Changes in mortality rates among extremely preterm infants born before 25 weeks? gestation: Comparison between the 2005 and 2010 nationwide surveys in Japan

Background: There is no consensus as to whether the outcomes of extremely preterm infants born <25 weeks? gestation have been constantly improving. Aims: Our study aimed to clarify changes in mortality during hospitalization among extremely preterm infants. Study design: Comparison of mortality rates between the 2005 and 2010 retrospective nationwide surveys in Japan. Subjects: Extremely preterm infants born <25 weeks? gestation in Japan and registered in the nationwide surveys, 802 infants in 2005 and 797 in 2010, respectively. Outcomes: Mortality rates stratified by gestational age. Results and conclusion: Mortality rates <25 weeks? gestation decreased from 36.4% to 25.6% (difference - 10.8% [95% confidence interval {CI}: -15.3%, -6.2%]) in 2010 compared to 2005. Gestational age-specific mortality rates were lower in 2010 compared to 2005, except for 24 weeks? gestation: 66.0% vs. 50.0% (difference: -16% [95% CI: -29.8%, -21.2%]) and 45.7% vs. 25.5% (difference: -20.2%, [95% CI: -28.1, -12.3%]) at 22, and 23 weeks? gestation, respectively. After adjusting for explanatory variables, the probability of death during hospitalization in 2010 was significantly lower in infants born <25 weeks? gestation (adjusted odds ratio [aOR] 0.597 [95% CI: 0.471, 0.757], but when stratified by gestational age, it was only significant for infants born at 23 weeks? gestation (aOR 0.439 [95% CI: 0.303, 0.636]). In conclusion, the mortality rates among infants born <25 weeks? gestation have been steadily improving from 2005 to 2010 in Japan, but the practice for infants born at 22 weeks? gestation is still challenging.