ORIGINAL ARTICLE Improved SNAPPE-II and CRIB II scores over a 15-year period F Groenendaal 1,2 , MC de Vos 1 , JB Derks 3 and EJH Mulder 3 OBJECTIVE: During the last decades mortality and morbidity of preterm infants have declined in the Western world. We hypothesized that the decrease in mortality in preterm infants was associated with a decrease in illness severity scores (SNAPPE-II and CRIB II scores). STUDY DESIGN: Subjects were inborn infants born between January 1997 and December 1999 (period 1) and between January 2006 and December 2011 (period 2) with a gestational age of 26+0 through 28+6 weeks and without congenital malformations (n = 394). SNAPPE-II, CRIB II scores, mortality, severe morbidity and survival without morbidity were recorded. Outcomes between the two periods were analyzed using multivariable analysis. RESULTS: SNAPPE-II, but not CRIB II, scores were significantly lower for all GAs in period 2 compared with period 1. The risk of mortality for identical SNAPPE-II scores and CRIB II scores did not differ between the two periods. The risk of morbidity for identical SNAPPE-II scores and CRIB II scores was significantly lower in period 2 versus period 1. Hence, the chance of survival without morbidity for identical SNAPPE-II scores and CRIB II scores increased significantly in period 2 versus period 1. CONCLUSIONS: SNAPPE-II, but not CRIB II, scores decreased over 15 years. The risk of mortality for identical SNAPPE-II and CRIB II scores did not change, but the risk of morbidity decreased and the chance of survival without morbidity increased for identical SNAPPE-II and CRIB II scores. These findings suggest substantial improvements in both obstetrical and neonatal care. Journal of Perinatology advance online publication, 26 January 2017; doi:10.1038/jp.2016.276 INTRODUCTION Illness severity scores have been designed to predict mortality among infants in neonatal intensive care units (NICUs) and to compare mortality rates among different NICUs. 1–5 During the last two decades new versions of these scoring systems, the Score for Neonatal Acute Physiology Perinatal Extension II (SNAPPE-II) and the Clinical Risk Index for Babies II (CRIB II), have been developed. 6–8 Perinatal care has improved over the last years resulting in a decrease in mortality of preterm infants. 9–13 In The Netherlands, the neonatal mortality rate of infants born between 26 and 32 weeks of gestation decreased from 10% in 2001 to 7% in 2012 (https://www.perined.nl/producten/publicaties/ jaarboeken). Besides the decrease in mortality, a decrease in the incidence of cerebral palsy in preterm infants was observed during the last years. 14–16 The decreased mortality rates are in part attributed to optimizing care of high risk pregnant women. Although different studies have shown a decrease in mortality rates during the last decade, it is unknown whether this decrease is associated with a decrease in SNAPPE-II and CRIB II scores, indicating improvements in antenatal and immediate postnatal care. 9–13,17 In addition, it is not known whether the association between SNAPPE-II and CRIB II scores and mortality and morbidity has changed over time, thereby changing the predictive value of these scores. The aim of our study was to determine whether the decrease in mortality in preterm infants was associated with a decrease in illness severity scores, and in addition whether the association between these scores and mortality, morbidity and survival without morbidity changed over time. MATERIALS AND METHODS The study population consisted of infants with a gestational age (GA) between 26+0 and 28+6 weeks born between 1 January 1997 and 31 December 1999 and between 1 January 2006 and 31 December 2011 in the Wilhelmina Children’s Hospital/University Medical Centre Utrecht, and admitted to the level III NICU. Subjects were divided into two birth cohorts: 1997 to 1999 (period 1) and 2006 to 2011 (period 2). In period 1, the maternity ward and the NICU were located in two different hospitals in the city of Utrecht, The Netherlands. In period 2, the units had been transformed into a single unit on one location. In period 2, active treatment was recommended for infants with a GA of 25 weeks and more. Before 2006 active treatment was initiated from a GA of 26 weeks and more. 13 This change may have affected treatment of all preterm born infants. Clinical data were collected from obstetric and neonatal patient records. Infants born outside our perinatal center and infants with severe congenital malformations were excluded from the analysis. SNAPPE-II and CRIB II scores were computed based on patient characteristics, and clinical and physiological data obtained within the first 12 h after birth. SNAPPE-II, a sum score, is composed of points assigned for birth weight, small for GA ( oP3), low Apgar score at 5 min ( o7), lowest blood pressure, lowest temperature, lowest arterial pH, lowest arterial PO 2 /FiO 2 ratio, lowest urine output and seizures, according to Dammann et al. 7 Possible SNAPPE-II scores range between 0 and 162. CRIB II was calculated according to Parry et al. 8 This sum score includes points for gender, birth weight, gestation in weeks, temperature and base excess at admission. Possible CRIB II values in our cohort range between 3 and 23. Mortality was defined as death within the first 120 days, since it covers mortality during NICU admission. Severe short-term morbidity includes intraventricular hemorrhage grades III and IV, including venous cerebral infarction, 18 cystic periventricular leukomalacia grades II and III, 19 perforated necrotizing enterocolitis, bronchopulmonary dysplasia defined as supplementary 1 Department of Neonatology, Birth Center, University Medical Center Utrecht/Wilhelmina Children’s Hospital, Utrecht, The Netherlands; 2 Brain Center Rudolf Magnus, University Medical Center Utrecht, Utrecht, The Netherlands and 3 Department of Neonatology, Birth Center, University Medical Center Utrecht/Wilhelmina Children’s Hospital, Utrecht, The Netherlands. Correspondence: Dr F Groenendaal, Department of Neonatology, Birth center, University Medical Center Utrecht/Wilhelmina Children’s Hospital, Room KE.04.123.1, Lundlaan 6, Utrecht 3584, EA, The Netherlands. E-mail: F.Groenendaal@umcutrecht.nl Received 29 August 2016; revised 22 December 2016; accepted 23 December 2016 Journal of Perinatology (2017) 00, 1 – 5 © 2017 Nature America, Inc., part of Springer Nature. All rights reserved 0743-8346/17 www.nature.com/jp