Effect of QUiPP prediction algorithm on treatment decisions in women with a previous preterm birth: a prospective cohort study

Abstract

Objective The QUiPP algorithm combines cervical length, quantitative fetal fibronectin (qfFN) and medical history to quantify risk of preterm birth. We assessed the utility of QUiPP to inform preterm birth prevention treatment decisions. Design A prospective cohort study with a subsequent impact assessment using the QUiPP risk of birth before 34 weeks' gestation. Setting A UK tertiary referral hospital. Sample In all, 119 women with previous spontaneous preterm birth (sPTB) or preterm premature rupture of membranes (PPROM) before 34 weeks' gestation. Methods Cervical length and qfFN were measured at 19(+0) to 23(+0) weeks' gestation. Clinical management was based on history and cervical length. After birth, clinicians were unblinded to qfFN results and QUiPP analysis was undertaken. Main outcome measures Predictive statistics of QUiPP algorithm using 10% risk of sPTB before 34(+0) weeks as treatment threshold. Results Fifteen of 119 women (13%) had PPROM or sPTB before 34 weeks. Of these, 53% (8/15) had QUiPP risk of sPTB before 34(+0) weeks above 10%. Applying this treatment threshold in practice would have doubled our treatment rate (20 versus 42%). QUIPP threshold of 10% had positive likelihood ratio (LR) of 1.3 (95% CI 0.76-2.18), and negative LR of 0.8 (95% CI 0.45-1.40) for predicting sPTB before 34(+0) weeks. Conclusions Use of the QUiPP algorithm in this population may lead to substantial increase in interventions without evidence that currently available treatment options are beneficial for this particular group. Tweetable abstract Independent study finds that the QUiPP algorithm could lead to substantial increases in treatment without evidence of benefit.