Souvik Mitra, Amish Jain, Joseph Y. Ting, Nadya Ben Fadel, Christine Drolet, Ayman Abou Mehrem, Amuchou Soraisham, Bonny Jasani, Deepak Louis, Anie Lapointe, Jon Dorling, Faiza Khurshid, Abbas Hyderi, Kumar Kumaran, Jennifer Toye, Andrei Harabor, Dany E. Weisz, Miroslav Stavel, Alyssa Morin, Soume Bhattacharya, Renjini Lalitha, Jehier Afifi, Sajit Augustine, Michael Castaldo, Tara Hatfield, Yi-Chen Su, Prakesh S. Shah, Canadian Neonatal Network Investigators, Marc Beltempo, Marco Zeid, Prakesh S. Shah, Thevanisha Pillay, Jonathan Wong, Miroslav Stavel, Rebecca Sherlock, Ayman Abou Mehrem, Joseph Ting, Carlos Fajardo, Andrei Harabor, Jaya Bodani, Lannae Strueby, Mary Seshia, Deepak Louis, Chelsea Ruth, Ann Yi, Amit Mukerji, Kevin Coughlin, Sajit Augustine, Kyong-Soon Lee, E Y Ng, Brigitte Lemyre, Eyad Bitar, Victoria Bizgu, Nina Nouraeyan, Keith Barrington, Anie Lapointe, Guillaume Éthier, Marie St-Hilaire, Valérie Bertelle, Édith Massé, Paloma Costa, Hala Makary, Ahmad Aziz, Gabriela de Carvalho Nunes, Wissam Alburaki, Jo-Anna Hudson, Jehier Afifi, Andrzej Kajetanowicz, Catherine Chang, Christine Drolet, Bruno Piedboeuf
Importance: Wide variation exists in patent ductus arteriosus (PDA) treatment practices in extremely preterm infants (defined in this study as born before 29 weeks' gestation). Commonly used treatment strategies include use of standard-dose and adjustable-dose ibuprofen, indomethacin, and acetaminophen as well as nonpharmacological conservative management. Objectives: To compare the relative effectiveness of different PDA pharmacotherapeutic regimens in extremely preterm infants, and to compare the clinical outcomes between infants who were treated with pharmacotherapy and those who received no pharmacotherapy (conservatively managed). Design, Setting, and Participants: This comparative effectiveness research study using clinical data from 19 tertiary neonatal intensive care units (NICUs) in Canada, which are part of the Canadian Neonatal Network, was conducted between January 1, 2020, and July 31, 2023. Infants born before 29 weeks' gestation with echocardiography-confirmed, predominantly left-right shunting, moderate- to large-sized (diameter: ≥1.5 mm) PDA were included. Infants who received PDA treatment based solely on clinical diagnosis without echocardiographic confirmation or underwent primary procedural closure were excluded. Prior to study initiation, each NICU self-selected 1 of the following 4 predefined interventions as their primary pharmacotherapy regimen: standard-dose ibuprofen, adjustable-dose ibuprofen, indomethacin, and acetaminophen. Data analyses were conducted from May 9, 2024, to February 23, 2026. Exposure: Primary PDA pharmacotherapy using 1 of the following regimens: standard-dose ibuprofen, adjustable-dose ibuprofen, indomethacin, and acetaminophen. Main Outcomes and Measures: The primary outcome was failure of primary pharmacotherapy, defined as the need for additional medical and/or procedural PDA treatment. The secondary outcomes included repeat pharmacotherapy course, surgical or interventional PDA closure, predischarge mortality, moderate to severe bronchopulmonary dysplasia (BPD), stage 2 or higher necrotizing enterocolitis (NEC), and definite sepsis. Results: A total of 1356 infants (mean [SD] gestational age [GA], 25.4 [1.6] weeks; mean [SD] birth weight, 828 [228] g; 746 males [55.0%]) were included, of whom 1097 (80.9%) received PDA pharmacotherapy and 259 (19.1%) did not. Among the participating NICUs, 7 selected standard-dose ibuprofen, 8 selected adjustable-dose ibuprofen, 1 selected indomethacin, and 3 selected acetaminophen as their primary pharmacotherapy. There was 28.1% (308 of 1097) protocol deviation, 75.0% of which was attributable to the use of acetaminophen by sites that did not intend to use acetaminophen. Among infants who received treatment, 464 (42.3%) experienced failure of the primary pharmacotherapy, and 456 (41.6%) received repeat pharmacotherapy course. There were no differences between the 4 treatment regimens in failure of primary pharmacotherapy after adjustment for confounders and accounting for clustering within each site. Compared with infants who received no pharmacotherapy, treated infants had higher odds of moderate to severe BPD (adjusted odds ratio [AOR], 1.91; 95% CI, 1.12-3.25) and NEC (AOR, 2.15; 95% CI, 1.55-2.99) but lower odds of mortality (AOR, 0.35; 95% CI, 0.21-0.58). Conclusions and Relevance: In this comparative effectiveness research study, the rate of PDA pharmacotherapy failure was high among extremely preterm infants, with no observed differences in effectiveness between commonly used pharmacotherapy regimens. Compared with conservative management, pharmacotherapy was associated with lower mortality and higher BPD and NEC odds; however, confounding by contraindication and survival bias cannot be ruled out.