A tool to predict spontaneous preterm birth, incorporating fetal fibronectin and cervical length, in symptomatic women and high-risk asymptomatic women
The QUiPP app is a clinical decision-making tool with the potential to revolutionize preterm birth prediction for women with symptoms of threatened preterm labour as well as asymptomatic high-risk women. Accurate diagnosis of preterm labour is desirable in order to prevent the maternal and fetal risks incurred to the majority of women who are over-managed, without missing true cases.
This application has been designed for health, allied health and health research professionals who look after pregnant women to calculate individualised % risks scores of delivery within pre-specified clinically relevant timeframes. It is designed to be used with women as an educational tool and to arrive at shared decisions regarding the management of their pregnancy.
It is designed for use in two clinical settings:
It relies on a relevant clinical history having been taken regarding the woman’s risk factors for preterm birth and her current symptoms. It relies on existing point-of-care testing: quantitative fetal fibronectin (fFN) sampling of the cervico-vaginal fluid and/or transvaginal ultrasound cervical length (CL) measurements. Therefore the user is expected to have significant midwifery or obstetric experience in order to use QUiPP app or is working closely with a team-member who does.
Recent publications about the QUiPP app.
The QUiPP App: a safe alternative to a treat-all strategy for threatened preterm labour.
H. A. WATSON, J. CARTER, P. T. SEED, R. M. TRIBE and A. H. SHENNAN
Objective: To evaluate the impact of triaging women at risk of spontaneous preterm birth (sPTB) using the QUiPP App, which incorporates a predictive model combining history of sPTB, gestational age and quantitative measurements of fetal fibronectin, compared with a treat-all policy (advocated by the UK National Institute for Health and Care Excellence) among women with threatened preterm labour before 30 weeks’ gestation.
Methods: Prospectively collected data from pregnant women presenting to a tertiary inner-city hospital with symptoms of preterm labour (abdominal pain or tightenings) at 24–34 weeks’ gestation were retrieved from the research databases of the EQUIPP and PETRA studies for sub-analysis. Each episode of threatened preterm labour was retrospectively assigned a risk for sPTB within 7 days using the QUiPP App. A primary outcome of delivery within 7 days was used to model the performance accuracy of the QUiPP App compared with a treat-all policy.
Results using a 5% risk of delivery within 7 days according to the QUiPP App as the threshold for intervention, 9/9 women with threatened preterm labour would have been treated correctly, giving a sensitivity of 100% (one-sided 97.5% CI, 66.4%) and a negative predictive value of 100% (95% CI, 98.9–100% ). The positive predictive value for delivery within 7 days was 30.0% (95% CI, 11.9–54.3%) for women presenting before 30 weeks and 20.0% (95% CI, 4.3–49.1%) for women presenting between 30 and 34 weeks. If this 5% threshold had been used to triage women presenting between 24 and 29+6 weeks, 89.4% (n=168) of admissions could have been safely avoided, compared with 0% for a treat-all strategy. No true case of preterm labour would have been missed, as no woman who was assigned a risk of <10% delivered within 7 days.
Conclusion: For women with threatened preterm labour, the QUiPP App can accurately guide management at risk thresholds for sPTB of 1%, 5% and 10%, allowing outpatient management in the vast majority of cases. A treat-all approach would not have avoided admission for any woman, exposed 188 mothers and their babies to unnecessary hospitalization and steroid administration and increased the burden on network and transport services owing to unnecessary in-utero transfers. Prediction of sPTB should be performed before 30 weeks to determine management until there is evidence that such a high level of unnecessary intervention, as suggested by the treat-all strategy, does less harm than the occurrence of rare false negatives.
Published by Ultrasound in Obstetrics & Gynaecology, July 2017
Miss Katy KUHRT, BSc; Elizabeth SMOUT, MBBS; Natasha HEZELGRAVE, Mr Paul T SEED, MSc; Andrew H SHENNAN, MD
Woman’s Health Academic Centre, Kings College London
Objective
Every year 15 million babies are born preterm (<37 weeks’ gestation)1 but prediction of spontaneous preterm birth (sPTB) is poor making it difficult to target interventions appropriately.
Recently, fetal fibronectin (fFN), though to act like a ‘glue’ between the maternal and fetal membranes,2 and cervical length (CL) assessment have superseded previous risk assessment, not only in women symptomatic of preterm birth, but in asymptomatic women where prophylactic preventative therapies can be attempted.2,3
Our aim was to develop a reliable and validated tool incorporating the improved quantitative fetal fibronectin (qfFN) measurement with other relevant risk factors to predict sPTB in asymptomatic women at high risk of preterm birth (one or more of: previous sPTB or previous preterm rupture of membranes (PPROM), previous late miscarriage (16-23+6), previous cervical surgery or cervical length measuring <25 mm in the current pregnancy) and symptomatic women presenting with symptoms suggestive of preterm labour (Abdominal pain and or threatened preterm labour:>1 painful contraction every 10 minutes for at least 30 minutes).
Method
Blinded prospective data was collected from asymptomatic women attending preterm surveillance clinic or women presenting with symptoms suggestive of preterm labour. Separate analyses were performed for asymptomatic and symptomatic women separately but the statistical methodology was the same.
Parametric survival models, with time updated covariates for sPTB, were compared for combinations of predictors and the best selected using the Akaike and Bayesian Information Criteria.4,5 The model was developed in a training set of women and tested in a validation. Probabilities of delivery before 5 gestations (30, 34, 37 weeks’ and within 2 or 4 weeks of test) were compared to actual event rates. Predictive statistics were calculated to compare training and validation sets.
Results
The final model for asymptomatic women incorporated cervical length, √fFN and previous sPTB/ PPROM as predictors and qfFN and previous sPTB/PPROM were included in the model for symptomatic women.
ROC areas ranged from 0.77-0.99 and 0.77-0.88 for asymptomatic and symptomatic models respectively, which is better than previously reported6, indicating good prediction across a range of fFN thresholds. Overall sensitivity and specificity are similar to the current literature 7 but for an individual woman the risk will be more accurate using the algorithm rather than having to use summary values across a range of women. The two models have been incorporated into an algorithm which is available as a smart phone app. This is intended for clinicians caring for women at risk of sPTB or where PTL is suspected. It will enable them to accurately determine a women’s risk at various clinically important gestations and to tailor management decisions appropriately.
References:
1. Blencowe H, Cousens S, Oestergaard MZ, Chou D, Moller AB, Narwal R
Adler A, Vera Garcia C, Rohde S, Say L, Lawn JE. National, regional, and worldwide estimates of preterm birth rates in the year 2010 with time trends since 1990 for selected countries: a systematic analysis and implications. Lancet 2012; 379: 2162-72.
2. Genc MR, Ford CE. The clinical use of inflammatory markers during pregnancy Curr Opin Obstet Gynecol. 2010; 22: 116-21
3. Bolt LA, Chandiramani M, de Greeff A, Seed PT, Kurtzman J, Shennan AH. The value of combined cervical length measurement and fetal fibronectin testing to predict spontaneous preterm birth in asymptomatic high risk women. J Matern Fetal Neonatal Med. 2011; 24: 928-932.
4. Akaike, H. Information Theory and the Extension of the Maximum Likelihood Principle. In Second International Symposium on Information Theory. Petrov VN, Csaki F (eds), Akailseoniai-Kiudo: Budapest, 1973; 267 -281.
5. Schwarz G. Estimating the dimension of a model. Ann Stat. 1978; 6: 461-464
6. Honest H, Bachmann LM, Gupta JK, Kleijnen J, Khan KS. Accuracy of cervicovaginal fetal fibronectin test in predicting risk of spontaneous preterm birth: systematic review. BMJ 2002; 325:301
5. Abbott DS, Radford SK, Seed PT, Tribe RM, Shennan AH. Evaluation of quantitative fetal fibronectin test for spontaneous preterm birth in symptomatic women. Am J Obstet Gynecol. 2012; 208: 122 e1-e6.