Stress-induced gastric findings in critically ill newborn infants: frequency and risk factors

Objective: To evaluate whether infants treated in neonatal intensive care u nits have stress-induced bleeding from gastrointestinal tract or gastric le sions and to define risk factors for these findings. Design: Part one: retr ospective; part two: prospective. Setting: Tampere University Hospital, neo natal intensive care unit. Patients and interventions: In part one, 100 con secutive newborn infants treated in intensive care were retrospectively eva luated for gastrointestinal tract bleeding and risk factors, and in part tw o 89 gastroscopied and mechanically ventilated infants were prospectively e valuated for further risk factors for gastric mucosal lesions. The statisti cal evaluation of risk factors was made by multivariate analysis using logi stic regression modeling. Main results: Of infants treated in the neonatal intensive care unit 20 % had signs of gastrointestinal bleeding. Mechanical ventilation was the only risk factor (OR = 4.06, 95 % confidence interval 1.21-12.3). In part two, when mechanically ventilated infants were prospect ively evaluated, 53 % had remarkable gastric mucosal lesions. The analysis showed three other risk factors: abnormal and delayed delivery and hypotens ion after birth. Conclusions: Newborn infants treated in the intensive care unit had a high frequency of stress-induced gastric hemorrhage with gastri c lesions similar to adults and children treated in intensive care. Mechani cal ventilation is the main risk factor. Also mode of delivery and hypotens ion after birth increase the risk of stress-induced gastric lesions. These infants should be the target for prophylactic gastroprotective treatment.