Reliability of a Noninvasive Measure of V./Q. Mismatch for Bronchopulmonary Dysplasia

RATIONALE: Currently used definitions of bronchopulmonary dysplasia (BPD) lack a continuous measure of disease severity. OBJECTIVES: To determine if an indirect measure of V./Q. mismatch is reliable when simplified to facilitate more widespread use for grading disease severity in BPD at 36 weeks postmenstrual age. METHODS: We used prospectively collected data from 32 preterm infants undergoing an oxygen reduction test at 36 weeks postmenstrual age to perform a simplified indirect assessment of V./Q. mismatch for each infant. Independent raters applied the model, and interrater reliability for a quantitative measure of mismatch was measured by intraclass correlation coefficient. A receiver operating characteristic curve evaluated the impact of increasing degrees of V./Q. mismatch on diagnosing BPD as defined by oxygen reduction test failure. MEASUREMENTS AND MAIN RESULTS: Concordance for the quantitative measure of V./Q. mismatch between independent raters improved from 0.72 (confidence interval [CI], 0.48-0.86) to 0.93 (CI, 0.87-0.96) after refinement of instructions for applying the simplified model. Higher degrees of mismatch were increasingly predictive of oxygen reduction test failure, with a receiver operating characteristic curve analysis area under the curve of 0.83 (CI, 0.68-0.99; P = 0.03). CONCLUSIONS: A simplified indirect measure of V./Q. mismatch for diagnosing and grading disease severity in BPD has high reliability and can be performed with data obtained during a standard oxygen reduction test. This should facilitate more widespread investigation of this model as a technique for characterizing BPD severity.