Elevated mid-trimester maternal corticotrophin-releasing hormone levels inpregnancies that delivered before 34 weeks

Objective To test whether maternal corticotrophin-releasing hormone levels are elevated in the midtrimester for those women who subsequently had spont aneous preterm delivery and to assess the clinical utility of the measureme nt in the prediction of preterm delivery. Design A prospective observational study. Setting Department of Obstetrics and Gynaecology, Prince of Wales hospital, Hong Kong. Population 1047 low risk pregnant women recruited at 15-20 weeks of gestati on. Methods Venous samples were assayed for levels of corticotrophin-releas ing hormone. The investigators responsible for the laboratory assay were bl inded to the obstetric outcome. Main outcome measures Incidence of preterm, term and post-term pregnancies. Results Those who were delivered spontaneously at a preterm gestational age (before 34 weeks) had significantly higher corticotrophin-releasing hormon e levels in the mid-trimester, compared with those who were delivered at te rm or post-term. There was a trend towards lower corticotrophin-releasing h ormone levels with more advanced gestational age at delivery. When the meas urement of corticotrophin-releasing hormone was used to predict delivery be fore 34 weeks, the best cut off was 1.9 MoM, which produced a sensitivity o f 72.7% and specificity of 78.4%. This translated to a positive predictive value of 3.6%, negative predictive value of 99.6% and relative risk of 9.4 when the background prevalence of spontaneous preterm delivery before 34 we eks was 1.1%. The likelihood ratio was 3.4. Conclusions Mid-trimester maternal corticotrophin-releasing hormone levels are elevated in pregnancies destined to deliver preterm before 34 weeks. Wh en used alone in a low risk population, the measurement has a low predictiv e power for preterm delivery. However, the likelihood ratio of 3.4 implies that in high risk populations the test may be considerably more valuable.