Journal of Perinatology (2014), 1–7 © 2014 Nature America, Inc. All rights reserved 0743-8346/14 www.nature.com/jp
ORIGINAL ARTICLE
Neonatal outcome of extremely preterm Asian infants ⩽ 28 weeks over a decade in the new millennium P Agarwal, B Sriram and VS Rajadurai OBJECTIVE: To evaluate predischarge neonatal mortality and morbidity and associated risk factors in extremely preterm Asian infants ⩽ 28 weeks, over a decade, so as to facilitate formulation of perinatal guidelines and counseling. STUDY DESIGN: Cohort study of 887 liveborn extremely preterm neonates between 2000 and 2009 at KKH, the centralized perinatal center in Singapore. Outcome measures were predischarge mortality, presence of one or more major neonatal morbidities and the composite outcome of mortality or neonatal morbidity. RESULT: Overall survival to discharge was 709/887 (80%) and was significantly higher with increasing gestational age (GA) (19% at 23 weeks to 93% at 28 weeks, P o 0.001). Survival remained unchanged between 78 and 86% during the decade with no significant secular trend. Overall incidence of major morbidities were bronchopulmonary dysplasia (29%), late onset sepsis (23%), severe retinopathy of prematurity (21%), Grade 3 to 4 intraventricular hemorrhage (12%) and necrotizing enterocolitis ⩾ Bells’ stage II/focal intestinal perforation (9%). Composite morbidity was seen in 465/835 (56%) neonatal intensive-care unit admissions, decreased with increasing GA (P o 0.001; odds ratio 0.65 (95% confidence interval 0.56 to 0.75) and was independently predicted by birth weight, Clinical Risk Index for Babies–revised version II score, male gender, presence of patent ductus arteriosus and airleaks. CONCLUSION: Although there was no significant trend in neonatal survival or composite morbidity over the decade, improved survival and morbidity were seen with increasing GA. Journal of Perinatology advance online publication, 5 February 2015; doi:10.1038/jp.2014.205
INTRODUCTION There has been a significant improvement in the survival of extremely preterm and extremely low birth weight infants in the decades of the eighties and nineties. This improvement is a reflection of the advances in antenatal, perinatal and neonatal care including increasing use of antenatal steroids, surfactant therapy, and improved respiratory and nutritional management. The neonatal research network from the National Institute of Child Health and Human Development (NICHD), USA has studied the trends in survival and suggested that a plateau in the neonatal survival of very low birth weight (VLBW) infants has been reached in the modern era.1 Concerns also remain that high morbidity may replace the higher mortality of earlier eras in extremely low gestational age (ELGA) newborns.2 It has been shown that major neonatal morbidities are associated with higher incidence of adverse neurodevelopmental outcome.3–5 Traditionally the neonatal outcomes of preterm infants have been reported according to birth weight among VLBW infants. However, there may be some inaccuracy in this because of inclusion of more mature intrauterine growth-restricted infants. Second, birth weight is available only after delivery and thus not useful for antenatal counseling and prognostication. There is a paucity of Asian data reporting neonatal outcomes in the extremely preterm infants. This data, regarding risk of pre discharge mortality and major morbidity, is commonly needed by neonatologists, obstetricians and parents for appropriate decision making.3–5 The primary aim of this study was to evaluate GA-specific neonatal mortality and morbidity in preterm VLBW infants born between 23 and 28 weeks GA in the single largest
tertiary perinatal center in Singapore over a decade in the new millennium. We also studied the trends in mortality over the decade and analyzed the predictive risk factors associated with mortality and morbidity in a large cohort of extremely preterm Asian infants. METHODS Study population The study was conducted at KK Women’s and Children’s Hospital (KKH), Singapore which is the largest tertiary referral perinatal center in Singapore with an annual delivery rate of approximately 12 000 per year. KKH provides care for more than two thirds of infants born at the threshold of viability (GA o 25 weeks) in Singapore who are predominantly of Chinese, Malay or Indian ethnicities (unpublished national network data). The study period was from January 2000 to December 2009 and the infants were part of the high-risk VLBW database. Perinatal/ neonatal data of all liveborn infants born o 29 weeks gestation and/or admitted to the KKH neonatal intensive-care unit (NICU) were collected. Data on stillbirth and miscarriage were not collected. The hospital follows a structured resuscitation policy wherein the limit of viability is set at 23 weeks gestation and no resuscitation is instituted except at specific parental request and after appropriate counseling. At 24 weeks gestation, resuscitation is commenced unless parents indicate otherwise. From 25 weeks onwards, all babies are resuscitated and intensive care offered. Babies who were born o25 weeks gestation and provided with comfort care were excluded from subsequent analysis of neonatal outcomes and morbidities, and analysis was restricted to babies provided with NICU admission and care. Comfort care was provided to babies born alive at threshold of viability and where after neonatal counseling families had chosen not to provide intensive care. Comfort care included palliative care and keeping the baby warm till demise.
Department of Neonatology, KK Women’s and Children’s Hospital, Singapore. Correspondence: Dr P Agarwal, Department of Neonatology, KK Women’s and Children’s Hospital, 100 Bukit Timah Road, Singapore 229899, Singapore. E-mail: [email protected] Received 14 May 2014; revised 5 September 2014; accepted 9 October 2014
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
2 Table 1.
Perinatal/neonatal demographics and delivery room information according to gestational age
Maternal/perinatal Race Chinese Malay Indian Others PIH PROM 424 h Chorioamnionitis Antenatal steroidsa Cesarean section delivery
23 weeks
24 weeks
25 weeks
26 weeks
27 weeks
28 weeks
Total
Significance OR (95% CI)/difference in mean
n = 23 (%)
n = 105 (%)
n = 125 (%)
n = 161 (%)
n = 194 (%)
n = 227 (%)
n = 835 (%)
P-value
8 (35) 14 (60) 1 (4) — — 6 (26) 17 (74) 18/47 (38) 1 (4)
57 (54) 73 (78) 100 (62) 122 (63) 135 (60) 495 (59) 36 (34) 26 (21) 33 (21) 35 (18) 35 (25) 200 (24) 17 (7) 15 (12) 19 (12) 21 (11) 21 (9) 84 (10) 3 (5) 11 (9) 9 (6) 16 (8) 16 (8) 54 (7) 8 (8) 11 (9) 24 (15) 30 (16) 41 (18) 114 (14) 38 (36) 50 (40) 60 (37) 79 (40) 61 (27) 292 (35) 66 (63) 73 (59) 67 42) 90 (47) 73 (33) 386 (47) 78/117 (67) 94/127 (74) 117/164 (71) 156/194 (80) 164/227 (73) 627/876 (72) 34 (32) 43 (34) 96 (60) 108 (56) 142 (63) 424 (51)
*
Neonatal Birth weight (mean ± 2s.d.)g 569 ± 64 Male 16 (70) SGA 1 (4) Delivery room resuscitation 23/47 (49) with intubationa 5 min Apgar score ⩽ 5 10 (43) CRIB II 17 ± 1.5
2.41 (1.96–2.94)*** 0.84 (0.74–0.96)* 0.59 (0.51–0.68)*** 1.54 (1.34–1.77)*** 2.38 (1.95–2.67)***
672 ± 86 769 ± 122 834 ± 163 955 ± 172 1057 ± 202 885 ± 216 96.7 (89.06–104) 62 (59) 76 (61) 82 (51) 108 (56) 123 (54) 467 (56) 0.82 (0.72–0.93 )** 5 (5) 10 (8) 28 (17) 27 (14) 38 (17) 109 (13) 7.66 (5.13–11.4)*** 99/117 (84) 104/127 (82) 130/164 (79) 121/194 (62) 106/227 (47) 583/876 (67) 0.91 (0.79–1.04) 16 (15) 15 ± 1.4
26 (21) 13 ± 1.4
13 (8) 11 ± 1.3
14 (7) 10 ± 1.3
16 (7) 9 ± 1.5
95 (12) 11.3 ± 2.8
0.71 (0.58–0.86)*** (−1.29 to − 1.14)***
Abbreviations: CI, confidence interval; CRIB II, Clinical Risk Index for Babies–revised version II; OR, odds ratio; PROM, premature rupture of membranes; SGA, small for gestational age. *P ⩽ 0.05, **P ⩽ 0.01, ***P ⩽ 0.001. P-values and OR (95% CI) were determined for differences according to gestational age, with adjustment for birth weight and year of birth. aComfort care cases included in the denominator for antenatal steroids and delivery room resuscitation. For all other variables comfort care cases were excluded from analysis.
This study was approved by the institutional review board of the hospital. Assessment of GA was done by the neonatologist based on the best obstetric estimate, date of last menstrual period and early dating scan. If no antenatal data were available, GA estimation was done by postnatal– neonatal assessment. Lubchenko’s growth charts were used to classify GA status through 2004 and Fenton’s growth charts were used beginning in 2005.6,7 Small for GA (SGA) was defined as birth weight less than the tenth centile for GA. Maternal obstetric data collected included data on antenatal steroid use, premature rupture of membranes 424 h, chorioamnionitis, pre-eclampsia and mode of delivery. Antenatal steroid therapy was considered to be complete if two doses of steroids were given to the mother at least 24 h prior to delivery and as no treatment if no steroids or partial course of steroids was given. Chorioamnionitis was diagnosed on either presence of clinical findings such as maternal fever, maternal/fetal tachycardia, uterine tenderness or foul smelling discharge or histopathological findings.8,9 On the basis of the American College of Obstetricians and Gynecologists guidelines, pre-eclampsia was defined as systolic blood pressure of ⩾ 140 mm Hg and diastolic blood pressure of ⩾ 90 mm Hg after 20 weeks of gestation in the presence of proteinuria (⩾1+ or ⩾ 0.3 g per day).10 Neonatal data on delivery room resuscitation, treatment modalities and complications were collected prospectively. Intensive delivery room resuscitation included endotracheal intubation, cardiac massage and use of adrenaline and excluded use of mask ventilation or free-flow oxygen.
Neonatal outcomes The Clinical Risk Index for Babies–revised version II (CRIB II) score is a riskadjusted instrument which includes birth weight, GA, sex, admission temperature and base deficit and is based on data available soon after birth. This tool is commonly used to assess neonatal illness severity in preterm VLBW infants and in predicting in hospital neonatal mortality, as well as neurodevelopmental impairment at 36 months.11–13 Surfactant was used as early rescue therapy for infants with respiratory distress syndrome which was defined by the presence of clinical and radiological features, along with the need for respiratory support in the first 24 h of life. A hemodynamically significant patent ductus arteriosus (PDA) on twodimentional echocardiography was treated with intravenous indomethacin or surgical ligation if medical treatment was unsuccessful or contraJournal of Perinatology (2014), 1 – 7
indicated. Chronic lung disease was defined as oxygen dependency at 36 weeks postmenstrual age. 14 Rescue steroids were used based on standard departmental protocol beyond 3 weeks of age to facilitate extubation in intubated ventilatordependent neonates with oxygen requirement 440% and significant radiological features of persistent lung disease. Serial cranial ultrasounds were done according to unit protocol at days 1, 3, 7, 14, 30, 60 and at term. The Papile grading system was used to grade intraventricular hemorrhage (IVH) while a standard definition of cystic periventricular leukomalacia (PVL) was used.15 Presence of grade III/IV IVH or PVL was considered to be a major ultrasonographic abnormality. A pediatric ophthalmologist examined surviving infants at 4 to 6 weeks postnatal age and established presence and staging of retinopathy of prematurity (ROP) according to the international classification.16 Stage III ROP or presence of plus disease was considered to be severe and treated with cryotherapy/laser photocoagulation when indicated. Necrotizing enterocolitis (NEC) was defined according to modified Bell’s criteria and ⩾ stage II was considered to be significant. Presence of focal intestinal perforation was also grouped together with NEC ⩾ stage II as a significant gastrointestinal morbidity.17 Nosocomial sepsis was defined as blood culture-positive sepsis occurring beyond 72 h of life whereas perinatally acquired early onset sepsis was defined as blood culture being positive o 72 h of life. Among infants admitted to the NICU, the major morbidities with potential long-term sequelae were cystic PVL, IVH III/IV chronic lung disease, severe ROP and NEC ⩾ 2 A/focal intestinal perforation and blood culture-positive sepsis. The presence of any one of the above morbidities was considered as a major morbidity.3–5,18,19 As mortality/neonatal morbidity may be competing outcomes in these highly immature infants, the presence of either mortality or morbidity as defined above was considered to be a composite morbidity. Statistics. Data were analyzed using SPSS for Windows, Version 16.0 (SPSS, Chicago, IL, USA). Categorical variables were compared using χ2-test or Fisher’s exact test where appropriate, whereas continuous variables were compared using Student’s t-test. Birth weight and GA are both strong and important predictors of neonatal outcome, but GA was moderately correlated with birth weight (Pearson r = 0.65). Thus we transformed GA variable into a categorical variable with GA ⩽ 25 weeks and GA 425 weeks as GA was our exposure © 2014 Nature America, Inc.
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
3 of interest in this study, and we wanted to examine the risks associated with GA in our cohort of VLBWs. CRIB II score was not included in the multivariate analysis as it was highly correlated with birth weight (Pearson r = 0.75) and GA (Pearson r = 0.86). Survival trends were analyzed using Cox regression models. The incidence of composite morbidity was evaluated for each GA. Crude and adjusted odds ratios would be reported for the identified predictive factors using logistic regression models. Year of birth was included in the multivariate model to look for a significant trend over the 10-year study period. Results are presented as odds ratio and 95% confidence intervals (95%). For the multivariate analysis, only the significant factors are presented in the table. All analysis performed were two tailed and the statistical significance was set at Po0.05.
RESULTS Between January 2000 and December 2009, 887 live born infants o29 weeks and o 1500 g were admitted to the KKH NICU or born alive. Of these 887 neonates, 11 (1.2%) were born at 22 weeks gestation and they were excluded from the data analysis. The mean birth weight of the 876 study neonates was 873 ± 220 g with a mean GA of 26.1 ± 1.5 weeks. Antenatal, maternal and neonatal demographics Perinatal and neonatal demographic data and neonatal outcomes are presented in Table 1 to show differences in maternal complications and neonatal demographic data by GA at birth having adjusted for birth weight and year of birth. Complete course of antenatal steroids was given in 72% of the mothers with 431 (49%) of the neonates being delivered by cesarean section and 583 (67%) being intubated for delivery room resuscitation. The rate of antenatal steroid use increased from 38% at 23 weeks, to 67% at 24 weeks and 72 to 80% between 25 and 28 weeks. There was also a sharp rise in the rates of cesarean section deliveries between 23 and 24 weeks (2% versus 30%) and between 25 and 26 weeks (35% versus 60%). Premature delivery due to maternal PIH and the proportion of small for GA neonates were significantly lower below 26 weeks. Overall 67% of the infants needed or were provided with intubation in the delivery suite. Half of the infants delivered at 28 weeks and a third at 27 weeks did not need intubation in the delivery room, whereas 79 to 85% of infants between 24 and 26 weeks needed intubation at birth. Apgar scores at 5 min were recorded only in the neonates provided with active resuscitation and thus the 41 neonates provided with comfort care were excluded. Incidence of Apgar score o5 at 5 min was significantly higher below 26 weeks. CRIB II scores were available for 723 of the neonates and scores decreased with increasing GA, but there was no significant change in the mean CRIB scores between 2000 and 2009 and ranged from 10.7 ± 2.7 to 11.9 ± 3.02 over the study period. Neonatal outcomes Trends in survival. Overall, 709 (81%) of the 876 infants survived to discharge in our entire cohort with a linear relation between survival and increasing GA. After adjusting for birth weight and year of birth, survival improved significantly with increasing GA at odds ratio 1.50 (95% confidence interval 1.28 to 1.74; P o0.001). Proportion of VLBW survivors increased from 9/47 (19%) at 23 weeks to 93% at 28 weeks (Figure 1). There was no increase in survival over the 10 years, either for the total cohort or for any GA group, and survival remained unchanged at 81.6% with interquartile range between 77 and 87% during the decade (Data not shown). Forty-one (5.5%) infants were born alive, but provided comfort care. The mean GA and birth weight of this group was significantly lower than the 835 babies who received NICU care, 23.6 ± 0.89 versus 26.2 ± 1.46 weeks and 608 ± 123 versus 885 ± 215 g, respectively (P o 0.001). Seventy-nine (9%) babies © 2014 Nature America, Inc.
Figure 1. Neonatal survival, mortality rates and comfort care according to gestational age.
died in the neonatal period and there were 47 (6%) late deaths in the post-neonatal period before hospital discharge. The respiratory morbidities and need for ventilator support in NICU admission are outlined in Table 2. Of the 835 babies who were provided NICU care, 591 (71%) needed surfactant therapy with significantly lower rates with each increasing week of gestation (Table 3). Only 102/835 (12%) of the cohort could be managed with nasal continuous positive airway pressure (CPAP) support alone with significantly increasing proportions as GA increased (Po 0.001). No baby at 23 or 24 weeks could be managed with CPAP support only. Of the 735 neonates who survived till 36 weeks postmenstrual age, 213(29%) had chronic lung disease and its incidence along with the use of postnatal steroids and home oxygen therapy were inversely proportional to GA at birth. Rates of other major neonatal morbidities were inversely proportional to GAs and are shown in Table 3. Of the 835 extremely preterm neonates provided with NICU care, six (0.7%) died within first 6 h before cranial ultrasounds could be performed. All the remaining neonates had a cranial ultrasound done with 451 (54%) being normal, 196 (24%) Grade I IVH, 84 (10%) Grade II, 53 (6%) Grade III IVH and 44 (5%) Grade IV IVH. PVL was seen in about 3% of the study cohort. Rates of severe IVH were greatest o25 weeks GA and then decreased with increasing GA. Rates of early and nosocomial late onset sepsis were 3% and 24%, respectively, for the whole cohort and the highest rates of late onset sepsis were seen below 25 weeks. Four hundred and forty-three (53%) of the study cohort needed medical treatment with indomethacin for a significant PDA whereas 114 (14%) additionally needed surgical ligation of the duct. The need for surgical ligation was highest at 23 and 24 weeks of GA. Twenty babies died prior to evaluating the duct. There were no significant differences in the incidence of NEC/focal intestinal perforation according to GA distribution. Of the 756 neonates who survived beyond 28 days, 19 died before the day of the first scheduled eye review and ophthalmological data were available for 731/737 (99%) infants. Three hundred and fifty-four (48%) had no ROP, 219 (30%) stage I and II ROP and 158 (21%) had severe ROP. Eightytwo (11%) of the babies needed laser treatment for the ROP. Neonatal morbidities were thus seen in 20 to 30% of the cohort as seen in Tables 3 and 4 and major morbidity comprising any of the previously defined morbidities was present in 439 (52%) of the patients. Composite morbidity comprising of either a death/major morbidity was seen in 465/835 (56%) neonates provided with NICU care and was inversely proportional to GA (odds ratio 0.65 (95% confidence interval 0.56 to 0.75; P o0.001; Table 3). Journal of Perinatology (2014), 1 – 7
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
4 Table 2.
Pulmonary morbidities according to gestational age in NICU admissions 23 weeks
24 weeks
25 weeks
26 weeks
27 weeks
28 weeks
Total
Significance adjusted OR (95% CI)
n = 23 (%)
n = 105 (%)
n = 125 (%)
n = 161 (%)
n = 194 (%)
n = 227 (%)
n = 835
P-value
Respiratory distress syndrome Surfactant therapy Air leaks CPAP support only Postnatal steroids
20 (87) 20 (87) 6 (26) 0 8/13 (62)
97 (94) 95 (91) 19 (19) 0 41/82 (50)
107 (86) 129 (81) 132 (68) 108 (86) 115 (71) 128 (66) 20 (16) 17 (11) 15 (8) 3 (2.4) 10 (6) 27 (14) 39/106 (37) 29/148 (20) 18/184 (10)
138 (61) 125 (55) 15 (7) 62 (27) 12/220 (6)
623 (75) 0.73 (0.62–0.86)*** 591 (71) 0.74 (0.64–0.86)** 92 (11) 0.86 (0.71–1.04) 102 (12) 1.72 (1.28–2.27)*** 147/753 (19.5) 0.68 (0.57–0.80)***
Survivors at PMA 36 weeks CLD Home O2
n = 12 10 (83) 4/10 (40)
n = 72 49 (68) 8/69 12)
n = 103 52 (50.5) 10/99 (10)
n = 219 25 (11) 2/212 (1)
n = 735 213/735 (29) 75/735 (10.2)
n = 145 46 (32) 8/139 (6)
n = 184 31 (17) 3/176 (1.7)
0.69 (0.59–0.81)*** 0.69 (0.52–0.94)**
Abbreviations: CI, confidence interval; CLD, chronic lung disease; CPAP, continuous positive airway pressure; NICU, neonatal intensive-care unit; OR, odds ratio; PMA, postmenstrual age. *P ⩽ 0.05, **P ⩽ 0.01, ***P ⩽ 0.001. P-values and OR (95% CI) were determined for differences according to gestational age, with adjustment for birth weight, year of birth. Comfort care cases were excluded from analysis.
Risk factors for composite morbidity (mortality/major morbidity) An analysis of risk factors for composite morbidity using multiple logistic regression showed GA o 26 weeks, male gender, presence of airleaks, significant PDA and respiratory distress syndrome were associated with higher odds for an adverse outcome, while increasing birth weight and cesarean delivery was associated with lower odds to an adverse outcome 5 min Apgar score ⩽ 5, hypotension in the first 72 h needing inotropes, small for GA were not independently predictive of composite morbidity though they were significant in univariate analysis (Table 4). The CRIB II score was not included in the multivariate analysis in view of the high multicollinearity with GA and birth weight with a Pearson’s of 0.75 for birth weight and 0.86 for GA. Year of birth was not significantly associated with an adverse outcome nor was there any protective effect of antenatal steroids. DISCUSSION In our Asian cohort of VLBWs, we found associations between higher survival and increasing GA. Lower birth weight, higher CRIB-II, male gender, presence of airleaks, significant PDA and respiratory distress syndrome were associated with higher odds to infant mortality or morbidity. However, cesarean delivery was protective against neonatal mortality and morbidity. Both birth weight and GA are considered to be predictors of outcome and traditionally the National Research Network of the NICHD has published outcomes of preterm babies based on birth weight.1,20 However, the inclusion of severely growth-restricted preterm infants may affect the reliability of the data. In addition, the birth weight is only known after birth and may thus be of limited use in perinatal counseling and parental decision making prior to delivery. GA based outcomes overcome these limitations and are therefore becoming increasingly the standard of reporting for outcomes of extremely preterm newborns.21–25 It is also important to evaluate trends in morbidity at extremely low gestation age to assess whether decreasing mortality is being replaced by increasing morbidity. This has been done in recent publications from EPICURE 2 and other international data.21,26–28 Antenatal, perinatal and delivery room data The rate of antenatal steroid use in our cohort was similar to the rates reported in the NICHD by Stoll et al. 21 whereas it was lower than the 85 to 94% usage reported from Norway29 A doubling of the rate of steroid use between 23 and 25 weeks from 38 to 74% was seen in our study though this was still lower than the 53% antenatal steroid use at 23 weeks reported by the NICHD.21 This Journal of Perinatology (2014), 1 – 7
lower use of antenatal steroids at 23 weeks was a strong reflection of our institutional policy where steroids were only recommended at 23+5 weeks of gestation and beyond, in keeping with the hospital guidelines of resuscitation and intensive care only at 24 weeks.The higher use of antenatal steroids and higher cesarean section rates beyond 24 weeks probably reflects the willingness of the obstetric and neonatal teams to provide active perinatal care only after this gestation. Our rates of cesarean section were very similar to those presented in the national data from Norway between 23 and 27 weeks with no baby at 23 weeks being delivered by cesarean section.29 In a survey of 1193 members of the American College of Obstetricians and Gynecologist regarding practices at the limits of viability it was seen that only 10% of respondents would perform a cesarean section for fetal distress at 23 weeks and 58% considered 24 weeks to be the earliest age of viability.30 Preterm delivery secondary to maternal pregnancy induced hypertension increased with increasing GA and probably reflected the perinatologist’s attempt at optimizing balance between maternal interests and fetal maturity in deciding timing of delivery. Rates of delivery room intubation were significantly lower at the extremes of gestation in our cohort. The lower rates of delivery room intubation below 24 weeks reflected the resuscitation policy for thresholds of viability in our department and were similar to the NICHD data.21 Conversely the lower rates of delivery room intubation at 27 and 28 weeks in our babies probably reflected their better physiologic condition at delivery and the widespread acceptance of early CPAP use in the delivery room in our center. Survival Overall survival in our extremely preterm cohort was at 80% with a significant reduction in mortality and improving survival noted with each increasing week of gestation from 19% at 23 weeks to 93% at 28 weeks. There was, however, no progressive reduction in mortality for the whole cohort, as well as at each individual gestational week over the 10 years of the study period. Similar findings with no improvement over time have also been reported by the neonatal research network in the NICHD between 2003 and 2007.21 The authors of the Epicure 2 study also did not reveal a significant improvement in survival at 22 to 23 weeks between 1997 and 2006 and this possibly reflects the biologically plausible limit to which the boundaries can be pushed by technological and medical advance.24–26 It could also be a reflection of the restrictive resuscitation policy at 23 weeks which is a common practice internationally in units in North America, UK and Europe.31–33 This © 2014 Nature America, Inc.
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
5 Table 3.
Other neonatal morbidities according to gestational age in NICU admissions
Sepsis Early onset sepsis Late onset sepsis
23 weeks
24 weeks
25 weeks
26 weeks
27 weeks
28 weeks
Total
Significance adjusted OR (95% CI)
n = 23 (%)
n = 105 (%)
n = 125 (%)
n = 161 (%)
n = 194 (%)
n = 227 (%)
n = 835 (%)
P-value
1 (4) 13 (56)
9 (7) 52 (50)
7 (6) 47 (38)
3 (2) 33 (21)
7 (4) 26 (13.4)
8 (3) 26 (11.5)
35 (4) 197 (24)
1.04 (0.92–1.17) 0.70 (0.61–0.84)***
2 (9)
9 (8.6) 5 (4.8)
12 (9.6) 5 (4)
7 (4.4) 6 (3.8)
7 (3.6) 5 (2.6)
13 (5.8) 6 (2.6)
50 (6) 27 (3)
0.94 (0.74–1.20) 1.12 (0.82–1.53)
12 (52) 6 (26)
58 (55) 33 (31)
65 (52) 23 (18)
96 (60) 21 (13)
113 (58) 19 (10)
99 (44) 12 (5)
443 (53) 114 (14)
0.72 (0.61–0.85)*** 0.69 (0.58–0.83)***
97 (12)
0.74 (0.57–0.97)*
23 (3) 158/731 (22)
1.11 (0.77–1.61) 0.56 (0.47–0.67)***
439 (53)
0.65 (0.56–0.75)***
465/835 (56)
0.65 (0.56–0.75)***
GIT morbidity ⩾ NEC II A Focal intestinal perforation Patent ductus arterious Medical treatment only Need for surgical ligation
Cranial ultrasound abnormalities ⩾ Stage III/IV 6 (26) 25 (24) 17 (14) 21 (13.2) 17 (9) 11 (5) intraventricular hemorrhage Periventricular leukomalacia 1 (1) 6 (5) 3 (2) 8 (4) 5 (2) ⩾ Stage III retinopathy of 10/12 (83.4) 46/74 (62) 42/101 (42) 26/143 (18) 21/183 (11) 13/220 (6) prematurity ⩾ 1 Major neonatal 18/19 (95) 92/100 (92) 91 (73) 86 (53) 81 (42) 71 (31) morbidity Composite morbidity 22/23 (94.7) 97/105 (93) 100/125 (79) 90/161 (54) 82/194 (41) 74/227 (31)
Abbreviations: CI, confidence interval; GIT, gastrointestinal tract; NEC, necrotizing enterocolitis; NICU, neonatal intensive-care unit; OR, odds ratio. *P ⩽ 0.05, **P ⩽ 0.01, ***P ⩽ 0.001. P-values and OR (95% CI) were determined for differences according to gestational age, with adjustment for birth weight, year of birth. Comfort care cases were excluded from analysis.
Table 4.
Analysis of risk factors associated with composite morbidity in NICU admissions
Antenatal Antenatal steroids Cesarean section Multiple pregnancy Chorioamnionitis Perinatal 5 min Apgar ⩽ 5 Gestational age o26 weeks Mean birth weight Male Small for gestational age (SGA) Neonatal factors Hypotension needing inotropes in first 72 h Air leaks Patent ductus arterious needing treatment Respiratory distress syndrome CRIB II
No adverse outcome
Adverse outcome
n = 370 (%)
n = 465 (%)
280 (76) 208 (56) 90 (24) 163 (44)
333 216 104 223
(71) (46) (22) (48)
Crude OR (95% CI)
0.78 (0.57–1.07) 0.67 (0.51–0.89)** 0.89 (0.65–1.24) 1.18 (0.89–1.55)
Adjusted OR (95% CI)
0.64 (0.44–0.93)*
20 (5) 34 (9) 995 ± 192 183 (49) 33 (9)
75 (16) 219 (47) 798 ± 192 284 (61) 76 (16)
3.36 (2.04–5.62)*** 8.77 (5.91–13.15)*** 0.995 (0.994–0.996)*** 1.60 (1.22–2.11)*** 1.99 (1.29–3.08)**
1.83 (0.98–3.46) 2.80 (1.63–4.80)*** 0.997 (0.996–0.998)*** 1.68 (1.20–2.39)*** 1.34 (0.72–2.49)
124 (34) 10 (2.7) 206 (56) 139 (38) 9.7 ± 1.8
273 (60) 82 (18) 351 (80) 290 (64) 12.5 ± 2.7
2.90 (2.18–3.86)*** 7.85 (4.01–15.38)*** 3.18 (2.33–4.33)*** 2.93 (2.21–3.91)*** 1.66(1.53–1.81)***
1.13 (0.94–1.35) 4.32 (2.00–9.06)*** 1.81 (1.23–2.66)** 1.66 (1.16–2.39)**
Abbreviations: CI, confidence interval; CRIB II, clinical risk index for babies-revised version II; NICU, neonatal intensive-care unit; OR, odds ratio. *P ⩽ 0.05, **P ⩽ 0.01, ***P ⩽ 0.001. Cesarean section, 5 min Apgar scores, gestional age o26 weeks, mean birth weight, male gender, small for gestational age, hypotension needing inotropes, airleaks, patent ductus arteriosus and respiratory distress syndrome and year of birth were entered into the multivariate model for adjusted odd ratio. Comfort care cases were excluded from analysis.
data is in contrast to that presented by the Japanese wherein the mortality rate was significantly lower in 2005 compared with 2000 for all gestations below 28 weeks.34 Kyser35 also demonstrated increasing odds of survival of extremely preterm infants below 25 weeks during the decade from 2000 to 2009. Our low survival rates of 19% at 23 weeks were very similar to data presented by © 2014 Nature America, Inc.
national or regional data from Switzerland, UK, Canada, Australia and NICHD, USA.21,24,36,37 The survival rates between 24 and 28 weeks in our study mirrored the figures reported by the NICHD between 2003 and 2007 at each week of gestation and were also similar to the Swedish data between 24 and 26 weeks and the Swiss data between 2005 and 2009.21,23,38 Unlike the Epicure 2 Journal of Perinatology (2014), 1 – 7
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
6 study which showed improved survival at 24 and 25 weeks between 1997 and 2006 we did not find any significant improvement in survival at these gestations from 2000 to 2009.26 However, our overall survival of 57% and 80% at 24 and 25 weeks, respectively, was slightly higher than the reported 47 and 67% survival at these two gestations in 2006 in Epicure 2.26 Our figures at these two gestations were also similar to the 63 and 79% survival at 24 and 25 weeks reported by Kutz et al.39 from a single center study in Germany. One possible reason for the higher survival in our study, as well as by Kutz et al.39 was that geographic population studies report lower survival figures than hospital-based studies. Our center, however, is the regional referral perinatal center and accounts for more than two thirds of infants born at the thresholds of viability in Singapore (unpublished data). The incidence of respiratory distress syndrome and need for, mechanical ventilation, surfactant therapy and postnatal steroids decreased with increasing GA whereas the use of exclusive CPAP support only increased significantly from 26 weeks onwards. The surfactant use in our cohort seemed lower and corresponding CPAP use higher in our study than the average NICHD figures between 26 and 28 weeks.21 The incidence of major morbidities such as CLD at 36 weeks postmenstrual age in survivors, culture-proven sepsis, severe ROP and severe IVH decreased with increasing gestation, whereas NEC/focal intestinal perforation did not show a similar trend across gestations. The presence of one or more major neonatal morbidities decreased with increasing GA similar to data from Sweden.3–5,18,19,40 Similar to data from Israel,41 we also looked at presence of mortality and morbidity as a composite morbidty in the babies provided with NICU care and it was seen that this adverse outcome was significantly associated with male sex, decreasing birth weight, the presence of significant hyaline membrane disease, airleaks and PDA. Delivery by cesarean section seemed to be protective and this has been reported by others at extreme prematurity.22,27,29 The increased adverse outcome in boys seen in our study has been well documented historically and from recent data the exact pathogenesis remains unclear.42 Intrauterine growth restriction has been reported as an independent risk factor for an adverse outcome in a recent meta-analysis by Damodaram et al.43 and Tyson et al.44 and in our study it was also seen to be predictive on univariate analysis. However, this significance did not persist on multivariate analysis. This additional knowledge of gender and growth restriction if available antenatally may be useful in refining the antenatal counseling in extreme prematurity. CRIB II is a tool commonly used to assess neonatal illness severity in preterm VLBW infants and in predicting in hospital neonatal mortality, as well as neurodevelopmental impairment at 36 months.11–13 In our study we also found CRIB II to be a predictor of composite morbidity on univariate analysis. Although birth weight, gestation and sex are not modifiable factors attempts to improve the body temperature or base excess could potentially improve CRIB scores and serve as a quality improvement tool. Similarly quality improvement tools aimed at reduction of airleaks may also help in reducing composite morbidity in our high-risk preterm population and is being evaluated. One of the strengths of our study was the systematic comprehensive data which was collected prospectively and spanned a period of 10 years. Adjustments were also made for the time effect thus minimizing the bias introduced due to a long study period. There was no change in either mortality or major morbidity over the 10 years of the study. There was, however, a significant reduction in mortality and morbidity noted with each increasing week of gestation. The other strength of our data is the inclusion of the live births not resuscitated in the delivery room in the denominator for survival estimation. This enabled us to overcome to a certain extent the evidence of selection bias reported by Evans45 in preterm survival studies with over estimation of survival in studies Journal of Perinatology (2014), 1 – 7
which did not take into account live births not resuscitated. Despite being a single-center study, our survival data are consistent with the findings reported from geographically defined populations and probably reflect the centralization of tertiary perinatal care in Singapore.36 Even within geographic areas, centralization of perinatal care for extremely preterm deliveries to centers of excellence has been recommended in Sweden, Finland and USA.38,46,47 Thus, our study has the advantage of reflecting centralized perinatal care in Singapore as KKH is the regional referral unit with the highest volume of extreme preterm births. This may to a certain extent overcome the limitation that our report does not describe a geographically based sample. Neonatal mortality and morbidity data are invaluable tools in the decision-making process along with the data on neurodevelopmental outcome in childhood. Thus, all extremely high-risk preterm infants need to be followed up for long-term neurodevelopmental morbidities and outcome, which is currently being done in our cohort up to 8 years of age. In conclusion, increasing survival and decreasing morbidity were seen with each increasing week in gestation without any significant alteration in survival trends over the decade, probably reflecting the limits of technological advances and biologic plausibility. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS We thank Lim Wai Yee for her assistance with the manuscript.
REFERENCES 1 Fanaroff AA, Stoll BJ, Wright LL, Carlo WA, Ehrenkranz RA et al. Trends in neonatal morbidity and mortality for very low birthweight infants. Am J Obstet Gynecol 2007; 196(2): 147.e1–8. 2 Lundqvist P, Kallen K, Hallstorm I, Westas LH. Trends in outcomes for very preterm infants in the southern region of Sweden over a 10 year period. Acta Paediatr 2009; 98: 648–653. 3 Schmidt B, Asztalos EV, Roberts RS, Robertson CMT, Sauve RS et al. Impact of bronchopulmonary dysplasia, brain injury and severe retinopathy on the outcome of extremely low birth weight infants at 18 months. Results from the trial of indomethacin prophylaxis in preterm. JAMA 2003; 289: 1124–1129. 4 Bassler D, Stoll BJ, Schmidt B, Asztalos EV, Roberts RS et al. Using a count of neonatal morbidities to predict poor outcome in extremely low birth weight infants: Added role of neonatal infection. Pediatrics 2009; 123: 313–318. 5 Gargus RA, Vohr BR, Tyson JE, High P, Higgins RD, Wrage LA et al. Unimpaired outcomes for extremely low birth weight infants at 18 to 22 months. Pediatrics 2009; 124: 112–121. 6 Lubcheno LO, Hansman C, Bay E. Intrauterine growth as estimated from liveborn birthweight data at 24 to 42 weeks of gestation. Pediatrics 1963; 32: 793–800. 7 Fenton TR. A new growth chart for preterm babies: Babson and Benda’s chart updated with recent data and a new format. BMC Pediatr 2003; 3: 13. 8 Tita ATN, Andrews WW. Diagnosis and management of clinical chorioamnionitis. Clin Perinatol 2010; 37: 339–354. 9 Redline RW. Inflammatory response in acute chorioamnionitis. Semin Fetal Neonatal Med 2012; 17: 20–25. 10 Diagnosis and management of preeclampsia and eclampsia. ACOG Practice Bulletin No. 33. American College of Obstetricians and Gynecologists. Obstet Gynecol 2002; 99: 159–167. 11 Parry G, Tucker J, Tarnow-Mordi W. CRIB II: an update of the clinical risk index for babies score. UK Neonatal Staffing Study Collaborative Group. Lancet 2003; 361 (9371): 1789–1791. 12 Gagliardi L, Cavazza A, Brunelli A, Battaglioli M, Merazzi D et al. Assessing mortality risk in very low birthweight infants: a comparison of CRIB, CRIB-II, and SNAPPE-II. Arch Dis Child Fetal Neonatal Ed 2004; 89(5): F419–F422. 13 Lodha A, Sauve R, Chen S, Tang S, Christianson H. Clinical Risk Index for babies score for the prediction of neurodevelopmental outcomes at 3 years of age in infants of very low birthweight. Dev Med Child Neurol 2009; 51(11): 895–900. 14 Jobe AH, Bancalari E. Bronchopulmonary Dysplasia. Am J Respir Crit Care Med 2001; 163: 1723–1729.
© 2014 Nature America, Inc.
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
7 15 Papile LA, Munsick-Bruno G, S Chaefer A. Relationship of cerebral intraventricular haemorrhage in early childhood neurological handicaps. J Pediatr 1983; 103: 293. 16 An International Committee for the Classification of Retinopathy of Prematurity. The International Classification of Retinopathy of Prematurity Revisited. Arch Ophthalmol. 2005; 123(7): 991–999. 17 Pumberger W, Mayr M, Kohlhauser C, Weninger M. Spontaneous localized intestinal perforation in very low birth weight infants: a distinct clinical entity different from necrotizing enterocolitis. J Am Coll Surg 2002; 195(6): 796–803. 18 Husain N, Galou M, Ehrerikranz RA, Henson VC, Rowe JC. Pre-discharge outcome of 22-27 weeks gestational age infants born at tertiary care centres in connecticut. Implications for perinatal management. Conn Med 1998; 62(3): 131–137. 19 Hintz SR, Poole WK, Wright LL. Changes in mortality and morbidities among infants born less than 25 weeks during the post-surfactant era. Arch Dis Child Fetal Neonatal Ed 2005; 90: F128–F133. 20 Lemons JA, Bauer CR, Oh W, Korones SB, Papile LA, Stoll BJ et al. Very low birth weight outcomes of the National Institue of Child Health and Human Development Neonatal Research Network, January 1995 through December 1996. NICHD Neonatal Research Network. Pediatrics 2001; 107: e1. 21 Stoll BJ, Hansen NI, Bell EF, Shankaran S, Laptok AR et al. Neonatal outcomes of extremely preterm infants from NICHD Neonatal Research Network. Pediatrics 2010; 126: 443. 22 Vanhaesebrouck TP, Allegaert K, Bottu J, Debauche C, Devlieger H et al. EPIBEL Study: Outcomes to discharge from hospital for extremely preterm infants in belgium for the EPIBEL study group. Pediatrics 2004; 114(3): 663–675. 23 Berger TM, Steurer MA, Woerner A, Schiffer PM, Adams M. Trends and centre to centre variability in survival rates of very preterm infants ( o 32 weeks) over a 10 year period in Switzerland. Arch Dis Child Fetal Neonatal Ed 2012; F1–F6. 24 Shah PS, Sankaran K, Aziz K, Allen AC, Seshia M, Ohlsson A et al. Canadian Neonatal Network Outcomes of preterm infants o 29 weeks gestation over 10-year period in Canada: a cause for concern? J Perinatol 32: 132–138. 25 Draper ES, Zeitlin J, Fenton AC. Investigating the variations in survival rates for very preterm infants in 10 European regions: the MOSAIC birth cohort. Arch Dis Child Fetal Neonatal Ed 2009; 94: F158–F163. 26 Costeloe K, Hennessy E, Haider S, Stacey F, Marlow N, Draper ES. Short term outcomes after extreme preterm birth in England: Comparison of two birth cohorts in 1995 and 2006 (the EPICURE studies). BMJ 2012; 345: e7976. 27 Hussain N, Rosenkrantz TS. Ethical considerations in the management of infants born at extremely low gestational age. Semin Perinatol 2003; 27(6): 458–470. 28 Tommiska V, Heinonen K, Lehtonen L, Renlund M, Saarela T, Tammela O et al. No improvement in outcome of nationwide extremely low birth weight infant populations between 1996-1997 and 1990-2000. Pediatrics 2007; 119: 29–36. 29 Markestad T, Kaaresen PI, Ronnestad A, Reigstad H, Lossius K, Medbo S et al. Norwegian extreme prematurity study group Early death, morbidity and need of treatment among extremely premature infants. Pediatrics 2005; 115: 1289–1298. 30 Morgan MA, Goldenberg RL, Schulkin J. Obstetrician-gynaecologist practices regarding preterm birth at the limit of viability. J Matern Fetal Neonatal Med 2008; 21(2): 115–121.
© 2014 Nature America, Inc.
31 Perinatal Care at the Threshold of Viability. ACOG Practice Bulletin. Obstet Gynecol 2002; 100: 617–624. 32 Nuffield Council on Bioethics Dilemmas in current practice: babies born at the borderline of viability. Critical Care Decisions in Fetal and Neonatal Medicine: Ethical Issues. Nuffield Council on Bioethics: London, UK, 2006, pp 67–87. 33 Pignotti MS, Donzelli G. Perinatal care at the threshold of viability: An international comparison of practical guidelines for the treatment of extremely preterm births. Pediatrics 2008; 121: e193. 34 Itabashi K, Horiuchi T, Kusuda S, Kabe K. Mortality rates for extremely low birthweight infants born in Japan in 2005. Pediatrics 2009; 123: 445–450. 35 Kyser KL, Morriss FH Jr, Bell EF, Klein JM, Dagle JM. Improving survival of extremely preterm infants born between 22 and 25 weeks of gestation. Obstet Gynecol 2012; 119(4): 795–800. 36 Field DJ, Dorling JS, Manktelow BN, Draper ES. Survival of extremely premature babies in a geographically defined population: prospective cohort study of 1994–1999 compared with 2000–2005. BMJ 2008; 336: 1221. 37 Fischer N, Steurer MA, Adams M, Berger TM. Survival rates of extremely preterm infants (gestational age o 26 weeks) in Switzerland: impact of the Swiss guidelines for the care of infants born at the limit of viability. Arch Dis Child Fetal Neonatal Ed 2009; 94(6): F407–F413. 38 The Express Team. One-year survival of extremely preterm infants after active perinatal care in Sweden. JAMA 2009; 301(21): 2225–2233. 39 Kutz P, Horsch S, Kuhn L, Roll C. Single centre vs population based outcome data of extremely preterm infants at the limits of viability. ACTA Paediatr 2009; 98: 1451–1455. 40 Stoll BJ, Hansen NI, Adams-Chapman I, Fanaroff AA, Hintz SR, Vohr B et al. Neurodevelopmental and growth impairment among extremely low birth weight infants with neonatal infection. JAMA 2004; 292: 2357–2365. 41 Kugelman A, Bader D, Lerner-Geva L, Boyko V, Levitzki O, Riskin A et al. Poor outcomes at discharge among extremely premature infants. Arch Pediatr Adolesc Med 2012; 166(6): 543–550. 42 Kent AL, Wright IM, Abdel-Latif ME. New South Wales and Australian Capital Territory Neonatal Intensive Care Units Audit Group Mortality and adverse neulogic outcomes are greater in preterm male infants. Pediatrics 2012; 129: 124–131. 43 Damodaram M, Story L, Kulinskaya E, Rutherford M, Kumar S. Early adverse perinatal complications in preterm growth-restricted fetuses. Aust N Z J Obstet Gynaecol 2011; 51: 204–209. 44 Tyson JE, Parikh NA, Langer J, Green C, Higgins RD. Intensive care for extreme prematurity—moving beyond gestational age. N Engl J Med 2008; 358: 1672–1681. 45 Evans DJ, Levene MI. Evidence of selection bias in preterm survival studies: a systematic review. Arch Dis Child Fetal Neonatal Ed 2001; 84: F79–F84. 46 Rautava L, Lehtonen L, Peltola M, Korvenranta E, Korvenranta H, Linna M et al. The effect of birth in secondary or tertiary level hospitals in Finland on mortality in very preterm infants: a birth-register study. Pediatrics 2007; 119: e257–e263. 47 Lorch SA, Baiocchi M, Ahlberg CE, Small DS. The differential impact of delivery hospital on the outcomes of premature infants. Pediatrics 2012; 130: 270–279.
Journal of Perinatology (2014), 1 – 7
ORIGINAL ARTICLE
Neonatal outcome of extremely preterm Asian infants ⩽ 28 weeks over a decade in the new millennium P Agarwal, B Sriram and VS Rajadurai OBJECTIVE: To evaluate predischarge neonatal mortality and morbidity and associated risk factors in extremely preterm Asian infants ⩽ 28 weeks, over a decade, so as to facilitate formulation of perinatal guidelines and counseling. STUDY DESIGN: Cohort study of 887 liveborn extremely preterm neonates between 2000 and 2009 at KKH, the centralized perinatal center in Singapore. Outcome measures were predischarge mortality, presence of one or more major neonatal morbidities and the composite outcome of mortality or neonatal morbidity. RESULT: Overall survival to discharge was 709/887 (80%) and was significantly higher with increasing gestational age (GA) (19% at 23 weeks to 93% at 28 weeks, P o 0.001). Survival remained unchanged between 78 and 86% during the decade with no significant secular trend. Overall incidence of major morbidities were bronchopulmonary dysplasia (29%), late onset sepsis (23%), severe retinopathy of prematurity (21%), Grade 3 to 4 intraventricular hemorrhage (12%) and necrotizing enterocolitis ⩾ Bells’ stage II/focal intestinal perforation (9%). Composite morbidity was seen in 465/835 (56%) neonatal intensive-care unit admissions, decreased with increasing GA (P o 0.001; odds ratio 0.65 (95% confidence interval 0.56 to 0.75) and was independently predicted by birth weight, Clinical Risk Index for Babies–revised version II score, male gender, presence of patent ductus arteriosus and airleaks. CONCLUSION: Although there was no significant trend in neonatal survival or composite morbidity over the decade, improved survival and morbidity were seen with increasing GA. Journal of Perinatology advance online publication, 5 February 2015; doi:10.1038/jp.2014.205
INTRODUCTION There has been a significant improvement in the survival of extremely preterm and extremely low birth weight infants in the decades of the eighties and nineties. This improvement is a reflection of the advances in antenatal, perinatal and neonatal care including increasing use of antenatal steroids, surfactant therapy, and improved respiratory and nutritional management. The neonatal research network from the National Institute of Child Health and Human Development (NICHD), USA has studied the trends in survival and suggested that a plateau in the neonatal survival of very low birth weight (VLBW) infants has been reached in the modern era.1 Concerns also remain that high morbidity may replace the higher mortality of earlier eras in extremely low gestational age (ELGA) newborns.2 It has been shown that major neonatal morbidities are associated with higher incidence of adverse neurodevelopmental outcome.3–5 Traditionally the neonatal outcomes of preterm infants have been reported according to birth weight among VLBW infants. However, there may be some inaccuracy in this because of inclusion of more mature intrauterine growth-restricted infants. Second, birth weight is available only after delivery and thus not useful for antenatal counseling and prognostication. There is a paucity of Asian data reporting neonatal outcomes in the extremely preterm infants. This data, regarding risk of pre discharge mortality and major morbidity, is commonly needed by neonatologists, obstetricians and parents for appropriate decision making.3–5 The primary aim of this study was to evaluate GA-specific neonatal mortality and morbidity in preterm VLBW infants born between 23 and 28 weeks GA in the single largest
tertiary perinatal center in Singapore over a decade in the new millennium. We also studied the trends in mortality over the decade and analyzed the predictive risk factors associated with mortality and morbidity in a large cohort of extremely preterm Asian infants. METHODS Study population The study was conducted at KK Women’s and Children’s Hospital (KKH), Singapore which is the largest tertiary referral perinatal center in Singapore with an annual delivery rate of approximately 12 000 per year. KKH provides care for more than two thirds of infants born at the threshold of viability (GA o 25 weeks) in Singapore who are predominantly of Chinese, Malay or Indian ethnicities (unpublished national network data). The study period was from January 2000 to December 2009 and the infants were part of the high-risk VLBW database. Perinatal/ neonatal data of all liveborn infants born o 29 weeks gestation and/or admitted to the KKH neonatal intensive-care unit (NICU) were collected. Data on stillbirth and miscarriage were not collected. The hospital follows a structured resuscitation policy wherein the limit of viability is set at 23 weeks gestation and no resuscitation is instituted except at specific parental request and after appropriate counseling. At 24 weeks gestation, resuscitation is commenced unless parents indicate otherwise. From 25 weeks onwards, all babies are resuscitated and intensive care offered. Babies who were born o25 weeks gestation and provided with comfort care were excluded from subsequent analysis of neonatal outcomes and morbidities, and analysis was restricted to babies provided with NICU admission and care. Comfort care was provided to babies born alive at threshold of viability and where after neonatal counseling families had chosen not to provide intensive care. Comfort care included palliative care and keeping the baby warm till demise.
Department of Neonatology, KK Women’s and Children’s Hospital, Singapore. Correspondence: Dr P Agarwal, Department of Neonatology, KK Women’s and Children’s Hospital, 100 Bukit Timah Road, Singapore 229899, Singapore. E-mail: [email protected] Received 14 May 2014; revised 5 September 2014; accepted 9 October 2014
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
2 Table 1.
Perinatal/neonatal demographics and delivery room information according to gestational age
Maternal/perinatal Race Chinese Malay Indian Others PIH PROM 424 h Chorioamnionitis Antenatal steroidsa Cesarean section delivery
23 weeks
24 weeks
25 weeks
26 weeks
27 weeks
28 weeks
Total
Significance OR (95% CI)/difference in mean
n = 23 (%)
n = 105 (%)
n = 125 (%)
n = 161 (%)
n = 194 (%)
n = 227 (%)
n = 835 (%)
P-value
8 (35) 14 (60) 1 (4) — — 6 (26) 17 (74) 18/47 (38) 1 (4)
57 (54) 73 (78) 100 (62) 122 (63) 135 (60) 495 (59) 36 (34) 26 (21) 33 (21) 35 (18) 35 (25) 200 (24) 17 (7) 15 (12) 19 (12) 21 (11) 21 (9) 84 (10) 3 (5) 11 (9) 9 (6) 16 (8) 16 (8) 54 (7) 8 (8) 11 (9) 24 (15) 30 (16) 41 (18) 114 (14) 38 (36) 50 (40) 60 (37) 79 (40) 61 (27) 292 (35) 66 (63) 73 (59) 67 42) 90 (47) 73 (33) 386 (47) 78/117 (67) 94/127 (74) 117/164 (71) 156/194 (80) 164/227 (73) 627/876 (72) 34 (32) 43 (34) 96 (60) 108 (56) 142 (63) 424 (51)
*
Neonatal Birth weight (mean ± 2s.d.)g 569 ± 64 Male 16 (70) SGA 1 (4) Delivery room resuscitation 23/47 (49) with intubationa 5 min Apgar score ⩽ 5 10 (43) CRIB II 17 ± 1.5
2.41 (1.96–2.94)*** 0.84 (0.74–0.96)* 0.59 (0.51–0.68)*** 1.54 (1.34–1.77)*** 2.38 (1.95–2.67)***
672 ± 86 769 ± 122 834 ± 163 955 ± 172 1057 ± 202 885 ± 216 96.7 (89.06–104) 62 (59) 76 (61) 82 (51) 108 (56) 123 (54) 467 (56) 0.82 (0.72–0.93 )** 5 (5) 10 (8) 28 (17) 27 (14) 38 (17) 109 (13) 7.66 (5.13–11.4)*** 99/117 (84) 104/127 (82) 130/164 (79) 121/194 (62) 106/227 (47) 583/876 (67) 0.91 (0.79–1.04) 16 (15) 15 ± 1.4
26 (21) 13 ± 1.4
13 (8) 11 ± 1.3
14 (7) 10 ± 1.3
16 (7) 9 ± 1.5
95 (12) 11.3 ± 2.8
0.71 (0.58–0.86)*** (−1.29 to − 1.14)***
Abbreviations: CI, confidence interval; CRIB II, Clinical Risk Index for Babies–revised version II; OR, odds ratio; PROM, premature rupture of membranes; SGA, small for gestational age. *P ⩽ 0.05, **P ⩽ 0.01, ***P ⩽ 0.001. P-values and OR (95% CI) were determined for differences according to gestational age, with adjustment for birth weight and year of birth. aComfort care cases included in the denominator for antenatal steroids and delivery room resuscitation. For all other variables comfort care cases were excluded from analysis.
This study was approved by the institutional review board of the hospital. Assessment of GA was done by the neonatologist based on the best obstetric estimate, date of last menstrual period and early dating scan. If no antenatal data were available, GA estimation was done by postnatal– neonatal assessment. Lubchenko’s growth charts were used to classify GA status through 2004 and Fenton’s growth charts were used beginning in 2005.6,7 Small for GA (SGA) was defined as birth weight less than the tenth centile for GA. Maternal obstetric data collected included data on antenatal steroid use, premature rupture of membranes 424 h, chorioamnionitis, pre-eclampsia and mode of delivery. Antenatal steroid therapy was considered to be complete if two doses of steroids were given to the mother at least 24 h prior to delivery and as no treatment if no steroids or partial course of steroids was given. Chorioamnionitis was diagnosed on either presence of clinical findings such as maternal fever, maternal/fetal tachycardia, uterine tenderness or foul smelling discharge or histopathological findings.8,9 On the basis of the American College of Obstetricians and Gynecologists guidelines, pre-eclampsia was defined as systolic blood pressure of ⩾ 140 mm Hg and diastolic blood pressure of ⩾ 90 mm Hg after 20 weeks of gestation in the presence of proteinuria (⩾1+ or ⩾ 0.3 g per day).10 Neonatal data on delivery room resuscitation, treatment modalities and complications were collected prospectively. Intensive delivery room resuscitation included endotracheal intubation, cardiac massage and use of adrenaline and excluded use of mask ventilation or free-flow oxygen.
Neonatal outcomes The Clinical Risk Index for Babies–revised version II (CRIB II) score is a riskadjusted instrument which includes birth weight, GA, sex, admission temperature and base deficit and is based on data available soon after birth. This tool is commonly used to assess neonatal illness severity in preterm VLBW infants and in predicting in hospital neonatal mortality, as well as neurodevelopmental impairment at 36 months.11–13 Surfactant was used as early rescue therapy for infants with respiratory distress syndrome which was defined by the presence of clinical and radiological features, along with the need for respiratory support in the first 24 h of life. A hemodynamically significant patent ductus arteriosus (PDA) on twodimentional echocardiography was treated with intravenous indomethacin or surgical ligation if medical treatment was unsuccessful or contraJournal of Perinatology (2014), 1 – 7
indicated. Chronic lung disease was defined as oxygen dependency at 36 weeks postmenstrual age. 14 Rescue steroids were used based on standard departmental protocol beyond 3 weeks of age to facilitate extubation in intubated ventilatordependent neonates with oxygen requirement 440% and significant radiological features of persistent lung disease. Serial cranial ultrasounds were done according to unit protocol at days 1, 3, 7, 14, 30, 60 and at term. The Papile grading system was used to grade intraventricular hemorrhage (IVH) while a standard definition of cystic periventricular leukomalacia (PVL) was used.15 Presence of grade III/IV IVH or PVL was considered to be a major ultrasonographic abnormality. A pediatric ophthalmologist examined surviving infants at 4 to 6 weeks postnatal age and established presence and staging of retinopathy of prematurity (ROP) according to the international classification.16 Stage III ROP or presence of plus disease was considered to be severe and treated with cryotherapy/laser photocoagulation when indicated. Necrotizing enterocolitis (NEC) was defined according to modified Bell’s criteria and ⩾ stage II was considered to be significant. Presence of focal intestinal perforation was also grouped together with NEC ⩾ stage II as a significant gastrointestinal morbidity.17 Nosocomial sepsis was defined as blood culture-positive sepsis occurring beyond 72 h of life whereas perinatally acquired early onset sepsis was defined as blood culture being positive o 72 h of life. Among infants admitted to the NICU, the major morbidities with potential long-term sequelae were cystic PVL, IVH III/IV chronic lung disease, severe ROP and NEC ⩾ 2 A/focal intestinal perforation and blood culture-positive sepsis. The presence of any one of the above morbidities was considered as a major morbidity.3–5,18,19 As mortality/neonatal morbidity may be competing outcomes in these highly immature infants, the presence of either mortality or morbidity as defined above was considered to be a composite morbidity. Statistics. Data were analyzed using SPSS for Windows, Version 16.0 (SPSS, Chicago, IL, USA). Categorical variables were compared using χ2-test or Fisher’s exact test where appropriate, whereas continuous variables were compared using Student’s t-test. Birth weight and GA are both strong and important predictors of neonatal outcome, but GA was moderately correlated with birth weight (Pearson r = 0.65). Thus we transformed GA variable into a categorical variable with GA ⩽ 25 weeks and GA 425 weeks as GA was our exposure © 2014 Nature America, Inc.
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
3 of interest in this study, and we wanted to examine the risks associated with GA in our cohort of VLBWs. CRIB II score was not included in the multivariate analysis as it was highly correlated with birth weight (Pearson r = 0.75) and GA (Pearson r = 0.86). Survival trends were analyzed using Cox regression models. The incidence of composite morbidity was evaluated for each GA. Crude and adjusted odds ratios would be reported for the identified predictive factors using logistic regression models. Year of birth was included in the multivariate model to look for a significant trend over the 10-year study period. Results are presented as odds ratio and 95% confidence intervals (95%). For the multivariate analysis, only the significant factors are presented in the table. All analysis performed were two tailed and the statistical significance was set at Po0.05.
RESULTS Between January 2000 and December 2009, 887 live born infants o29 weeks and o 1500 g were admitted to the KKH NICU or born alive. Of these 887 neonates, 11 (1.2%) were born at 22 weeks gestation and they were excluded from the data analysis. The mean birth weight of the 876 study neonates was 873 ± 220 g with a mean GA of 26.1 ± 1.5 weeks. Antenatal, maternal and neonatal demographics Perinatal and neonatal demographic data and neonatal outcomes are presented in Table 1 to show differences in maternal complications and neonatal demographic data by GA at birth having adjusted for birth weight and year of birth. Complete course of antenatal steroids was given in 72% of the mothers with 431 (49%) of the neonates being delivered by cesarean section and 583 (67%) being intubated for delivery room resuscitation. The rate of antenatal steroid use increased from 38% at 23 weeks, to 67% at 24 weeks and 72 to 80% between 25 and 28 weeks. There was also a sharp rise in the rates of cesarean section deliveries between 23 and 24 weeks (2% versus 30%) and between 25 and 26 weeks (35% versus 60%). Premature delivery due to maternal PIH and the proportion of small for GA neonates were significantly lower below 26 weeks. Overall 67% of the infants needed or were provided with intubation in the delivery suite. Half of the infants delivered at 28 weeks and a third at 27 weeks did not need intubation in the delivery room, whereas 79 to 85% of infants between 24 and 26 weeks needed intubation at birth. Apgar scores at 5 min were recorded only in the neonates provided with active resuscitation and thus the 41 neonates provided with comfort care were excluded. Incidence of Apgar score o5 at 5 min was significantly higher below 26 weeks. CRIB II scores were available for 723 of the neonates and scores decreased with increasing GA, but there was no significant change in the mean CRIB scores between 2000 and 2009 and ranged from 10.7 ± 2.7 to 11.9 ± 3.02 over the study period. Neonatal outcomes Trends in survival. Overall, 709 (81%) of the 876 infants survived to discharge in our entire cohort with a linear relation between survival and increasing GA. After adjusting for birth weight and year of birth, survival improved significantly with increasing GA at odds ratio 1.50 (95% confidence interval 1.28 to 1.74; P o0.001). Proportion of VLBW survivors increased from 9/47 (19%) at 23 weeks to 93% at 28 weeks (Figure 1). There was no increase in survival over the 10 years, either for the total cohort or for any GA group, and survival remained unchanged at 81.6% with interquartile range between 77 and 87% during the decade (Data not shown). Forty-one (5.5%) infants were born alive, but provided comfort care. The mean GA and birth weight of this group was significantly lower than the 835 babies who received NICU care, 23.6 ± 0.89 versus 26.2 ± 1.46 weeks and 608 ± 123 versus 885 ± 215 g, respectively (P o 0.001). Seventy-nine (9%) babies © 2014 Nature America, Inc.
Figure 1. Neonatal survival, mortality rates and comfort care according to gestational age.
died in the neonatal period and there were 47 (6%) late deaths in the post-neonatal period before hospital discharge. The respiratory morbidities and need for ventilator support in NICU admission are outlined in Table 2. Of the 835 babies who were provided NICU care, 591 (71%) needed surfactant therapy with significantly lower rates with each increasing week of gestation (Table 3). Only 102/835 (12%) of the cohort could be managed with nasal continuous positive airway pressure (CPAP) support alone with significantly increasing proportions as GA increased (Po 0.001). No baby at 23 or 24 weeks could be managed with CPAP support only. Of the 735 neonates who survived till 36 weeks postmenstrual age, 213(29%) had chronic lung disease and its incidence along with the use of postnatal steroids and home oxygen therapy were inversely proportional to GA at birth. Rates of other major neonatal morbidities were inversely proportional to GAs and are shown in Table 3. Of the 835 extremely preterm neonates provided with NICU care, six (0.7%) died within first 6 h before cranial ultrasounds could be performed. All the remaining neonates had a cranial ultrasound done with 451 (54%) being normal, 196 (24%) Grade I IVH, 84 (10%) Grade II, 53 (6%) Grade III IVH and 44 (5%) Grade IV IVH. PVL was seen in about 3% of the study cohort. Rates of severe IVH were greatest o25 weeks GA and then decreased with increasing GA. Rates of early and nosocomial late onset sepsis were 3% and 24%, respectively, for the whole cohort and the highest rates of late onset sepsis were seen below 25 weeks. Four hundred and forty-three (53%) of the study cohort needed medical treatment with indomethacin for a significant PDA whereas 114 (14%) additionally needed surgical ligation of the duct. The need for surgical ligation was highest at 23 and 24 weeks of GA. Twenty babies died prior to evaluating the duct. There were no significant differences in the incidence of NEC/focal intestinal perforation according to GA distribution. Of the 756 neonates who survived beyond 28 days, 19 died before the day of the first scheduled eye review and ophthalmological data were available for 731/737 (99%) infants. Three hundred and fifty-four (48%) had no ROP, 219 (30%) stage I and II ROP and 158 (21%) had severe ROP. Eightytwo (11%) of the babies needed laser treatment for the ROP. Neonatal morbidities were thus seen in 20 to 30% of the cohort as seen in Tables 3 and 4 and major morbidity comprising any of the previously defined morbidities was present in 439 (52%) of the patients. Composite morbidity comprising of either a death/major morbidity was seen in 465/835 (56%) neonates provided with NICU care and was inversely proportional to GA (odds ratio 0.65 (95% confidence interval 0.56 to 0.75; P o0.001; Table 3). Journal of Perinatology (2014), 1 – 7
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
4 Table 2.
Pulmonary morbidities according to gestational age in NICU admissions 23 weeks
24 weeks
25 weeks
26 weeks
27 weeks
28 weeks
Total
Significance adjusted OR (95% CI)
n = 23 (%)
n = 105 (%)
n = 125 (%)
n = 161 (%)
n = 194 (%)
n = 227 (%)
n = 835
P-value
Respiratory distress syndrome Surfactant therapy Air leaks CPAP support only Postnatal steroids
20 (87) 20 (87) 6 (26) 0 8/13 (62)
97 (94) 95 (91) 19 (19) 0 41/82 (50)
107 (86) 129 (81) 132 (68) 108 (86) 115 (71) 128 (66) 20 (16) 17 (11) 15 (8) 3 (2.4) 10 (6) 27 (14) 39/106 (37) 29/148 (20) 18/184 (10)
138 (61) 125 (55) 15 (7) 62 (27) 12/220 (6)
623 (75) 0.73 (0.62–0.86)*** 591 (71) 0.74 (0.64–0.86)** 92 (11) 0.86 (0.71–1.04) 102 (12) 1.72 (1.28–2.27)*** 147/753 (19.5) 0.68 (0.57–0.80)***
Survivors at PMA 36 weeks CLD Home O2
n = 12 10 (83) 4/10 (40)
n = 72 49 (68) 8/69 12)
n = 103 52 (50.5) 10/99 (10)
n = 219 25 (11) 2/212 (1)
n = 735 213/735 (29) 75/735 (10.2)
n = 145 46 (32) 8/139 (6)
n = 184 31 (17) 3/176 (1.7)
0.69 (0.59–0.81)*** 0.69 (0.52–0.94)**
Abbreviations: CI, confidence interval; CLD, chronic lung disease; CPAP, continuous positive airway pressure; NICU, neonatal intensive-care unit; OR, odds ratio; PMA, postmenstrual age. *P ⩽ 0.05, **P ⩽ 0.01, ***P ⩽ 0.001. P-values and OR (95% CI) were determined for differences according to gestational age, with adjustment for birth weight, year of birth. Comfort care cases were excluded from analysis.
Risk factors for composite morbidity (mortality/major morbidity) An analysis of risk factors for composite morbidity using multiple logistic regression showed GA o 26 weeks, male gender, presence of airleaks, significant PDA and respiratory distress syndrome were associated with higher odds for an adverse outcome, while increasing birth weight and cesarean delivery was associated with lower odds to an adverse outcome 5 min Apgar score ⩽ 5, hypotension in the first 72 h needing inotropes, small for GA were not independently predictive of composite morbidity though they were significant in univariate analysis (Table 4). The CRIB II score was not included in the multivariate analysis in view of the high multicollinearity with GA and birth weight with a Pearson’s of 0.75 for birth weight and 0.86 for GA. Year of birth was not significantly associated with an adverse outcome nor was there any protective effect of antenatal steroids. DISCUSSION In our Asian cohort of VLBWs, we found associations between higher survival and increasing GA. Lower birth weight, higher CRIB-II, male gender, presence of airleaks, significant PDA and respiratory distress syndrome were associated with higher odds to infant mortality or morbidity. However, cesarean delivery was protective against neonatal mortality and morbidity. Both birth weight and GA are considered to be predictors of outcome and traditionally the National Research Network of the NICHD has published outcomes of preterm babies based on birth weight.1,20 However, the inclusion of severely growth-restricted preterm infants may affect the reliability of the data. In addition, the birth weight is only known after birth and may thus be of limited use in perinatal counseling and parental decision making prior to delivery. GA based outcomes overcome these limitations and are therefore becoming increasingly the standard of reporting for outcomes of extremely preterm newborns.21–25 It is also important to evaluate trends in morbidity at extremely low gestation age to assess whether decreasing mortality is being replaced by increasing morbidity. This has been done in recent publications from EPICURE 2 and other international data.21,26–28 Antenatal, perinatal and delivery room data The rate of antenatal steroid use in our cohort was similar to the rates reported in the NICHD by Stoll et al. 21 whereas it was lower than the 85 to 94% usage reported from Norway29 A doubling of the rate of steroid use between 23 and 25 weeks from 38 to 74% was seen in our study though this was still lower than the 53% antenatal steroid use at 23 weeks reported by the NICHD.21 This Journal of Perinatology (2014), 1 – 7
lower use of antenatal steroids at 23 weeks was a strong reflection of our institutional policy where steroids were only recommended at 23+5 weeks of gestation and beyond, in keeping with the hospital guidelines of resuscitation and intensive care only at 24 weeks.The higher use of antenatal steroids and higher cesarean section rates beyond 24 weeks probably reflects the willingness of the obstetric and neonatal teams to provide active perinatal care only after this gestation. Our rates of cesarean section were very similar to those presented in the national data from Norway between 23 and 27 weeks with no baby at 23 weeks being delivered by cesarean section.29 In a survey of 1193 members of the American College of Obstetricians and Gynecologist regarding practices at the limits of viability it was seen that only 10% of respondents would perform a cesarean section for fetal distress at 23 weeks and 58% considered 24 weeks to be the earliest age of viability.30 Preterm delivery secondary to maternal pregnancy induced hypertension increased with increasing GA and probably reflected the perinatologist’s attempt at optimizing balance between maternal interests and fetal maturity in deciding timing of delivery. Rates of delivery room intubation were significantly lower at the extremes of gestation in our cohort. The lower rates of delivery room intubation below 24 weeks reflected the resuscitation policy for thresholds of viability in our department and were similar to the NICHD data.21 Conversely the lower rates of delivery room intubation at 27 and 28 weeks in our babies probably reflected their better physiologic condition at delivery and the widespread acceptance of early CPAP use in the delivery room in our center. Survival Overall survival in our extremely preterm cohort was at 80% with a significant reduction in mortality and improving survival noted with each increasing week of gestation from 19% at 23 weeks to 93% at 28 weeks. There was, however, no progressive reduction in mortality for the whole cohort, as well as at each individual gestational week over the 10 years of the study period. Similar findings with no improvement over time have also been reported by the neonatal research network in the NICHD between 2003 and 2007.21 The authors of the Epicure 2 study also did not reveal a significant improvement in survival at 22 to 23 weeks between 1997 and 2006 and this possibly reflects the biologically plausible limit to which the boundaries can be pushed by technological and medical advance.24–26 It could also be a reflection of the restrictive resuscitation policy at 23 weeks which is a common practice internationally in units in North America, UK and Europe.31–33 This © 2014 Nature America, Inc.
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
5 Table 3.
Other neonatal morbidities according to gestational age in NICU admissions
Sepsis Early onset sepsis Late onset sepsis
23 weeks
24 weeks
25 weeks
26 weeks
27 weeks
28 weeks
Total
Significance adjusted OR (95% CI)
n = 23 (%)
n = 105 (%)
n = 125 (%)
n = 161 (%)
n = 194 (%)
n = 227 (%)
n = 835 (%)
P-value
1 (4) 13 (56)
9 (7) 52 (50)
7 (6) 47 (38)
3 (2) 33 (21)
7 (4) 26 (13.4)
8 (3) 26 (11.5)
35 (4) 197 (24)
1.04 (0.92–1.17) 0.70 (0.61–0.84)***
2 (9)
9 (8.6) 5 (4.8)
12 (9.6) 5 (4)
7 (4.4) 6 (3.8)
7 (3.6) 5 (2.6)
13 (5.8) 6 (2.6)
50 (6) 27 (3)
0.94 (0.74–1.20) 1.12 (0.82–1.53)
12 (52) 6 (26)
58 (55) 33 (31)
65 (52) 23 (18)
96 (60) 21 (13)
113 (58) 19 (10)
99 (44) 12 (5)
443 (53) 114 (14)
0.72 (0.61–0.85)*** 0.69 (0.58–0.83)***
97 (12)
0.74 (0.57–0.97)*
23 (3) 158/731 (22)
1.11 (0.77–1.61) 0.56 (0.47–0.67)***
439 (53)
0.65 (0.56–0.75)***
465/835 (56)
0.65 (0.56–0.75)***
GIT morbidity ⩾ NEC II A Focal intestinal perforation Patent ductus arterious Medical treatment only Need for surgical ligation
Cranial ultrasound abnormalities ⩾ Stage III/IV 6 (26) 25 (24) 17 (14) 21 (13.2) 17 (9) 11 (5) intraventricular hemorrhage Periventricular leukomalacia 1 (1) 6 (5) 3 (2) 8 (4) 5 (2) ⩾ Stage III retinopathy of 10/12 (83.4) 46/74 (62) 42/101 (42) 26/143 (18) 21/183 (11) 13/220 (6) prematurity ⩾ 1 Major neonatal 18/19 (95) 92/100 (92) 91 (73) 86 (53) 81 (42) 71 (31) morbidity Composite morbidity 22/23 (94.7) 97/105 (93) 100/125 (79) 90/161 (54) 82/194 (41) 74/227 (31)
Abbreviations: CI, confidence interval; GIT, gastrointestinal tract; NEC, necrotizing enterocolitis; NICU, neonatal intensive-care unit; OR, odds ratio. *P ⩽ 0.05, **P ⩽ 0.01, ***P ⩽ 0.001. P-values and OR (95% CI) were determined for differences according to gestational age, with adjustment for birth weight, year of birth. Comfort care cases were excluded from analysis.
Table 4.
Analysis of risk factors associated with composite morbidity in NICU admissions
Antenatal Antenatal steroids Cesarean section Multiple pregnancy Chorioamnionitis Perinatal 5 min Apgar ⩽ 5 Gestational age o26 weeks Mean birth weight Male Small for gestational age (SGA) Neonatal factors Hypotension needing inotropes in first 72 h Air leaks Patent ductus arterious needing treatment Respiratory distress syndrome CRIB II
No adverse outcome
Adverse outcome
n = 370 (%)
n = 465 (%)
280 (76) 208 (56) 90 (24) 163 (44)
333 216 104 223
(71) (46) (22) (48)
Crude OR (95% CI)
0.78 (0.57–1.07) 0.67 (0.51–0.89)** 0.89 (0.65–1.24) 1.18 (0.89–1.55)
Adjusted OR (95% CI)
0.64 (0.44–0.93)*
20 (5) 34 (9) 995 ± 192 183 (49) 33 (9)
75 (16) 219 (47) 798 ± 192 284 (61) 76 (16)
3.36 (2.04–5.62)*** 8.77 (5.91–13.15)*** 0.995 (0.994–0.996)*** 1.60 (1.22–2.11)*** 1.99 (1.29–3.08)**
1.83 (0.98–3.46) 2.80 (1.63–4.80)*** 0.997 (0.996–0.998)*** 1.68 (1.20–2.39)*** 1.34 (0.72–2.49)
124 (34) 10 (2.7) 206 (56) 139 (38) 9.7 ± 1.8
273 (60) 82 (18) 351 (80) 290 (64) 12.5 ± 2.7
2.90 (2.18–3.86)*** 7.85 (4.01–15.38)*** 3.18 (2.33–4.33)*** 2.93 (2.21–3.91)*** 1.66(1.53–1.81)***
1.13 (0.94–1.35) 4.32 (2.00–9.06)*** 1.81 (1.23–2.66)** 1.66 (1.16–2.39)**
Abbreviations: CI, confidence interval; CRIB II, clinical risk index for babies-revised version II; NICU, neonatal intensive-care unit; OR, odds ratio. *P ⩽ 0.05, **P ⩽ 0.01, ***P ⩽ 0.001. Cesarean section, 5 min Apgar scores, gestional age o26 weeks, mean birth weight, male gender, small for gestational age, hypotension needing inotropes, airleaks, patent ductus arteriosus and respiratory distress syndrome and year of birth were entered into the multivariate model for adjusted odd ratio. Comfort care cases were excluded from analysis.
data is in contrast to that presented by the Japanese wherein the mortality rate was significantly lower in 2005 compared with 2000 for all gestations below 28 weeks.34 Kyser35 also demonstrated increasing odds of survival of extremely preterm infants below 25 weeks during the decade from 2000 to 2009. Our low survival rates of 19% at 23 weeks were very similar to data presented by © 2014 Nature America, Inc.
national or regional data from Switzerland, UK, Canada, Australia and NICHD, USA.21,24,36,37 The survival rates between 24 and 28 weeks in our study mirrored the figures reported by the NICHD between 2003 and 2007 at each week of gestation and were also similar to the Swedish data between 24 and 26 weeks and the Swiss data between 2005 and 2009.21,23,38 Unlike the Epicure 2 Journal of Perinatology (2014), 1 – 7
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
6 study which showed improved survival at 24 and 25 weeks between 1997 and 2006 we did not find any significant improvement in survival at these gestations from 2000 to 2009.26 However, our overall survival of 57% and 80% at 24 and 25 weeks, respectively, was slightly higher than the reported 47 and 67% survival at these two gestations in 2006 in Epicure 2.26 Our figures at these two gestations were also similar to the 63 and 79% survival at 24 and 25 weeks reported by Kutz et al.39 from a single center study in Germany. One possible reason for the higher survival in our study, as well as by Kutz et al.39 was that geographic population studies report lower survival figures than hospital-based studies. Our center, however, is the regional referral perinatal center and accounts for more than two thirds of infants born at the thresholds of viability in Singapore (unpublished data). The incidence of respiratory distress syndrome and need for, mechanical ventilation, surfactant therapy and postnatal steroids decreased with increasing GA whereas the use of exclusive CPAP support only increased significantly from 26 weeks onwards. The surfactant use in our cohort seemed lower and corresponding CPAP use higher in our study than the average NICHD figures between 26 and 28 weeks.21 The incidence of major morbidities such as CLD at 36 weeks postmenstrual age in survivors, culture-proven sepsis, severe ROP and severe IVH decreased with increasing gestation, whereas NEC/focal intestinal perforation did not show a similar trend across gestations. The presence of one or more major neonatal morbidities decreased with increasing GA similar to data from Sweden.3–5,18,19,40 Similar to data from Israel,41 we also looked at presence of mortality and morbidity as a composite morbidty in the babies provided with NICU care and it was seen that this adverse outcome was significantly associated with male sex, decreasing birth weight, the presence of significant hyaline membrane disease, airleaks and PDA. Delivery by cesarean section seemed to be protective and this has been reported by others at extreme prematurity.22,27,29 The increased adverse outcome in boys seen in our study has been well documented historically and from recent data the exact pathogenesis remains unclear.42 Intrauterine growth restriction has been reported as an independent risk factor for an adverse outcome in a recent meta-analysis by Damodaram et al.43 and Tyson et al.44 and in our study it was also seen to be predictive on univariate analysis. However, this significance did not persist on multivariate analysis. This additional knowledge of gender and growth restriction if available antenatally may be useful in refining the antenatal counseling in extreme prematurity. CRIB II is a tool commonly used to assess neonatal illness severity in preterm VLBW infants and in predicting in hospital neonatal mortality, as well as neurodevelopmental impairment at 36 months.11–13 In our study we also found CRIB II to be a predictor of composite morbidity on univariate analysis. Although birth weight, gestation and sex are not modifiable factors attempts to improve the body temperature or base excess could potentially improve CRIB scores and serve as a quality improvement tool. Similarly quality improvement tools aimed at reduction of airleaks may also help in reducing composite morbidity in our high-risk preterm population and is being evaluated. One of the strengths of our study was the systematic comprehensive data which was collected prospectively and spanned a period of 10 years. Adjustments were also made for the time effect thus minimizing the bias introduced due to a long study period. There was no change in either mortality or major morbidity over the 10 years of the study. There was, however, a significant reduction in mortality and morbidity noted with each increasing week of gestation. The other strength of our data is the inclusion of the live births not resuscitated in the delivery room in the denominator for survival estimation. This enabled us to overcome to a certain extent the evidence of selection bias reported by Evans45 in preterm survival studies with over estimation of survival in studies Journal of Perinatology (2014), 1 – 7
which did not take into account live births not resuscitated. Despite being a single-center study, our survival data are consistent with the findings reported from geographically defined populations and probably reflect the centralization of tertiary perinatal care in Singapore.36 Even within geographic areas, centralization of perinatal care for extremely preterm deliveries to centers of excellence has been recommended in Sweden, Finland and USA.38,46,47 Thus, our study has the advantage of reflecting centralized perinatal care in Singapore as KKH is the regional referral unit with the highest volume of extreme preterm births. This may to a certain extent overcome the limitation that our report does not describe a geographically based sample. Neonatal mortality and morbidity data are invaluable tools in the decision-making process along with the data on neurodevelopmental outcome in childhood. Thus, all extremely high-risk preterm infants need to be followed up for long-term neurodevelopmental morbidities and outcome, which is currently being done in our cohort up to 8 years of age. In conclusion, increasing survival and decreasing morbidity were seen with each increasing week in gestation without any significant alteration in survival trends over the decade, probably reflecting the limits of technological advances and biologic plausibility. CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS We thank Lim Wai Yee for her assistance with the manuscript.
REFERENCES 1 Fanaroff AA, Stoll BJ, Wright LL, Carlo WA, Ehrenkranz RA et al. Trends in neonatal morbidity and mortality for very low birthweight infants. Am J Obstet Gynecol 2007; 196(2): 147.e1–8. 2 Lundqvist P, Kallen K, Hallstorm I, Westas LH. Trends in outcomes for very preterm infants in the southern region of Sweden over a 10 year period. Acta Paediatr 2009; 98: 648–653. 3 Schmidt B, Asztalos EV, Roberts RS, Robertson CMT, Sauve RS et al. Impact of bronchopulmonary dysplasia, brain injury and severe retinopathy on the outcome of extremely low birth weight infants at 18 months. Results from the trial of indomethacin prophylaxis in preterm. JAMA 2003; 289: 1124–1129. 4 Bassler D, Stoll BJ, Schmidt B, Asztalos EV, Roberts RS et al. Using a count of neonatal morbidities to predict poor outcome in extremely low birth weight infants: Added role of neonatal infection. Pediatrics 2009; 123: 313–318. 5 Gargus RA, Vohr BR, Tyson JE, High P, Higgins RD, Wrage LA et al. Unimpaired outcomes for extremely low birth weight infants at 18 to 22 months. Pediatrics 2009; 124: 112–121. 6 Lubcheno LO, Hansman C, Bay E. Intrauterine growth as estimated from liveborn birthweight data at 24 to 42 weeks of gestation. Pediatrics 1963; 32: 793–800. 7 Fenton TR. A new growth chart for preterm babies: Babson and Benda’s chart updated with recent data and a new format. BMC Pediatr 2003; 3: 13. 8 Tita ATN, Andrews WW. Diagnosis and management of clinical chorioamnionitis. Clin Perinatol 2010; 37: 339–354. 9 Redline RW. Inflammatory response in acute chorioamnionitis. Semin Fetal Neonatal Med 2012; 17: 20–25. 10 Diagnosis and management of preeclampsia and eclampsia. ACOG Practice Bulletin No. 33. American College of Obstetricians and Gynecologists. Obstet Gynecol 2002; 99: 159–167. 11 Parry G, Tucker J, Tarnow-Mordi W. CRIB II: an update of the clinical risk index for babies score. UK Neonatal Staffing Study Collaborative Group. Lancet 2003; 361 (9371): 1789–1791. 12 Gagliardi L, Cavazza A, Brunelli A, Battaglioli M, Merazzi D et al. Assessing mortality risk in very low birthweight infants: a comparison of CRIB, CRIB-II, and SNAPPE-II. Arch Dis Child Fetal Neonatal Ed 2004; 89(5): F419–F422. 13 Lodha A, Sauve R, Chen S, Tang S, Christianson H. Clinical Risk Index for babies score for the prediction of neurodevelopmental outcomes at 3 years of age in infants of very low birthweight. Dev Med Child Neurol 2009; 51(11): 895–900. 14 Jobe AH, Bancalari E. Bronchopulmonary Dysplasia. Am J Respir Crit Care Med 2001; 163: 1723–1729.
© 2014 Nature America, Inc.
Neonatal outcome below 28 weeks of gestational age P Agarwal et al
7 15 Papile LA, Munsick-Bruno G, S Chaefer A. Relationship of cerebral intraventricular haemorrhage in early childhood neurological handicaps. J Pediatr 1983; 103: 293. 16 An International Committee for the Classification of Retinopathy of Prematurity. The International Classification of Retinopathy of Prematurity Revisited. Arch Ophthalmol. 2005; 123(7): 991–999. 17 Pumberger W, Mayr M, Kohlhauser C, Weninger M. Spontaneous localized intestinal perforation in very low birth weight infants: a distinct clinical entity different from necrotizing enterocolitis. J Am Coll Surg 2002; 195(6): 796–803. 18 Husain N, Galou M, Ehrerikranz RA, Henson VC, Rowe JC. Pre-discharge outcome of 22-27 weeks gestational age infants born at tertiary care centres in connecticut. Implications for perinatal management. Conn Med 1998; 62(3): 131–137. 19 Hintz SR, Poole WK, Wright LL. Changes in mortality and morbidities among infants born less than 25 weeks during the post-surfactant era. Arch Dis Child Fetal Neonatal Ed 2005; 90: F128–F133. 20 Lemons JA, Bauer CR, Oh W, Korones SB, Papile LA, Stoll BJ et al. Very low birth weight outcomes of the National Institue of Child Health and Human Development Neonatal Research Network, January 1995 through December 1996. NICHD Neonatal Research Network. Pediatrics 2001; 107: e1. 21 Stoll BJ, Hansen NI, Bell EF, Shankaran S, Laptok AR et al. Neonatal outcomes of extremely preterm infants from NICHD Neonatal Research Network. Pediatrics 2010; 126: 443. 22 Vanhaesebrouck TP, Allegaert K, Bottu J, Debauche C, Devlieger H et al. EPIBEL Study: Outcomes to discharge from hospital for extremely preterm infants in belgium for the EPIBEL study group. Pediatrics 2004; 114(3): 663–675. 23 Berger TM, Steurer MA, Woerner A, Schiffer PM, Adams M. Trends and centre to centre variability in survival rates of very preterm infants ( o 32 weeks) over a 10 year period in Switzerland. Arch Dis Child Fetal Neonatal Ed 2012; F1–F6. 24 Shah PS, Sankaran K, Aziz K, Allen AC, Seshia M, Ohlsson A et al. Canadian Neonatal Network Outcomes of preterm infants o 29 weeks gestation over 10-year period in Canada: a cause for concern? J Perinatol 32: 132–138. 25 Draper ES, Zeitlin J, Fenton AC. Investigating the variations in survival rates for very preterm infants in 10 European regions: the MOSAIC birth cohort. Arch Dis Child Fetal Neonatal Ed 2009; 94: F158–F163. 26 Costeloe K, Hennessy E, Haider S, Stacey F, Marlow N, Draper ES. Short term outcomes after extreme preterm birth in England: Comparison of two birth cohorts in 1995 and 2006 (the EPICURE studies). BMJ 2012; 345: e7976. 27 Hussain N, Rosenkrantz TS. Ethical considerations in the management of infants born at extremely low gestational age. Semin Perinatol 2003; 27(6): 458–470. 28 Tommiska V, Heinonen K, Lehtonen L, Renlund M, Saarela T, Tammela O et al. No improvement in outcome of nationwide extremely low birth weight infant populations between 1996-1997 and 1990-2000. Pediatrics 2007; 119: 29–36. 29 Markestad T, Kaaresen PI, Ronnestad A, Reigstad H, Lossius K, Medbo S et al. Norwegian extreme prematurity study group Early death, morbidity and need of treatment among extremely premature infants. Pediatrics 2005; 115: 1289–1298. 30 Morgan MA, Goldenberg RL, Schulkin J. Obstetrician-gynaecologist practices regarding preterm birth at the limit of viability. J Matern Fetal Neonatal Med 2008; 21(2): 115–121.
© 2014 Nature America, Inc.
31 Perinatal Care at the Threshold of Viability. ACOG Practice Bulletin. Obstet Gynecol 2002; 100: 617–624. 32 Nuffield Council on Bioethics Dilemmas in current practice: babies born at the borderline of viability. Critical Care Decisions in Fetal and Neonatal Medicine: Ethical Issues. Nuffield Council on Bioethics: London, UK, 2006, pp 67–87. 33 Pignotti MS, Donzelli G. Perinatal care at the threshold of viability: An international comparison of practical guidelines for the treatment of extremely preterm births. Pediatrics 2008; 121: e193. 34 Itabashi K, Horiuchi T, Kusuda S, Kabe K. Mortality rates for extremely low birthweight infants born in Japan in 2005. Pediatrics 2009; 123: 445–450. 35 Kyser KL, Morriss FH Jr, Bell EF, Klein JM, Dagle JM. Improving survival of extremely preterm infants born between 22 and 25 weeks of gestation. Obstet Gynecol 2012; 119(4): 795–800. 36 Field DJ, Dorling JS, Manktelow BN, Draper ES. Survival of extremely premature babies in a geographically defined population: prospective cohort study of 1994–1999 compared with 2000–2005. BMJ 2008; 336: 1221. 37 Fischer N, Steurer MA, Adams M, Berger TM. Survival rates of extremely preterm infants (gestational age o 26 weeks) in Switzerland: impact of the Swiss guidelines for the care of infants born at the limit of viability. Arch Dis Child Fetal Neonatal Ed 2009; 94(6): F407–F413. 38 The Express Team. One-year survival of extremely preterm infants after active perinatal care in Sweden. JAMA 2009; 301(21): 2225–2233. 39 Kutz P, Horsch S, Kuhn L, Roll C. Single centre vs population based outcome data of extremely preterm infants at the limits of viability. ACTA Paediatr 2009; 98: 1451–1455. 40 Stoll BJ, Hansen NI, Adams-Chapman I, Fanaroff AA, Hintz SR, Vohr B et al. Neurodevelopmental and growth impairment among extremely low birth weight infants with neonatal infection. JAMA 2004; 292: 2357–2365. 41 Kugelman A, Bader D, Lerner-Geva L, Boyko V, Levitzki O, Riskin A et al. Poor outcomes at discharge among extremely premature infants. Arch Pediatr Adolesc Med 2012; 166(6): 543–550. 42 Kent AL, Wright IM, Abdel-Latif ME. New South Wales and Australian Capital Territory Neonatal Intensive Care Units Audit Group Mortality and adverse neulogic outcomes are greater in preterm male infants. Pediatrics 2012; 129: 124–131. 43 Damodaram M, Story L, Kulinskaya E, Rutherford M, Kumar S. Early adverse perinatal complications in preterm growth-restricted fetuses. Aust N Z J Obstet Gynaecol 2011; 51: 204–209. 44 Tyson JE, Parikh NA, Langer J, Green C, Higgins RD. Intensive care for extreme prematurity—moving beyond gestational age. N Engl J Med 2008; 358: 1672–1681. 45 Evans DJ, Levene MI. Evidence of selection bias in preterm survival studies: a systematic review. Arch Dis Child Fetal Neonatal Ed 2001; 84: F79–F84. 46 Rautava L, Lehtonen L, Peltola M, Korvenranta E, Korvenranta H, Linna M et al. The effect of birth in secondary or tertiary level hospitals in Finland on mortality in very preterm infants: a birth-register study. Pediatrics 2007; 119: e257–e263. 47 Lorch SA, Baiocchi M, Ahlberg CE, Small DS. The differential impact of delivery hospital on the outcomes of premature infants. Pediatrics 2012; 130: 270–279.
Journal of Perinatology (2014), 1 – 7
