, and a 21 clinical study report will be prepared in accordance with GCP guidelines. Results will 22 be communicated to the academic community via the scientific literature, attendance 23 at conferences and invited presentations. Summaries of results will also be made 24 available to investigators for dissemination within clinics. Social media will be used to 25 signpost publications and conference presentations and highlight important findings. 26 Twitter and Facebook will be used to disseminate findings to professional 27 organizations, charities, stakeholders and the public. Communication to the general 28 public will further be facilitated by our close links with charities such as Tommy's. [22] 29  1 We anticipate that the decision support will be made available as web based 2 application that will be made freely available so clinicians can access it easily and it 3 can be readily translatable into UK practice. If it is found to be effective in ruling out 4 preterm delivery, it is likely that it will decrease unnecessary costly, and potentially 5 harmful treatments in women who have symptoms suggestive of preterm labour but 6 do not deliver early. 7 8

PEER REVIEW 9
The study was extensively peer reviewed as part of the process of gaining grant 10 funding from the NIHR HTA (14/32/01).         Background Preterm birth, defined as birth prior to 37 weeks gestation, occurs in 6-7% of pregnancies in Europe 1 and was recorded as 5.78% in England in 2013/14, equating to over 37,000 births. 2 Preterm birth is associated with a high risk of mortality, wide-ranging short-and long-term morbidities, 3,4 and significant economic costs to the NHS compared with birth at term. 5 Reducing the detrimental impact of preterm birth relies on the provision of timely and appropriate perinatal interventions.
However, accurate prediction of preterm birth is challenging, even when the clinical symptoms are suggestive of preterm labour. In randomised trials approximately 80% of women diagnosed with preterm labour remained pregnant after 7 days. 6,7 Interventions in preterm labour and preparations for preterm birth may include administration of corticosteroids to accelerate fetal lung maturation 8,9 and magnesium sulphate for fetal neuroprotection, 10 in utero transfer to a facility with appropriate maternity and neonatal services, and tocolysis to optimise time before birth to enable these. 11 Whilst such interventions can improve outcomes for mothers and babies who do experience preterm birth, they are not necessarily benign, especially for those in whom preterm birth does not occur.
The maximal beneficial impact of corticosteroids occurs with administration between 48 hours and seven days before birth, thus timing is especially important in optimising benefit for the neonate. For women who remain at risk of preterm birth after seven days of the initial dose, repeated doses reduce respiratory distress in the neonate 9 but have been found to be associated with a dosedependent reduction in birthweight. 12,13 A five-year follow-up study of women who received repeated doses of antenatal corticosteroids due to risk of preterm birth found an increased risk of neurodevelopment impairment in infants born at term. 14 Therefore developing a strategy to establish the optimal time to give steroids is a research priority. Magnesium sulphate administration immediately prior to birth has been shown to reduce cerebral palsy, 10 but there is a risk of magnesium toxicity leading to respiratory depression in the mother and, theoretically, the neonate. 15 Whilst there is no clear beneficial effect of tocolytics on the incidence or outcome of preterm birth, 16 their use is recommended if the days gained prior to preterm birth can be used appropriately, for example transfer to a suitable maternity unit or the administration of drugs to protect the neonate. 11 Tocolysis is linked with various maternal and neonatal complications, 17 hence the need for therapy targeted only for those at risk of preterm birth and close monitoring of the mother and fetus throughout.
Often, inpatient admission is recommended if preterm labour is suspected. Previous literature has highlighted the social isolation and support needs that women with high-risk pregnancies who are hospitalised experience. 18 In some cases, in-utero transfer is indicated to ensure that birth takes place in a specialist unit with appropriate neonatal care facilities. This policy has been shown to reduce mortality 19,20 and morbidity 21 in preterm neonates, especially those born very premature.
Qualitative research has indicated that women generally acknowledge the potential benefit of in utero transfer to their baby and, hence, are willing to endure the inconvenience and upheaval that it entails. 22,23 However, the experience is associated with an emotional, social and financial burden on women and their families, especially for the substantial proportion of women who do not deliver prematurely following in utero transfer. When describing their experiences of in utero transfer, women expressed shock at the prospect of the transfer, feeling socially isolated, and having no control over the situation, in addition to the practical difficulties experienced particularly by women who already had children. 22,24,25 In a large survey of women who had experienced in utero transfer, over a quarter lamented the financial cost 24 particularly with respect to their partner's outlay for travel, food, accommodation, and phone bills, exacerbated with requiring time off work. 22 Furthermore, in utero transfer is costly to maternity services. Securing a maternal and neonatal bed in another unit is a time-consuming task that often falls to delivery suite midwives to arrange, whilst also continuing to provide care to the woman. 26 In a large observational study of all in utero transfers that took place in Scotland in a six-month period, nearly one third of all transfers were due to threatened preterm labour. 27 Under half of the women transferred from one consultant-led unit to another gave birth within 48 hours. 27 Such unnecessary transfers are costly to women, their families and maternity services. Qualitative research into women's experiences of preterm labour have highlighted the need for caregivers to create an environment where women are enabled to discuss their fears 28 and exert control over how they manage their preterm labour care. 25 Accurate prediction of preterm birth could reduce the burdens and risks associated with unnecessary interventions, and enable women and their clinicians to make informed decisions regarding their care. Numerous diagnostic tests have been used in preterm labour, including biochemical tests of vaginal secretions and cervical length. 29 One such test is fetal fibronectin, a near-bedside test that provides a positive or negative result and has excellent negative predictive value. 30 Thus fetal fibronectin can identify which women will not benefit and may be put at risk by the interventions described previously, and reduce costs to maternity services. 31 Developments in fetal fibronectin testing have led to a quantitative test that provides a concentration of fetal fibronectin in vaginal secretions, giving women and clinicians more information on which to base their management decisions. 32 Qualitative evidence has indicated that women feel a sense of increased responsibility to their babies and themselves during a high risk pregnancy, such as threatened preterm labour. 33 Women want to be involved in decision making about their care to different degrees and feel most satisfied when their caregiver supports them to make decisions in the way they felt most comfortable. 33 Previous literature on decision making and preterm birth has focussed on diagnostic tests 6,[28][29][30][31][32]34 and the care of the preterm infant. 35,36 To date, there has been no investigation of what women, their partners and caregivers would like to know in order to make informed decisions about the care that is provided following the signs and symptoms of preterm labour.
Funding has been received from the National Institute for Health Research Health Technology Assessment Programme for a large, multicentre trial to develop a mobile application decision support tool for the management of women with symptoms and signs of preterm labour, based on a validated model using quantitative fetal fibronectin testing. This study is the precursor to that trial, with the aim of determining the decisional needs of pregnant women with the symptoms and signs of preterm labour, their families and caregivers, using a qualitative framework approach. The outcomes of this qualitative study will inform the development of the mobile application decision support tool, using the findings from an individual patient data meta-analysis. The tool will then be externally validated and refined in the multi-centre trial, QUIDS. Methods A qualitative framework approach will be used, based on data collected from focus groups and semistructured telephone interviews.

Setting
Focus groups will take place in three maternity units: Liverpool Women's NHS Foundation Trust, Birmingham Women's NHS Foundation Trust and Royal Edinburgh Hospital, NHS Lothian. There will be focus groups for women and a separate focus group for partners. Clinicians who care for women with threatened preterm birth will be interviewed by telephone.

Sample
A purposive sample of women and partners will be recruited to cover a variety of experiences of preterm labour and birth. Women will be stratified by their prior experience and relevant characteristics, including ethnicity, previous obstetric history, living in an urban or rural setting and proximity to a tertiary neonatal referral centre. Two focus groups of 4-8 women will be conducted at each site; one for pregnant women who are at high risk of preterm birth, and one for postnatal women who have recently experienced preterm birth. One partners' focus group will be conducted at one of the sites. If women or partners are unable to attend a focus group but still wish to participate, a semi-structured telephone interview will be offered.
Up to 10 obstetricians, including trainees, midwives, and neonatologists will be purposefully recruited to cover a range of professional backgrounds and experience. Semi-structured telephone interviews will be used to collect the data.

Principal inclusion criteria for women's antenatal focus groups
Women who are currently pregnant who: • Have previously experienced preterm birth following preterm labour, • Have experienced threatened preterm labour in this pregnancy, • Are at high risk of preterm birth for another clinical reason, such as prior cervical surgery.

Principal inclusion criteria for women's postnatal focus groups
Women who have experienced preterm birth following preterm labour at <34 weeks whose babies are stable and well and are receiving care on the special care baby unit or neonatal intensive care unit.

Principal inclusion criteria for partners' focus groups
Partners of women who fit the eligibility criteria for either focus group.

Principal exclusion criteria for the focus groups
Non-English speaking individuals.

Principal inclusion criteria for clinician interviews
Clinicians who care for pregnant women i.e. obstetricians (including trainees), neonatologists and midwives.

Women and partners
Eligible women will be identified by clinicians in the preterm birth clinic and other antenatal clinics, and antenatal, triage or labour wards (for the antenatal focus groups) and the special care baby unit or postnatal clinics (for the postnatal focus groups) at each site. Eligible partners will be identified by the same method. Clinicians who are aware of and understand the research aims will approach women and partners to request consent for a researcher to contact them. Importantly, only postnatal parents whose babies are being cared for on the SCBU who are considered stable and well by the clinicians will be approached. With consent the researcher will make contact to talk to the women and/or their partners about the research, either face-to-face or over the telephone. Potential participants will be given the participant information sheet (PIS) (appendix _) that is relevant to them and given verbal information about the study. Each participant will be given time to read the information and the opportunity to have any questions answered. Willing participants will be asked to provide their written consent prior to the focus groups.

Clinicians
Eligible clinicians will be approached by the researchers, via email or face-to-face. Clinicians will be given the clinician PIS (appendix _) and the opportunity to read the information and have any questions answered. Willing clinicians will be asked to provide their written consent prior to the interviews. All participants (women, partners and clinicians) will be reassured that they are not compelled to participate, that they can withdraw from the study at any time, and that non-participation will not affect their care or employment in any way.

Data collection
The primary aim of this research is to determine the decisional requirements of women, their partners and clinicians for the management of preterm labour. Qualitative semi-structured interviews, in a focus-group setting or individual telephone interviews, provide a means of collecting rich, in-depth data with a specific focus. 37 Hence, structured topic guides will be used to initiate and concentrate the discussion (appendices 7-10).
Focus groups are the preferred format for eliciting the view of women and women's partners.
Encouraging discussion among a homogenous group with a shared interest is likely to provide rich insight and understanding into the group's experiences, beliefs and norms as a result of their social interaction. 38 Conversely, interviewing clinicians individually avoids the potential pitfall of professional embarrassment stifling ideas in a group setting. Interviewing individual clinicians with a range of professional experience should ensure that the decisional requirements of clinicians at all levels of experience are understood.
Demographic details and baseline characteristics will be collected prior to the interviews, either as a self-completion questionnaire, or questions asked by the researcher over the telephone. All interviews will be audio recorded, with the participants' consent, and field notes taken. The focus groups will be facilitated by at least two researchers. This is to ensure that all pre-specified areas of interest are covered and that non-verbal communication and group interactions are documented within the field-notes, which will provide context for the data analysis. Recapping will be used to

Analysis plan
A framework approach to data analysis will be used. This approach was developed to manage and interpret large volumes of data collected to inform health policy, meaning they had focussed aims and objectives. 37 Likewise, this research has clear aims, as described previously, in addition to the methodological aim of collecting rich data about the experiences and beliefs of women, their partners and clinicians in relation to managing preterm labour.
Framework analysis follows specific, clearly documented stages of analysis that are transparent so that others can review the interpretation processes and understand how the findings were reached. 39 Transparency is particularly important in this study as the findings will inform the development of an application to aid management decisions in clinical practice. Following verbatim transcription of the interview recordings, the researchers will become familiar with the data by reading the transcripts and field-notes several times. The next stage is to develop a theoretical framework by re-reading the transcripts and making notes as recurring characteristics are recognised. The characteristics will then be collated into themes, which are based on the text itself, supported by the field-notes. The resulting thematic framework will be applied back to the transcripts and field-notes to check that it reflects the context of the original data. The transcripts will be coded, so that portions of text are linked to a discrete theme. A sample of transcripts will be independently coded by two people. The data will be charted and indexed to identify the preterm labour or professional experience of the participant, thus enabling the attribution of themes to a particular group. Finally, the content of the charts will be interpreted and mapped against each other to devise themes and sub-themes categories. Once again, this will involve review of the original data. Explanatory accounts will be developed to clarify the data and quotable sections of data will be identified. The final categories will be discussed between the researchers until consensus is met. The researchers will maintain reflexive journals throughout the data collection and analysis stages, recognising and ameliorating, as far as possible, the fact that their presence and assumptions impact on the data and the findings. 40 This method of data analysis creates a clear audit trail thus ensuring rigour. Each stage of analysis refers back to the original data so that context and meaning is not lost in the final framework of themes and subthemes. The data analysis process will be managed using NVivo software, a qualitative data analysis tool.

Safety
The physical safety of participants will be ensured through adhering to the health and safety policies of the host units where the focus groups take place.
The emotional wellbeing of the participants will be safeguarded by following the Distress Policy (see appendix 11). The Supervisors of Midwives (SOM) team in each unit will be informed of the study and women and their partners will be given the SOM team contact details, should they become distressed or upset as a result of talking about their experiences. Participants will also be given the contact details for accessing local counselling services.

Informed consent
All participants will be fully informed about the study and the subsequent QUIDS trial via verbal and written communication. All eligible individuals will be given the participant information sheet (appendix __) and provided with an opportunity to have any questions answered. Written consent will then be gained prior to the commencement of the focus groups/interviews.

Confidentiality
Demographic information will be collected from participants to attribute themes from the data to particular groups within the analysis and dissemination of findings. Demographic information, which will contain potentially identifiable information, will be kept in a secure lockable cabinet. Audio recordings will be stored on an encryptable audio device only until they are transcribed. Once transcribed the audio recordings will be deleted. Transcription services are provided by '1 st Class Secretarial', who subscribe to the Data Protection Act and have also signed the Code of Practice on Data Handling. Hard copies of audio transcripts and field-notes will be kept in a separate secure  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  lockable cabinet to the demographic information. The transcripts and field-notes will be coded to identify which participant provided that data; the codes will only be known by the researchers.
Participant's data will not be used for any purpose other than this study and the subsequent QUIDS trial.

Data Protection
Participants will be informed that publications from this study will contain direct quotes from the focus groups/interviews and categorisation of their experience of preterm labour (e.g. experienced preterm birth), which could enable personal identification.
All researchers involved in this study must comply with the requirements of the Data Protection Act 1998 with regard to the collection, storage, processing and disclosure of personal information and uphold the Act's core principles. All computers used for processing data are password protected and subject to the strict data protection policies of the researcher's institution.

Good clinical practice training
All researchers involved in this study must hold evidence of recent Good Clinical Practice training.

Safety of researchers
An individualised risk assessment will be conducted to identify any risks to researchers or participants involved in this study. The lone working policy of the institution will be adhered to at all times. The only anticipated lone working will be during travel to and from the interview sites.
The lone working policy of the researcher's institutions mandates that researchers wear a GPS tracking and audio transmitting device during all lone-working, off-site research activity with participants. Participants will be informed if this device is being used.

Insurance / Indemnity
The researcher's institution holds public liability insurance and professional indemnity insurance (appendices 12, 13 and 14).

STRENGTHS AND LIMITATIONS OF THIS STUDY 25
Strengths 26

•
Validation of a prognostic model in a separate prospective cohort study 27  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  The overall aim of the QUIDS study is to develop a decision support tool for the 2 management of women with symptoms and signs of preterm labour, based on a 3 validated prognostic model using quantitative fFN testing. The study has been 4 conceptually divided into two parts. In this, the protocol for QUIDS Part Two, we 5 detail the protocol for a prospective cohort study. This will externally validate a 6 prognostic model developed in QUIDS Part One.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   6 The aim of the QUIDS study is to develop a decision support tool for the 1 management of women with symptoms and signs of preterm labour, based on a 2 validated prognostic model using quantitative fFN testing. 3 4 The study protocol has been divided into two parts (see flow chart Figure 1). The 5 protocols for Parts One and Two are reported in separate manuscripts. 6 7 In QUIDS Protocol Part One we have described how we will perform (i) an Individual 8 Participant Data (IPD) meta-analysis, and (ii) and Economic Analysis. The protocol 9 details how we will develop and internally validate a prognostic model using 10 quantitative fFN (as a continuous variable) and other risk (prognostic) factors and to 11 evaluate the added value of quantitative fFN toward this prognostic model 12 performance. We will also provide an economic rationale for the prognostic model 13 and analyze its cost-effectiveness from the perspective of the NHS. 14 15 In this, the QUIDS Protocol Part Two, we will detail the prospective cohort study to 16 externally validate and, if necessary, refine the prognostic model. This will be 17 performed in at least eight UK hospitals with different settings (rural/urban) and 18 different levels of neonatal care facilities. In addition, acceptability of quantitative fFN 19 testing, and effects on maternal anxiety will be performed. We will assess the 20 potential cost-effectiveness of the final prognostic model/decision support tool. This  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   7 The primary endpoint of the prognostic model is spontaneous preterm delivery within 1 seven days of qfFN test, in women less than 36 weeks' gestation. This was 2 influenced by the preceding QUIDS Qualitative Study, which included focus group 3 consultation to determine the decisional needs of women, their partners and 4 clinicians (Supplementary Material). It is also a recognised clinically important 5 endpoint, as antenatal steroids (which significantly reduce morbidity and mortality in 6 preterm babies [4]) are most effective if delivery occurs within seven days of 7 administration.   The prospective cohort study will include women with signs and symptoms of 24 preterm labour at 22 +0 to 34 +6 weeks gestation in whom admission, transfer or 25 treatment is being considered. These will be recruited from at least eight sites with a 26 mix of rural/urban settings, and have different levels of neonatal care facilities, over 27 12 months. 28 (antenatal steroids, tocolysis or magnesium sulphate) is being considered due 10 to signs of pre-term labour. 11 • Women aged 16 years or above. 12 The broad inclusion criteria reflect current clinical practice and enable the 13 generalisability of the results of the trial for routine clinical care. We will include 14 women who re-attend seven days or more after initial recruitment with signs and 15 symptoms of preterm labour and also women who remain symptomatic but 16 undelivered seven days later in whom repeat testing by the clinician is deemed to be 17 appropriate. This will be in line with manufacturer's recommendation for fFN testing. 18 19 The following inclusion criteria will apply on speculum examination: 20 • Cervical dilation ≤ 3cm 21 • Intact membranes 22 • No significant vaginal bleeding, as judged by the clinician. 23 • Once it has been established that the women meets the above criteria, on 24 speculum examination, the fFN swab can be taken. 25 Participants that sign the consent but are not eligible upon examination to have an 26 fFN swab taken will still be enrolled and have outcome data collected. 27 The following exclusion criteria will apply: 2 • Contraindication to vaginal examination (e.g. placenta praevia). 3 • Higher order multiple pregnancy (triplets or more). 4 • Moderate or severe vaginal bleeding. 5 • Cervical dilatation greater than 3cm. 6 • Confirmed rupture of membranes. 7 • Sexual intercourse, vaginal examination or transvaginal ultrasound in the 8 preceding 24 hours factors may invalidate results. These women will be 9 initially excluded from the study, but can be included if still symptomatic after 10 24 hours, when fFN accuracy will be restored. 11 12

Co-Enrolment 13
This trial involves validating a decision support tool relating to a test that is currently 14 commonly used in clinical practice. As such, there are no additional interventions. 15 Co-enrolment in other non-interventional trials will be allowed. Co-enrolment in trials 16 of tocolytic treatments or other management strategies that may influence timing of 17 delivery as a primary outcome will not be allowed. Participation in QUIDs would not 18 preclude babies being subsequently involved in interventional trials. Co-enrolment 19 will be recorded in the electronic case report form (eCRF). 20

Setting 22
The prospective cohort study will take place in at least eight consultant-led obstetric 23 units in the UK. More than 93% of pregnant women in the UK deliver in consultant-24 led units. [5,6] The vast majority of women with symptoms of preterm labour will 25 present to a consultant-led unit for assessment, either directly or following advice 26 from their community midwife or General Practitioner. 27 The study will not include any community maternity units (staffed by midwives, with 2 or without involvement of non-obstetric medical staff), which cover a small proportion 3 of women, mainly in remote and rural areas. In the Perinatal Collaborative Transport 4 Study (CoTS study) of perinatal transfers in Scotland, [7] which involved 52,727 5 births, only 69 (0.13%) women were transferred to a consultant-led obstetric unit 6 from community maternity units, and only a proportion of these were for suspected 7 preterm labour. The small number of women cared for in community maternity units 8 means their inclusion would not be an efficient use of study resources.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Posters and leaflets will be situated in antenatal areas of participating hospitals to 8 alert women that the study is taking place, and women will be allowed as much time 9 as possible to consider participation without unduly delaying further clinical 10 assessment. Participants will receive adequate oral and written information and 11 appropriate participant information and informed consent forms will be provided.  Certain exclusion criteria can only be assessed at speculum examination (for 1 example vaginal bleeding or evidence of ruptured membranes), so a proportion of 2 women will not be eligible for fFN testing after consent is given. These women will 3 still be enrolled and delivery outcomes collected. The decision whether to use this 4 data for analysis will be the decision of the Chief Investigator and Statisticians. 5 6 Withdrawal Of Study Participants 7 Women will be able to withdraw consent for us of their data at any time until the end 8 of the study. 9 10 Study Assessments (See Table 1) 11

Eligibility Assessment (Screening And Recruitment) 12
Women presenting with signs and symptoms of pre-term labour will be identified on 13 presentation to obstetric services. The doctor or midwife assessing the woman will 14 identify potentially eligible participants and provide an invitation letter and short 15 information leaflet. 16 17 After the woman has had the opportunity to consider whether she would like to 18 participate, she will be asked to complete the Screening and Consent Form. The 19 clinician will then decide whether the fFN test can be carried out. If the test can be 20 carried out (according to manufacturer's guidelines), then the participant will be fully 21 enrolled and that their delivery outcomes will still be collected. 22 23 If the woman declines to participate and she is willing to provide a reason for this, the 24 reason given will be entered on to an anonymous log. Baseline demographics will be 25 collected on consenting women, together with height and weight, information on 26 medical history, obstetric history, estimated date of delivery and presenting signs and 27 symptoms. 28 The original consent form will be stored in the Investigator Site File (ISF) file, a copy 2 is given to the woman, a copy added to the medical notes and a copy sent to the 3 Trial Office. 4 5 After providing consent, the participant will be asked to complete a short State Trait 6 Anxiety Inventory (STAI) questionnaire and complete a contact details form. They will 7 also be issued with a letter thanking them for taking part in the trial and giving details 8 of the second questionnaire to be completed. 9 10

Sample Collection 11
Samples for analysis will be taken with a fFN specimen collection kit, which consists 12 of a sterile polyester tipped swab and a specimen transport tube containing 1 ml  The sample taken will be run at a near bedside Hologic Rapid fFN 10Q analyser, 1 specially adapted for the QUIDS study. As fFN (or other similar biochemical tests of 2 preterm labour) are part of standard care, it would be unethical to blind clinicians 3 from the qualitative fFN result. The analyser will thus reveal a qualitative fFN result 4 (positive/negative/invalid based on a 50ng/ml threshold) for clinicians to base clinical 5 decision-making on, according to local protocols. The quantitative fFN result 6 however, will be stored as a three-letter code, blinding caregivers from the result. 7 Samples will be run as per manufacturers instructions (described above in the 8 section "Health technologies being assessed"). 9 10

Repeat fFN Tests 11
If there is clinical indication for further fFN tests (eg because of ongoing symptoms of 12 preterm labour after seven days), the results will also be recorded. 13 14

Labour/Delivery/ Neonatal Assessments 15
Admission for delivery will not be a formal study visit but data will be collected using 16 information recorded in the participant's notes. Delivery data will be collected on the 17 maternal outcomes of delivery, including method of delivery, indication for delivery 18 method, onset of labour, date and gestation of delivery and blood loss. 19 20

Questionnaires 21
All participants who are eligible to participate will be asked to complete a STAI 22 questionnaire before the speculum examination. The same questionnaire will be 23 repeated 24-48 hours post examination. The second questionnaire will be provided 24 on paper with a pre-paid envelope to be returned by post to the Trial Office.  The Hologic Rapid fFN 10Q analyzer has integrated quality control measures, and 2 we will keep records of these as well as any additional staff training that occurs after 3 the study starts. It is recommended that a daily pre-calibrated reusable quality control 4 cassette be inserted and analysed every 24 hours to verify that the analyser 5 performance is within specification. A daily quality control (QC) should be performed 6 if one has not been done in the preceding 24 hours before a patient test is to be 7 done. Logs of results are stored on the machine and can be downloaded, and we will 8 also ask the participating sites to keep a monthly paper log of QC tests done. Each 9 patient test has an internal quality control, with a procedural control line that verifies 10 the threshold level of signal by the instrument. Sample flow detection ensures the 11 sample travels across the cassette properly, and confirms absence of conjugate 12 aggregation. We believe that these measures will help ensure the validity of results. 13 However, to provide further evidence of integrity and comparability of results from 14 each site we will request that all participating sites enrol in the Wales External Quality 15 Assurance Scheme (WEQAS) Point of Care Quality Assurance Scheme. WEQAS will 16 provide a sample for analysis to each site bimonthly, and provide reports on analyser 17 performance and variability. [ Screening data and data about quantitative fFN testing will be collected on paper 8 based CRFs and research midwives will input these into the web based electronic 9 database. Clinical outcome data will be collected from the medical records. 10 11

Maternal Acceptability and Anxiety 12
Maternal anxiety will be measured pre and post-test (24-48h) using the validated 13 State Trait Anxiety Inventory (STAI) questionnaire. Acceptability of fFN testing and 14 the decision support will be assessed using follow up interviews (face to face or 15 telephone, according to maternal preference) which will be conducted with a sub-16 group of participants (n=30) purposively sampled and stratified according to 17 geographical location, outcome (preterm labour or not) and anxiety scores. 18 Acceptability will also be assessed in a cohort of clinicians (n=30). 19 20

Statistics and Sample Size Calculation 21
Guidance for external validation suggests at least ten events (preterm delivery within 22 seven days of test) are required for each covariate included in a prognostic 23 model. [9,10] Data from the cohorts included in our IPD meta-analysis suggests an 24 event rate of between 6 and 12%. [1] Based on these estimates a sample size of 25 1,600 will provide 96 and 192 events (preterm delivery within 7 days). 26 27 A UK study has shown that 8.9% of pregnant women present with symptoms of 28 preterm labour and are eligible for quantitative fFN [11] and we anticipate 50% 29  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   19 recruitment rate is achievable, thus overall 4.5% of maternities could be recruited. 1 We will initially include eight units in the cohort study with a combined delivery rate of 2 approximately 36,000 per annum. We anticipate that we will achieve target 3 recruitment within 12 months (1 year * 36,000 * 0.089 * 0.5 = 1,602). If however, the 4 recruitment rate or event rate is lower than predicted, we will increase the number of 5 sites included in the study and/or the recruitment period, to ensure that a minimum of 6 60 events (preterm delivery within 7 days of test) are achieved, allowing for external 7 validation of at least six covariates in our model. 8 9 It is possible that the IPD meta-analysis will find there is potential added value of 10 combining quantitative fFN testing with cervical length measurement. [12,13] As 11 cervical length measurement has significant resource requirement (estimated NHS 12 cost £68.16 per test) and lack of out of hours provision further limits availability in 13 many NHS hospitals, we think it is very unlikely that cervical length scanning will 14 improve performance of the prognostic model to such a degree as to make it cost 15 effective. We will assess the incremental costs and effects of cervical length 16 measurement in the proposed health economic model performed in parallel with the 17 IPD meta-analysis, and will feed into design considerations during the first iteration of 18 the prognostic model. 19 20 If inclusion of cervical length ultrasound is found to be potentially cost-effective, we 21 will assess the feasibility of including it in the prospective cohort study. We anticipate 22 that including cervical length measurement in the prospective cohort study would be 23 extremely difficult in the current NHS setting as the majority of units do not have 24 24 hour availability of transvaginal ultrasound and/or trained personnel to perform scans.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y The prognostic model developed in the IPD will be externally validated using data 3 collected in the prospective cohort data, using the measures of discrimination and 4 calibration described in QUIDS Protocol Part One, [1] including R 2 , C statistic, 5 calibration slope, calibration-in-the-large, and calibration plots of observed versus 6 predicted risks across deciles (with Loess smoother). The average performance of 7 the model will be summarised across the centers in the cohort study. Between-center 8 heterogeneity in performance will also be summarised, and reduced (if necessary) by 9 recalibration techniques regarding the strategy for the choice of baseline risk 10 (intercept). That is, the predictor effects will not be modified from the IPD meta-11 analysis model, but the intercept may need to be tailored to improve validation in UK 12 centers (e.g. for rural settings). Based on the findings, a final model and its 13 implementation strategy will then be recommended for use. 14 15

Economic Analysis 16
The economic model will be refined, integrated and updated with data from the 17 prospective study cohort, so as the most up to date and validated evidence is used to 18 inform a cost-effectiveness decision. Such an iterative approach to economic 19 evaluation is now well established. [14,15] The care pathway following diagnosis will 20 be included in the economic analysis, using data from the cohort study such as the 21 diagnostic test accuracy data, resource use data (i.e. steroid use, other medications, 22 time in hospital, hospital transfer) and secondary outcome data (i.e., treatment of 23 side-effects, morbidity, mortality) so as to capture the full costs and effect impacts 24 (quality of life, morbidity and mortality) for both the mother and baby. Resource use 25 data will be combined with unit cost information from the British National 26 Formulary [16] and NHS reference costs. [17,18] Outcomes will be reported as the 27 incremental cost per correct diagnosis, and incremental cost per Quality Adjusted 28 Life Year (QALY) gained of the qfFN prognostic model compared to current practice 29  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60   F  o  r  p  e  e  r  r  e  v  i  e  w  o  n  l  y   22 model will also be assessed with women and a group of clinicians. All interviews will 1 be audio recorded with consent, and field notes taken to ensure an audit trail. 2 3 Decision Support 4 We will develop a decision support tool in accordance with the guidelines produced 5 by the International Patient Decision Aid Standards (IPDAS) Collaboration. [21] 6 Scoping of decisional requirements and how data should be presented was 7 performed during focus group consultation as part of QUIDS Qualitative 8 (Supplementary Material). A prototype decision support tool incorporating the initial 9 prognostic model developed as part of the IPD-meta-analysis, will be tested with 10 women and clinicians, as part of the acceptability studies described above. A final 11 version will be updated with the validated (and, if necessary revised) prognostic 12 model generated from the prospective cohort study. The multidisciplinary trial 13 steering committee will oversee the development process, and decide how material 14 is selected for inclusion. 15 16

ETHICS AND DISSEMINATION 4
The study will be conducted in accordance with the principles of Good Clinical 5 Practice (GCP). A favorable ethical opinion has been obtained from the appropriate 6 REC (reference 16/WS/0068) and local R&D approval will be obtained prior to 7 commencement of the study at each site. 8 9 On completion of the study, the study data will be analysed and tabulated, and a 10 clinical study report will be prepared in accordance with GCP guidelines. Results will 11 be communicated to the academic community via the scientific literature, attendance 12 at conferences and invited presentations. Summaries of results will also be made 13 available to investigators for dissemination within clinics. We anticipate that the 14 decision support will be made available as web based application that will be made 15 freely available so clinicians can access it easily and it can be readily translatable 16 into UK practice. If it is found to be effective in ruling out preterm delivery, it is likely 17 that it will decrease unnecessary costly, and potentially harmful treatments in women 18 who have symptoms suggestive of preterm labour but do not deliver early. 19 20
However, accurate prediction of preterm birth is challenging, even when the clinical symptoms are suggestive of preterm labour. In randomised trials approximately 80% of women diagnosed with preterm labour remained pregnant after 7 days. 6,7 Interventions in preterm labour and preparations for preterm birth may include administration of corticosteroids to accelerate fetal lung maturation 8,9 and magnesium sulphate for fetal neuroprotection, 10 in utero transfer to a facility with appropriate maternity and neonatal services, and tocolysis to optimise time before birth to enable these. 11 Whilst such interventions can improve outcomes for mothers and babies who do experience preterm birth, they are not necessarily benign, especially for those in whom preterm birth does not occur.
The maximal beneficial impact of corticosteroids occurs with administration between 48 hours and seven days before birth, thus timing is especially important in optimising benefit for the neonate. For women who remain at risk of preterm birth after seven days of the initial dose, repeated doses reduce respiratory distress in the neonate 9 but have been found to be associated with a dosedependent reduction in birthweight. 12,13 A five-year follow-up study of women who received repeated doses of antenatal corticosteroids due to risk of preterm birth found an increased risk of neurodevelopment impairment in infants born at term. 14 Therefore developing a strategy to establish the optimal time to give steroids is a research priority.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Magnesium sulphate administration immediately prior to birth has been shown to reduce cerebral palsy, 10 but there is a risk of magnesium toxicity leading to respiratory depression in the mother and, theoretically, the neonate. 15 Whilst there is no clear beneficial effect of tocolytics on the incidence or outcome of preterm birth, 16 their use is recommended if the days gained prior to preterm birth can be used appropriately, for example transfer to a suitable maternity unit or the administration of drugs to protect the neonate. 11 Tocolysis is linked with various maternal and neonatal complications, 17 hence the need for therapy targeted only for those at risk of preterm birth and close monitoring of the mother and fetus throughout.
Often, inpatient admission is recommended if preterm labour is suspected. Previous literature has highlighted the social isolation and support needs that women with high-risk pregnancies who are hospitalised experience. 18 In some cases, in-utero transfer is indicated to ensure that birth takes place in a specialist unit with appropriate neonatal care facilities. This policy has been shown to reduce mortality 19,20 and morbidity 21 in preterm neonates, especially those born very premature.
Qualitative research has indicated that women generally acknowledge the potential benefit of in utero transfer to their baby and, hence, are willing to endure the inconvenience and upheaval that it entails. 22,23 However, the experience is associated with an emotional, social and financial burden on women and their families, especially for the substantial proportion of women who do not deliver prematurely following in utero transfer. When describing their experiences of in utero transfer, women expressed shock at the prospect of the transfer, feeling socially isolated, and having no control over the situation, in addition to the practical difficulties experienced particularly by women who already had children. 22,24,25 In a large survey of women who had experienced in utero transfer, over a quarter lamented the financial cost 24 particularly with respect to their partner's outlay for travel, food, accommodation, and phone bills, exacerbated with requiring time off work. 22 Furthermore, in utero transfer is costly to maternity services. Securing a maternal and neonatal bed  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  in another unit is a time-consuming task that often falls to delivery suite midwives to arrange, whilst also continuing to provide care to the woman. 26 In a large observational study of all in utero transfers that took place in Scotland in a six-month period, nearly one third of all transfers were due to threatened preterm labour. 27 Under half of the women transferred from one consultant-led unit to another gave birth within 48 hours. 27 Such unnecessary transfers are costly to women, their families and maternity services. Qualitative research into women's experiences of preterm labour have highlighted the need for caregivers to create an environment where women are enabled to discuss their fears 28 and exert control over how they manage their preterm labour care. 25 Accurate prediction of preterm birth could reduce the burdens and risks associated with unnecessary interventions, and enable women and their clinicians to make informed decisions regarding their care. Numerous diagnostic tests have been used in preterm labour, including biochemical tests of vaginal secretions and cervical length. 29 One such test is fetal fibronectin, a near-bedside test that provides a positive or negative result and has excellent negative predictive value. 30 Thus fetal fibronectin can identify which women will not benefit and may be put at risk by the interventions described previously, and reduce costs to maternity services. 31 Developments in fetal fibronectin testing have led to a quantitative test that provides a concentration of fetal fibronectin in vaginal secretions, giving women and clinicians more information on which to base their management decisions. 32 Qualitative evidence has indicated that women feel a sense of increased responsibility to their babies and themselves during a high risk pregnancy, such as threatened preterm labour. 33 Women want to be involved in decision making about their care to different degrees and feel most satisfied when their caregiver supports them to make decisions in the way they felt most comfortable. 33  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59 35,36 To date, there has been no investigation of what women, their partners and caregivers would like to know in order to make informed decisions about the care that is provided following the signs and symptoms of preterm labour.
Funding has been received from the National Institute for Health Research Health Technology Assessment Programme for a large, multicentre trial to develop a mobile application decision support tool for the management of women with symptoms and signs of preterm labour, based on a validated model using quantitative fetal fibronectin testing. This study is the precursor to that trial, with the aim of determining the decisional needs of pregnant women with the symptoms and signs of preterm labour, their families and caregivers, using a qualitative framework approach. The outcomes of this qualitative study will inform the development of the mobile application decision support tool, using the findings from an individual patient data meta-analysis. The tool will then be externally validated and refined in the multi-centre trial, QUIDS.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  Methods A qualitative framework approach will be used, based on data collected from focus groups and semistructured telephone interviews.

Setting
Focus groups will take place in three maternity units: Liverpool Women's NHS Foundation Trust, Birmingham Women's NHS Foundation Trust and Royal Edinburgh Hospital, NHS Lothian. There will be focus groups for women and a separate focus group for partners. Clinicians who care for women with threatened preterm birth will be interviewed by telephone.

Sample
A purposive sample of women and partners will be recruited to cover a variety of experiences of preterm labour and birth. Women will be stratified by their prior experience and relevant characteristics, including ethnicity, previous obstetric history, living in an urban or rural setting and proximity to a tertiary neonatal referral centre. Two focus groups of 4-8 women will be conducted at each site; one for pregnant women who are at high risk of preterm birth, and one for postnatal women who have recently experienced preterm birth. One partners' focus group will be conducted at one of the sites. If women or partners are unable to attend a focus group but still wish to participate, a semi-structured telephone interview will be offered.

Principal inclusion criteria for women's antenatal focus groups
Women who are currently pregnant who: • Have previously experienced preterm birth following preterm labour, • Have experienced threatened preterm labour in this pregnancy, • Are at high risk of preterm birth for another clinical reason, such as prior cervical surgery.

Principal inclusion criteria for women's postnatal focus groups
Women who have experienced preterm birth following preterm labour at <34 weeks whose babies are stable and well and are receiving care on the special care baby unit or neonatal intensive care unit.

Principal inclusion criteria for partners' focus groups
Partners of women who fit the eligibility criteria for either focus group.

Principal exclusion criteria for the focus groups
Non-English speaking individuals.

Women and partners
Eligible women will be identified by clinicians in the preterm birth clinic and other antenatal clinics, and antenatal, triage or labour wards (for the antenatal focus groups) and the special care baby unit or postnatal clinics (for the postnatal focus groups) at each site. Eligible partners will be identified by the same method. Clinicians who are aware of and understand the research aims will approach women and partners to request consent for a researcher to contact them. Importantly, only postnatal parents whose babies are being cared for on the SCBU who are considered stable and well by the clinicians will be approached. With consent the researcher will make contact to talk to the women and/or their partners about the research, either face-to-face or over the telephone. Potential participants will be given the participant information sheet (PIS) (appendix _) that is relevant to them and given verbal information about the study. Each participant will be given time to read the information and the opportunity to have any questions answered. Willing participants will be asked to provide their written consent prior to the focus groups.

Clinicians
Eligible clinicians will be approached by the researchers, via email or face-to-face. Clinicians will be given the clinician PIS (appendix _) and the opportunity to read the information and have any questions answered. Willing clinicians will be asked to provide their written consent prior to the interviews.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  All participants (women, partners and clinicians) will be reassured that they are not compelled to participate, that they can withdraw from the study at any time, and that non-participation will not affect their care or employment in any way.

Data collection
The primary aim of this research is to determine the decisional requirements of women, their partners and clinicians for the management of preterm labour. Qualitative semi-structured interviews, in a focus-group setting or individual telephone interviews, provide a means of collecting rich, in-depth data with a specific focus. 37 Hence, structured topic guides will be used to initiate and concentrate the discussion (appendices 7-10).
Focus groups are the preferred format for eliciting the view of women and women's partners.
Encouraging discussion among a homogenous group with a shared interest is likely to provide rich insight and understanding into the group's experiences, beliefs and norms as a result of their social interaction. 38 Conversely, interviewing clinicians individually avoids the potential pitfall of professional embarrassment stifling ideas in a group setting. Interviewing individual clinicians with a range of professional experience should ensure that the decisional requirements of clinicians at all levels of experience are understood.
Demographic details and baseline characteristics will be collected prior to the interviews, either as a self-completion questionnaire, or questions asked by the researcher over the telephone. All interviews will be audio recorded, with the participants' consent, and field notes taken. The focus groups will be facilitated by at least two researchers. This is to ensure that all pre-specified areas of interest are covered and that non-verbal communication and group interactions are documented within the field-notes, which will provide context for the data analysis. Recapping will be used to  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  clarify aspects and avoid misinterpretation. To enable all participants to talk freely, the researchers will be unknown to the participants and not working clinically in the unit where the interview is conducted. Clinicians will be interviewed by a researcher who is unknown to them.

Analysis plan
A framework approach to data analysis will be used. This approach was developed to manage and interpret large volumes of data collected to inform health policy, meaning they had focussed aims and objectives. 37 Likewise, this research has clear aims, as described previously, in addition to the methodological aim of collecting rich data about the experiences and beliefs of women, their partners and clinicians in relation to managing preterm labour.
Framework analysis follows specific, clearly documented stages of analysis that are transparent so that others can review the interpretation processes and understand how the findings were reached. 39 Transparency is particularly important in this study as the findings will inform the development of an application to aid management decisions in clinical practice. Following verbatim transcription of the interview recordings, the researchers will become familiar with the data by reading the transcripts and field-notes several times. The next stage is to develop a theoretical framework by re-reading the transcripts and making notes as recurring characteristics are recognised. The characteristics will then be collated into themes, which are based on the text itself, supported by the field-notes. The resulting thematic framework will be applied back to the transcripts and field-notes to check that it reflects the context of the original data. The transcripts will be coded, so that portions of text are linked to a discrete theme. A sample of transcripts will be independently coded by two people. The data will be charted and indexed to identify the preterm labour or professional experience of the participant, thus enabling the attribution of themes to a particular group. Finally, the content of the charts will be interpreted and mapped against each other to devise themes and sub-themes categories. Once again, this will involve review of the original data. Explanatory accounts will be developed to clarify the data and quotable sections of data will be identified. The final categories will be discussed between the researchers until consensus is met. The researchers will maintain reflexive journals throughout the data collection and analysis stages, recognising and ameliorating, as far as possible, the fact that their presence and assumptions impact on the data and the findings. 40 This method of data analysis creates a clear audit trail thus ensuring rigour. Each stage of analysis refers back to the original data so that context and meaning is not lost in the final framework of themes and subthemes. The data analysis process will be managed using NVivo software, a qualitative data analysis tool.

Safety
The physical safety of participants will be ensured through adhering to the health and safety policies of the host units where the focus groups take place.
The emotional wellbeing of the participants will be safeguarded by following the Distress Policy (see appendix 11). The Supervisors of Midwives (SOM) team in each unit will be informed of the study and women and their partners will be given the SOM team contact details, should they become distressed or upset as a result of talking about their experiences. Participants will also be given the contact details for accessing local counselling services.

Informed consent
All participants will be fully informed about the study and the subsequent QUIDS trial via verbal and written communication. All eligible individuals will be given the participant information sheet (appendix __) and provided with an opportunity to have any questions answered. Written consent will then be gained prior to the commencement of the focus groups/interviews.

Confidentiality
Demographic information will be collected from participants to attribute themes from the data to particular groups within the analysis and dissemination of findings. Demographic information, which will contain potentially identifiable information, will be kept in a secure lockable cabinet. Audio recordings will be stored on an encryptable audio device only until they are transcribed. Once transcribed the audio recordings will be deleted. Transcription services are provided by '1 st Class Secretarial', who subscribe to the Data Protection Act and have also signed the Code of Practice on Data Handling. Hard copies of audio transcripts and field-notes will be kept in a separate secure  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  lockable cabinet to the demographic information. The transcripts and field-notes will be coded to identify which participant provided that data; the codes will only be known by the researchers.
Participant's data will not be used for any purpose other than this study and the subsequent QUIDS trial.

Data Protection
Participants will be informed that publications from this study will contain direct quotes from the focus groups/interviews and categorisation of their experience of preterm labour (e.g. experienced preterm birth), which could enable personal identification.
All researchers involved in this study must comply with the requirements of the Data Protection Act 1998 with regard to the collection, storage, processing and disclosure of personal information and uphold the Act's core principles. All computers used for processing data are password protected and subject to the strict data protection policies of the researcher's institution.

Good clinical practice training
All researchers involved in this study must hold evidence of recent Good Clinical Practice training.

Safety of researchers
An individualised risk assessment will be conducted to identify any risks to researchers or participants involved in this study. The lone working policy of the institution will be adhered to at all times. The only anticipated lone working will be during travel to and from the interview sites.
The lone working policy of the researcher's institutions mandates that researchers wear a GPS tracking and audio transmitting device during all lone-working, off-site research activity with participants. Participants will be informed if this device is being used.

Insurance / Indemnity
The researcher's institution holds public liability insurance and professional indemnity insurance (appendices 12, 13 and 14).