Does Assessment of Cervical Phosphorylated Insulin-like Growth Factor Binding Protein-1 by Bedside Vaginal Swab Test Really Predict Preterm Birth?

Abstract

Preterm birth is the first cause of neonatal mortality and is associated with elevated risks of long-term complications such as neurodevelopmental impairment. Prediction of spontaneous preterm birth, one of the biggest challenges in obstetrics, aims at delaying birth in order to allow corticosteroid therapy and, if necessary, transfer of patient to a higher-level maternity care unit. We aimed to assess the predictive role of phIGFBP-1 (Actim® Partus) diagnostic test on patients at risk of preterm labor, routinely used in our institution. We conducted a retrospective cohort study on 99 patients admitted in the high-risk pregnancy unit of our institution from June 2012 to November 2014. The primary outcome measures were delivery before 34+0 and 37+0 weeks. Data analysis allowed measure of Actim® Partus test sensitivity (Se), specificity (Sp), positive and negative predictive values (PPV and NPV), diagnostic efficiency as well as positive and negative likelihood ratios. Actim® Partus test features (Se, Sp, PPV and NPV) were 53.3, 67.9, 23.5 and 88.7% respectively for deliveries occurring ≤\u200934+0 weeks and 54.2, 75.4, 55.8, and 74.2%, respectively, for deliveries occurring ≤\u200937+0 weeks. Diagnostic efficiency of the test was 65.7% (≤\u200934+0 weeks) and 67.7% (≤\u200937+0 weeks). Positive likelihood ratios were 1.6 (≤\u200934+0 weeks) and 2.2 (≤\u200937+0 weeks). Negative likelihood ratios were 0.7 (≤\u200934+0 weeks) and 0.6 (≤\u200937+0 weeks). Results of our study show that phIGFBP-1 diagnostic test is not accurate enough in predicting preterm birth before 34+0 or 37+0 weeks, and therefore, there is little clinical interest in its everyday use.