PIH3 VARIATIONS IN DURATION OF MECHANICAL VENTILATION IN VERY LOW BIRTH WEIGHT INFANTS IN JAPAN

Abstract

Respiratory care is one of the essential cares in very low birth weight (VLBW) infants. Insufficient care sometimes leads to chronic lung disease, which influences their health and quality of life for the rest of their lives. Duration of initial mechanical ventilation (MV) in VLBW was considered as a candidate for quality indicators. However, an institutional variation of risk-adjusted initial MV duration was not assessed in Japanese setting. The objective of this study was to investigate the variation in VLBW infants. Data pertaining to VLBW infants (birth weight <1500) who stayed at neonatal intensive care unit between April 2014 and March 2016 were retrieved from the Japanese national inpatient database and retrospectively analyzed. Duration of MV was calculated from the database. Hospital-based risk-adjusted duration of initial MV was estimated by the multiple regression analysis. We compared duration of initial MV among hospitals adjusted for patient-level factors. We identified 8905 VLBW infants from the database. Individual-level characteristics are as follows: i)male:female ratio; 52:48 , ii) mean(sd) birth weight; 996.2g(308.0g), and iii) mean(sd) birth week; 28(3). Overall mean(sd) duration of initial MV; 37.6days(39.5days). Duration of initial MV was significantly affected by several patient factors such as birth week, sex, chromosomal abnormalities, and small for gestational age. There were 210 hospitals covering these patients after excluding hospitals treating <5 VLBW infants. The risk-adjusted duration of initial MV varied widely among the hospitals; showing bimodal distribution (two peaks; 18 days and 24 days). The proportions of hospitals with risk-standardized mean initial duration of MV categories of <=14days, 15-18days, 19-22days, 23-26days, 27-30days, and 32>=days were 11.0%, 16.7%, 16.7%, 29.0%, 18.6%, and 8.1%, respectively. Our results suggest that there may be institutional difference in therapeutic strategies for respiratory care in VLBW infants in Japan. More efforts are needed to provide further validation.