The evaluation of the QUiPP app for triage and transfer (EQUIPTT)

Student thesis: Doctoral ThesisDoctor of Philosophy


Introduction Preterm delivery (before 37 weeks of gestation) is the single most important contributor to neonatal death and morbidity, with lifelong repercussions. However, the majority of women who present with preterm labour symptoms do not deliver imminently. Accurate prediction of preterm labour is needed in order ensure correct management of those most at risk of preterm birth and to prevent the maternal and fetal risks incurred by unnecessary interventions given to the majority. The QUIPP App aims to support clinical decision-making for women in threatened preterm labour by combining quantitative fetal fibronectin (qfFn) values, cervical length and significant preterm birth risk factors to create an individualised percentage risk of delivery. This PhD project describes the demand for such an app, the development process and its evaluation through a cluster randomised trial and a qualitative sub-study.
Methods This methodological development of this project includes improvement of the QUiPP app statistical algorithms,the app build process and quality assurance measures. A multi-site parallel cluster randomised-controlled trial(RCT) evaluated whether the QUIPP app reduced unnecessary management for threatened preterm labour. The13 participating centres were randomly allocated to receive either intervention or control. The composite primary outcome composed of proportion of unnecessary admission decisions and the proportion of unnecessary in-utero transfer decisions/actions. When the QUIPP app risk of delivery within 7 days is less than5%, the guidance was that interventions (e.g. admission, antenatal corticosteroids, in-utero transfer) could be withheld. The 13 participating centres were randomly allocated to receive either intervention or control.Women’s experiences of threatened preterm labour assessment were explored using self-completed questionnaires, with a subset of participants being invited to semi-structured interview. Interviews were also conducted with clinicians to explore implementation outcomes.
Results There were 1872 women in cluster analysis period (761 intervention, 1111 control), 1794 of which had sufficient data for analysis and 63/1794 (3.5%) delivered within 7 days. Unnecessary management of TPTL was 11.3% at the intervention sites versus 11.5% at control sites (OR 0.972 95% CI 0.66- 1.42). This outcome was largely driven by unnecessary admissions (10.7% vs 10.8%). Inappropriate discharge home was 0.4% versus 0.5% and there were no women who delivered <30 weeks’ gestation outside the hospital. Of women who delivered infants before 36 weeks’, 27.6% vs 20.5% received necessary antenatal corticosteroids (>24 h and <7 days from delivery). External validation of the symptomatic qfFn QUiPP algorithm confirmed its high predictive accuracy for risk of delivery within 7 days (ROC 0.898 95% CI 0.850-0.946). If the QUiPP app had been used as per protocol, unnecessary admissions or discharges would have been 7.4% vs 9.9% (AOR 0.72 95% CI 0.454-1.156). For womenseen with QUiPP scores prior to 30 weeks’ gestation, a comparison of the QUiPP 5% threshold with a treat-all policy demonstrated a significant reduction in unnecessary admissions (93.9%, CI 90.5 to 96.3). At intervention sites, uptake of QUiPP was 69%, 90% of inputted QUiPP values were correct and adherence to the 5%management guidance was 85%. Our qualitative study revealed a lack of shared decision making in TPTL practice, and minimal awareness of the significance of preterm birth or which tests were used to guide decision making. Clinicians who used the app found it easy to use and valued its simplicity, accessibility, and decision support. 
Conclusion This thesis has proven that the QUiPP Version 2 accurately identifies which women with preterm labour symptoms are at true risk of imminent delivery and who can be safely reassured. The trial has provided evidence that the QUiPP decision-support app is acceptable to clinicians and appropriate to United Kingdom (UK)maternity settings. This work has also revealed the prevalence of non-compliance with UK guidance for preterm labour. The low rates of unnecessary management of TPTL in both the EQUIPTT control and intervention sites alike, supports the safe use of quantitative fFn with or without the QUiPP app, to predict preterm birth in symptomatic women. As the research and technology evolve, future versions of the QUiPP may incorporate cervicovaginal- concentrations of new biomarkers such as host-defence peptides, microbiota, or their metabolites in order to refine prediction further. However, for QUiPP to deliver its full impact currently, it needs to be delivered with an implementation strategy that addresses the complexity of decisions around sPTB in real life settings.  
Date of Award1 Apr 2021
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorAndrew Shennan (Supervisor) & Rachel Tribe (Supervisor)

Cite this