ABSTRACT. Background. The most important factor determining anatomic remodeling and permanent closure of the ductus arteriosus is the degree of ductus constriction after indomethacin treatment. Muscular constriction produces a region of ischemic hypoxia in the middle of the ductus muscle media that initiates the process of permanent closure. Previous studies have shown that infants delivered before 28 weeks' gestation, who still have evidence of ductus flow on Doppler examination (performed after the standard 3-dose course of indomethacin), have a high likelihood (>85% chance) of reopening their ductus in the future. In contrast, if there is no evidence of luminal patency on the posttreatment Doppler examination, the incidence of ductus reopening is <20%.In the following study, we examined infants who still had a patent ductus on Doppler examination after a 3-dose course of indomethacin, to identify which factors might be associated with permanent ductus closure. We hypothesized that infants who received additional doses of indomethacin after the standard 3-dose course might develop an even tighter degree of ductus constriction and increase their chance of developing permanent closure.Methods. We performed a retrospective cohort study of preterm infants (<26 6 ⁄7 weeks' gestation) who were treated with indomethacin. Between 12 and 24 hours after the third dose of indomethacin, infants were examined for the presence or absence of ductus-related signs, and an echocardiogram was performed. Infants responded to the initial 3 doses of indomethacin in 1 of 3 ways: 1) the ductus was closed clinically (absent clinical signs) with no evidence of luminal flow on Doppler examination ("clinically closed"; n ؍ 214); 2) the ductus was closed clinically, but a small amount of left-to-right luminal flow was evident on Doppler examination ("partially closed"; n ؍ 69); 3) or the ductus was open clinically and echocardiographically ("nonresponder"; n ؍ 30). Nonresponders underwent surgical ligation (n ؍ 30). Infants with a partially closed ductus formed our study population.We used a hierarchical regression model to identify which, if any, of the following factors might be associated with permanent anatomic closure in the 69 infants with a partially closed ductus: 1) gestational age, 2) exposure to antenatal steroids, 3) birth weight, 4) sex, 5) presence and severity of respiratory distress, 6) fluid administration during the first 96 hours after birth, 7) indomethacin treatment approach (prophylactic vs symptomatic), 8) year of birth, 9) use of surfactant, 10) preeclampsia, 11) chorioamnionitis, 12) bacterial septicemia, 13) necrotizing enterocolitis, or 14) duration of indomethacin treatment (standard 3-dose course vs prolonged 6-dose course). Infants who received the standard 3-dose course of indomethacin treatment were given 0.2, 0.1, and 0.1 mg/kg indomethacin during a 48-hour period. Infants who received the prolonged 6-dose course of indomethacin treatment were given a fourth, fifth, and sixth dose of 0.1 mg/kg at 2...