Am Fam Physician. 2002;66(1):148
Despite a lack of evidence showing that frequency of uterine contractions somehow predicts the risk of preterm delivery, many physicians continue to use this “common sense” indicator, allowing it to influence clinical decisions about management of preterm delivery risk. Iams and colleagues conducted a large, multicenter, observational study to determine whether the frequency of uterine contractions provides any useful data in predicting preterm delivery.
Women with a history of preterm delivery or second-trimester bleeding, who are known to be at increased risk for subsequent preterm delivery, were recruited for the study. Of 2,205 women who were initially screened, 306 were ultimately included in the study. Almost 25 percent of those screened declined to participate, and about one third of the enrollees did not comply adequately with home contraction monitoring. Starting at 22 weeks of gestation, trial participants were seen every two to three weeks until reaching 35 weeks of gestation or preterm delivery. Home monitoring of uterine contractions was done at least twice weekly and included one daytime session and one nighttime session of at least one hour each. At each follow-up visit, a speculum examination was done to collect cervical fluid for measurement of fetal fibronectin, and digital and ultrasound measurements of cervical length were done at every other visit.
The population enrolled was at high risk for preterm delivery, which occurred in 35 percent of cases. Analysis of 34,908 hours of monitoring showed an association between preterm delivery and increased frequency of uterine contractions. However, the ability to predict preterm delivery (based on a maximum frequency of four or more contractions per hour) was fairly limited, with a sensitivity ranging from 9 to 27 percent, depending on the gestational age at the time of monitoring. Fetal fibronectin presence in cervical fluid also had limited predictive value, with a sensitivity ranging from 19 to 41 percent. Shortened cervical length (25 mm or less by ultrasound or a Bishop score of 4 or higher) was the only predictor that identified at least one half of preterm deliveries (i.e., sensitivity of 50 percent or more, on average).
The authors conclude that uterine contraction frequency, fetal fibronectin assessment, and cervical length assessment are not useful screening tests for preterm delivery in asymptomatic women.
editor's note: While none of the screening tests had useful positive results, several had respectable negative predictive values. Absence of fetal fibronectin and no evidence of cervical length shortening by ultrasound or digital examination had negative predictive values of more than 90 percent across the range of gestational ages. Therefore, women with a negative result on one of these tests had less than a 10 percent chance of preterm delivery.—b.z.