Gestational Age and Birth Weight as Predictors of NICU Admission Probability and Length of Stay in a Jordanian Neonatal Cohort.

Background Predicting the length of stay (LOS) in the neonatal intensive care unit (NICU) is crucial for resource allocation and clinical management. Objective This study aimed to develop and validate a simple, practical predictive model using gestational age (GA) and birth body weight (BBW) in a Jordanian cohort. Methods This retrospective study analyzed 218 neonates admitted to the NICU between January 2022 and December 2023. Binary logistic regression was employed to predict prolonged LOS (≥7 days), multiple linear regression to estimate actual LOS, Cox proportional hazards regression to model time-to-discharge, and receiver operating characteristic (ROC) analysis to assess discriminative capacity. Subgroup analyses were performed using GA categories and small for gestational age (SGA) / appropriate for gestational age (AGA) / large for gestational age (LGA) classifications. Temporal validation was conducted on an independent 2024 cohort (n = 87). Results The cohort's mean GA was 33.15 ± 2.81 weeks, mean BBW was 2.61 ± 0.52 kg, and median LOS was seven days (IQR: 5-11 days). The multiple linear regression model was highly significant (R² = 0.71, p < 0.001): LOS = 34.12 - (0.59 × GA) - (3.85 × BBW). BBW demonstrated superior predictive performance (AUC = 0.81) compared to GA (AUC = 0.75). SGA infants had significantly longer LOS (11.8 ± 4.6 days) compared to AGA infants (7.9 ± 3.8 days, p < 0.001), with 3.4-fold higher odds of prolonged stay. Temporal validation confirmed model stability (validation R² = 0.694, AUC = 0.821). Conclusions GA and BBW are strong, quantifiable predictors of NICU LOS. The developed regression models provide a practical bedside tool for early risk stratification, potentially improving resource planning and targeted care for neonates at a high risk of extended hospitalization in Jordan.