246 - Clinical Outcomes of Neonates with Birth Asphyxia Treated with bCPAP: A 10-year retrospective cohort study from Uganda

Publication Number: 1233.246

Vimathi S. Gummalla, University of Washington School of Medicine, Seattle, WA, United States; Rebecca L. Jones, Kiwoko Hospital, LUWERO, Kampala, Uganda; Maneesh Batra, University of Washington School of Medicine, Seattle, WA, United States; James Mwangwa. Nyonyintono, Kiwoko hospital, Nakaseke, Nakaseke, Uganda; Anna Hedstrom, University of Washington, Seattle, WA, United States

Background: Birth asphyxia is a leading cause of respiratory failure and neonatal death worldwide, with the highest burden in low resource settings. Bubble CPAP (bCPAP) is increasingly used globally to treat respiratory failure and has been shown to reduce mortality in preterm neonates. Many sites in Sub-Saharan Africa, however, do not use their limited bCPAP devices for neonates with birth asphyxia, believing they are unlikely to benefit. As a result, data on outcomes among these patients is limited. Kiwoko Hospital in rural central Uganda, however, provides bCPAP for all neonates with respiratory failure, regardless of the underlying cause, making it an ideal setting for study.

Objective: We aimed to evaluate outcomes for neonates with birth asphyxia treated with bCPAP over a 10-year period. We compare these outcomes alongside those for preterm neonates, who are more commonly treated with bCPAP.

Design/Methods: This was a retrospective cohort study using a hospital dataset from a level two neonatal unit in Uganda, which uses bCPAP as the highest level of care. Our analysis included neonates with birth asphyxia and prematurity treated with bCPAP during their birth admission from 2014 to 2023. Diagnoses were made clinically by the doctor at discharge or death. We excluded neonates born outside the hospital and those transferred elsewhere. We performed t- and Mood's median tests on admission characteristics, mortality, and length of stay among survivors.

Results: 141 neonates with birth asphyxia and 541 preterm neonates were treated with bCPAP. On admission before bCPAP was initiated, 43% of neonates with birth asphyxia had an oxygen saturation < 70%, and the median respiratory severity score was 5/10 (Figure 1). Neonates with birth asphyxia had a mortality of 46% vs. 30% for preterm neonates (p < 0.001). This difference is influenced by a sharp increase in birth asphyxia mortality during the COVID-19 pandemic (Figure 2). Outside this period, birth asphyxia mortality rates remained similar year to year. Among those who survived to discharge, neonates with birth asphyxia were hospitalized for a median length of stay of 8 days vs. 25 days for preterm neonates (p < 0.0001, Figure 3).

Conclusion(s): Almost half of neonates with birth asphyxia treated with bCPAP died during admission, highlighting the severity of this illness. However, their length of stay when treated successfully was much shorter than that of preterm neonates. This suggests these patients require significantly less unit resources for treatment and could be considered for bCPAP care in low resource settings.

Figure 1. Admission assessment and clinical course of neonates treated with bCPAP from 2014-2023 by diagnosis.
Most neonates with birth asphyxia required PPV at delivery. Other markers of respiratory status such as respiratory rate and oxygen saturation on admission are similar compared to preterm neonates.

Figure 2. Mortality of neonates treated with bCPAP in Kiwoko Hospital over a 10-year period by diagnosis.
Birth asphyxia mortality rates remained similar year to year apart from 2020-2022 during the COVID-19 pandemic. During this period, pregnant women needed permission from their town council to travel to the hospital during labor, increasing risks of peripartum morbidity.

Figure 3. Length of stay of neonates treated with bCPAP in Kiwoko Hospital who survived to discharge from 2014-2023 by diagnosis.
Among patients who were discharged home alive, neonates with birth asphyxia were hospitalized in Kiwoko Hospital for a median of 8 days vs. 25 days for preterm neonates. In the graph, the bolded numbers represent the median length of stay, while the x represents the mean.