Three biomarker tests to help diagnose preterm labour: a systematic review and economic evaluation
Jo Varley-Campbell,Rubén Mújica-Mota,Helen Coelho,Neel Ocean,Max Barnish,David Packman,Sophie Dodman,Chris Cooper,Tristan Snowsill,Tracey Kay,Neil Liversedge,Michelle Parr,Lisa Knight,Chris Hyde,Andrew Shennan,Martin Hoyle
DOI: https://doi.org/10.3310/hta23130
Abstract:Background: Preterm birth may result in short- and long-term health problems for the child. Accurate diagnoses of preterm births could prevent unnecessary (or ensure appropriate) admissions into hospitals or transfers to specialist units. Objectives: The purpose of this report is to assess the test accuracy, clinical effectiveness and cost-effectiveness of the diagnostic tests PartoSure™ (Parsagen Diagnostics Inc., Boston, MA, USA), Actim® Partus (Medix Biochemica, Espoo, Finland) and the Rapid Fetal Fibronectin (fFN)® 10Q Cassette Kit (Hologic, Inc., Marlborough, MA, USA) at thresholds ≠50 ng/ml [quantitative fFN (qfFN)] for women presenting with signs and symptoms of preterm labour relative to fFN at 50 ng/ml. Methods: Systematic reviews of the published literature were conducted for diagnostic test accuracy (DTA) studies of PartoSure, Actim Partus and qfFN for predicting preterm birth, the clinical effectiveness following treatment decisions informed by test results and economic evaluations of the tests. A model-based economic evaluation was also conducted to extrapolate long-term outcomes from the results of the diagnostic tests. The model followed the structure of the model that informed the 2015 National Institute for Health and Care Excellence guidelines on preterm labour diagnosis and treatment, but with antenatal steroids use, as opposed to tocolysis, driving health outcomes. Results: Twenty studies were identified evaluating DTA against the reference standard of delivery within 7 days and seven studies were identified evaluating DTA against the reference standard of delivery within 48 hours. Two studies assessed two of the index tests within the same population. One study demonstrated that depending on the threshold used, qfFN was more or less accurate than Actim Partus, whereas the other indicated little difference between PartoSure and Actim Partus. No study assessing qfFN and PartoSure in the same population was identified. The test accuracy results from the other included studies revealed a high level of uncertainty, primarily attributable to substantial methodological, clinical and statistical heterogeneity between studies. No study compared all three tests simultaneously. No clinical effectiveness studies evaluating any of the three biomarker tests were identified. One partial economic evaluation was identified for predicting preterm birth. It assessed the number needed to treat to prevent a respiratory distress syndrome case with a 'treat-all' strategy, relative to testing with qualitative fFN. Because of the lack of data, our de novo model involved the assumption that management of pregnant women fully adhered to the results of the tests. In the base-case analysis for a woman at 30 weeks' gestation, Actim Partus had lower health-care costs and fewer quality-adjusted life-years (QALYs) than qfFN at 50 ng/ml, reducing costs at a rate of £56,030 per QALY lost compared with qfFN at 50 ng/ml. PartoSure is less costly than Actim Partus while being equally effective, but this is based on diagnostic accuracy data from a small study. Treatment with qfFN at 200 ng/ml and 500 ng/ml resulted in lower cost savings per QALY lost relative to fFN at 50 ng/ml than treatment with Actim Partus. In contrast, qfFN at 10 ng/ml increased QALYs, by 0.002, and had a cost per QALY gained of £140,267 relative to fFN at 50 ng/ml. Similar qualitative results were obtained for women presenting at different gestational ages. Conclusion: There is a high degree of uncertainty surrounding the test accuracy and cost-effectiveness results. We are aware of four ongoing UK trials, two of which plan to enrol > 1000 participants. The results of these trials may significantly alter the findings presented here. Study registration: The study is registered as PROSPERO CRD42017072696. Funding: The National Institute for Health Research Health Technology Assessment programme.