|| Checking for direct PDF access through Ovid
To assess the accuracy of vaginal fetal fibronectin (FFN) as a screening test for preterm delivery in a community hospital.A prospective cohort of patients at high risk for preterm delivery at a community hospital underwent testing with FFN over a 15 month-period (March 2004-May 2005). Indications for testing were preterm labor, multiple pregnancies, cervical shortening, and cerclage. Pregnancy characteristics were retrieved on all women with positive FFN results and controls in a 1:2 ratio. Outcome variables included interval to delivery; length of hospital stay; and rates of preterm delivery <37 weeks. In the presence of serial FFN testing, only the initial result was used for calculation of diagnostic indices. Statistical analysis utilized t-test, Fisher's exact test and logistic regression analyis to control for gestational age at testing, with P < 0.05 or odds ratio (OR) with 95% confidence interval (CI) not inclusive of the unity considered significant.Two hundred and fifty seven FFN tests were performed in 230 women, of which 33 (14.3%) had positive FFN results. Duration of hospital stay was significantly shorter for patients with negative than positive results (8 h vs. 2.1 days, P = 0.011). Women with positive FFN were more likely to deliver within 14 days (OR = 6.5, 95% CI 1.4; 30.7), within 21 days (OR = 4.8; 95% CI 1.4; 16.6), before 34 weeks (OR = 5.0, 95% CI 1.7; 14.8) and before 37 weeks (OR = 3.1; 95% CI 1.3; 7.1) than women with negative results.A negative FFN result provides enough reassurance to allow shorter hospital stay. In a real-world setting (a community hospital with a population heterogenous for risk factors for preterm delivery, and in a non-protocol setting) the performance of FFN testing closely mirrors that obtained in academic institutions, where the test was studied in more uniform populations under strict protocols.The performance of vaginal fibronectin in patients with heterogeneous risk factors for preterm delivery closely mirrors that obtained in studies conducted in populations with homogeneous risk factors.