Neonatal tetanus in undocumented migrant families in Saudi Arabia: outcomes and critical care resource utilization in a 10-year tertiary PICU cohort.

Background: Neonatal tetanus (NT) is an entirely preventable disease; however, it persists among underserved populations even within high-resource health systems. Contemporary data detailing outcomes in referral pediatric intensive care units (PICUs)-specifically regarding the intersection of low mortality and substantial resource utilization-remain limited. Methods: We conducted a retrospective cohort study of 74 neonates with clinically confirmed NT admitted to a tertiary academic PICU in Saudi Arabia between 2014 and 2023. The cases met World Health Organization clinical criteria. We evaluated admission severity (Ablett grade), presenting clinical features and baseline laboratory indices as prognostic factors. Notably, postadmission therapies were excluded to preserve baseline prognostic interpretation and minimize confounding by indication. In-hospital mortality was analyzed using Firth bias-reduced logistic regression. Furthermore, resource-utilization outcomes-including invasive mechanical ventilation (IMV) duration, PICU length of stay (LOS), and hospital LOS-were modeled using accelerated failure time (AFT) regression. Internal validation for parsimonious mortality models utilized bootstrap optimism correction. Results: In-hospital mortality was 4.1% (3/74), occurring exclusively among infants graded Ablett III at admission. Conversely, resource utilization was substantial: median (IQR) IMV duration was 25.0 (16.0-30.0) days, PICU LOS was 32.0 (20.0-37.0) days, and hospital LOS was 40.0 (27.3-57.5) days. In parsimonious admission-only Firth models, higher birth weight showed an inverse association with lower mortality [adjusted odds ratios (aOR) 0.35 per SD, 95% confidence interval (CI) 0.09-0.85], while admission C-reactive protein showed a risk-increasing trend (aOR 1.81 per SD, 95% CI 0.87-3.92); given the fact that there were only three deaths, these estimates should be interpreted as directional signals rather than definitive independent predictors. In AFT models, septic shock on admission was associated with a shorter observed IMV duration (time ratio 0.05, 95% CI 0.002-0.947), most consistent with a mortality-related truncation of observed duration rather than accelerated clinical recovery. At 3-6 months, neurodevelopmental follow-up was available for 71 survivors, of whom 10 screened positive for developmental concerns (10/71; 14.1%). Conclusion: In this tertiary PICU cohort of neonatal tetanus in Saudi Arabia, mortality was low but critical-care utilization and early morbidity were substantial. Admission severity and inflammatory burden showed higher-risk signals, whereas birth weight showed a protective association; however, the small number of deaths limits precise inference. These findings quantify the PICU burden of a preventable disease and reinforce prevention priorities, including maternal vaccination, antenatal care, and safe delivery practices.