Objective To verify if “physiologic definition” of BPD is more useful than classic definition.

Methods Neonates with BW 500–1250 g were prospectively studied to evaluate incidence of BPD. Respiratory outcome at 28 days and/or 36 wks PMA was defined in 2 ways: a clinic one based on O₂ administration and a physiologic one based on O₂ administration and saturations. Neonates who received FiO₂ ⩽0.30 with or without CPAP were eligible. The test was divided into 3 parts, baseline, challenge and post-test: every minute we recordered ptcO₂, ptcCO₂, HR, RR, SpO₂. Presence of apnoea (breathing cessation >20 sec) and bradycardia (HR<80 bpm<10 sec) was noted. Failure was defined as SpO₂ 80–90% for 5 minutes with ptcO₂ <50 mmHg, or SpO₂ <80% for 1 minute, or occurrence of an adverse event during the test or a FiO25% increasing one hour after reduction.

Results 21 neonates were tested, 2 were examined at 28 days and 36 wks PMA.

At 28° day BPD incidence passed from 50.8% (clinical definition) to 44.2% (physiological definition), because 4/15 (27%) passed the test.

At 36 wks PMA BPD incidence passed from 21.3% to 15%, because 4/8 (50%) passed the test.

The mean ptcCO₂ was significantly higher in neonates that failed the test (p<0.05) in respect to those who overcame it both at 28 days and 36 wks PMA, while the mean SpO₂ was always lower but significant only at 28 days (p<0.05).

Conclusions In our study the test was used as a clinical guide and performed not only at 36 wks PMA as reported in the literature but also at 28 days and in neonates with CPAP. The ptcCO₂ and SpO₂ resulted in being good criteria for predicting test’s failures.