Prevention of respiratory distress syndrome: Cost-effectiveness for managing idiopathic preterm labor

Abstract

117 PREVENTION OF RESPIRATORY DISTRESS SYNDROME: COST-EFFECTIVENESS FOR MANAGING IDIOPATHIC PRETERM LABOR. ER Myers ~, JG Alvarez ~, DK Richardson ~, J Ludmir. Beth Israel Hospital, Harvard Medical School, Boston, MA. OBJECTIVE: To determine the relative cost-ett~ctiveness of three strategies for preventing respiratory distress syndrome (RDS) resulting from preterm labor and delivery~ (1) empiric tocolysis with betamimetic agonists and treatment with corticosteroids (TREATALL), (2) anmiocentesis and testing for fetal lung maturity, with treatment based on tests results (TESTALL), and (3) no treatment (TREATNONE). METHODS: We used a Markov decision analytic model to estimate the outcomes of each strategy. We took a health care system perspective and used a one-week tinle ti-allle. We assumed there were no fetal or maternal indications for delivery, and that flm patients were not diabetic. Probabilit) variables were obtained from the literature and included the probability of delivery with preterm labor and of RDS by gestational age, the efficacy of betamimetics and corticosteroids, and the sensitivity and specificity of the test for fetal lung maturity. Cost variables were obtained from the Costs and Payments Department at Beth Israel Hospital and included the costs of amniocentesis, fetal lung maturity tests, betarnimeties, corticosteroids, maternal hospital costs, costs associated with RDS and other neonatal costs. Sensitivity analysis was performed on all variables. RESULTS: The most cost-effective strategy varied with the probability of RDS. At probabilities ranging from 65% to 14%, corresponding to gestational ages fi-om 28 to 34 wks, TREATALL was the most cost-effective strategy. TESTALL was most cost-effi~ctive from 14% to 1% (34 to 36 wks) and TREATNONE at probabilities less than 1% (>36 wks). TREATALL was more highly favored as flae cost of RDS increased, while TESTAI.L was more favored as the specificity of the test increased. CONCLUSION: Although testing for fetal lung maturity is usefid in many clinical situations, the cost-effi~etiveness of such testing in the setting of idiopathic preterm labor appears to be limited to a narrow time fi-ame. 119 THE RELATIONSHIP BETWEEN DIFFERENCES IN BODY MASS INDEX AT DELIVERY AND RACIAL DIFFERENCES IN PREMATURITY RATE. E.C Lamplev ~, S.A. Myers, R. Ku~zel. Chicago Med, Dept OB/GYN Mt Sinai Hospital, Chicago, 1L. OBJECTIVE: Low body mass index (BMI) is reported to be associated with preterm deliver), (PTD). The purpose of the current study is to evaluate the relationship beiaveen maternal body mass index at delivery and the incidence of PTD in two sub-populations at the same institution. STUDY DESIGN: Patient age, race, gravity, past history of PTD, drug use, height, weight at delivery, and gestational age at delivery were obtained from the Mt. Sinai perinatal database from 1987-1994. The BMI was calculated for 14,031 black and 6,831 Hispanic patients (kg/m'2). Data was grouped according to race and stratified by gestational age at delivery, maternal age, previous preterm delivery, and drug use. The mean and standard deviation of the BMI and the proportion of patients with BMI < 25 Kg/M ~ were calculated. T-test, X ~, fisher exact were used. RESULTS: Compared to Hispanic patients, the black patients were at higher risk for preterm deliver T, 18% vs 10% (p < 0.001) and BMI < 25 Kg/M 2 21% vs 16.5% (p < 0.01). However, confounding factors were dissimilar between groups. As a result, 7,243 black patients with no prior histo~ of preterm birth, drug use, and maternal age 1%30 were compared to 1,531 Hispanics with the same history. Amongst black patients, BMI was lower for preterm than that of the term patients 27.4 + 5.9 vs 29.4 + 6.0 (p < 0.001) and patients were 1.9 times (95 CI 1.6-2.2) more likely to have a BM1 less than 25 then term patients. PTD was 1.4 times (CI 1.2, 1.7) more likely. Amongst Hispanics there was no difference in BMI between term and preterm patients, 28.6 • 7.8 vs 27.8 • 8.9 (NS). Also, preterm patients were 1.6 times (95 CI 1.0-2.5 N.S.) more likely to have a BMI <25 than the term patients. CONCLUSION: These data demonstrate that the relationship between PTD and BM1 is not the same for black and Hispanic patients in this population. Race-specific and population-specific analysis may be necessary beftire generalizations to larger populations are made.