Physiology and predictors of impaired gas exchange in infants with bronchopulmonary dysplasia

Rationale: A sensitive outcome measure for infants with bronchopulmonary dysplasia would facilitate clinical benchmarking and enhance epidemiologic understanding, evaluation of clinical interventions, and outcome prediction.

Objectives: Noninvasive assessment of pulmonary gas exchange in preterm infants with and without bronchopulmonary dysplasia to grade disease severity and to identify determinants of impaired gas exchange.

Methods: This is a prospective observational study in very preterm infants. Inspired oxygen partial pressure (PIO2) was decreased stepwise to achieve oxygen saturation as measured by pulse oximetry (Sp(O2)) that decreased from 95% to 86%. Right shift, (V) over dot(A)/(Q) over dot, and right-left shunt were derived from the resulting Sp(O2) versus PIO2 curve and compared with current disease severity classification. Potential determinants of shift, (V) over dot(A)/(Q) over dot, and shunt were identified using principal components analysis and multiple linear regression.

Measurements and Main Results: A total of 219 infants with median (interquartile range) gestation of 28 weeks and 0 days (26 weeks and 0 days to 29 weeks and 0 days) had a valid study at 35 weeks and 4 days (34 weeks and 1 day to 39 weeks and 3 days) of postmenstrual age. Shift increased and (V) over dot(A)/(Q) over dot decreased as severity of bronchopulmonary dysplasia increased. Infants with moderate-severe disease also had increased shunt. Extent of impaired gas exchange overlapped between severity groups. Infants requiring mechanical support but no supplemental oxygen at 36 weeks' postmenstrual age had similar values of shift, (V) over dot(A)/(Q) over dot, and shunt to preterm infants without bronchopulmonary dysplasia. Lower gestation and increased duration of invasive ventilation independently predicted increased shift, decreased (V) over dot(A)/(Q) over dot, and increased shunt. Shift was the most sensitive and specific index of the severity of bronchopulmonary dysplasia.

Conclusions: Most infants with bronchopulmonary dysplasia have impaired oxygenation quantified by a simple, sensitive bedside test. Shift of the Sp(O2)/PIO2 curve may be useful for prediction and measurement of preterm infant respiratory outcomes.