Preventability and contributing factors in neonatal deaths in protracted refugee camp settings: A five-year audit among Syrian refugees in Jordan

Background Individual-level death audits can identify modifiable failures that routine mortality statistics miss. We analyzed 5 years of neonatal death audits from the Zaatari and Azraq refugee camps to characterize health-system delays, contributing factors, and preventability. Methods This individual-level analysis included 177 audited neonatal deaths from January 1, 2021, through December 31, 2025. Structured audits used caregiver interviews, maternal and neonatal records, antenatal care (ANC) cards, and provider discussions, followed by sequential independent clinical review and UNHCR validation. Contributing factors were organized using an adapted Three Delays operational taxonomy, and preventability was assigned through the audit review and adjudication process. Results Concurrent Delay 1 and Delay 3 factors were recorded in 164 deaths (92.7%). Although 99.4% of mothers received ANC, suboptimal ANC was documented in 70.6% of audited deaths, poor management of high-risk pregnancy in 57.6%, and clinician non-adherence to standards in 46.3%. Term neonates accounted for 30.1% of deaths and normal-birth-weight neonates for 30.4%. Seventy-four deaths (41.8%) were classified as definitely preventable, 60 (33.9%) as probably preventable, and 43 (24.3%) as possibly preventable; no deaths were classified as unlikely or not preventable. Conclusions Near-universal ANC attendance coexisted with frequent audit-identified deficiencies in ANC and high-risk pregnancy management, highlighting an access-quality gap. Three-quarters of audited deaths were classified as definitely or probably preventable through the audit process. These findings identify priorities for quality improvement but do not establish causal effects or the comparative impact of quality versus service coverage.