Objective
To determine whether survival is different in “early” (23 0/7 -23 3/7 weeks) vs “late” (23 4/7 -23 6/7 weeks) infants.
Study Design
Records of 126 consecutive liveborn infants delivered at 23 0/7 to 23 6/7 weeks’ gestation from 2001-2010 were examined using the Vermont Oxford Network database. Infants born at 23 0/7 to 23 3/7 weeks were grouped into “early” and those at 23 4/7 to 23 6/7 weeks were “late.” Clinical characteristics were compared between groups and multivariate analyses were used to predict survival.
Results
Seventy-two infants were early and 54 were late. Survival was 25% vs 56%, respectively ( P < .001). The early group was less likely to receive steroids (43% vs 65%; P = .016) and had a lower mean birthweight (547 g vs 596 g; P < .001). No difference in other factors was seen between groups. No change in survival was observed during the study period in either group.
Conclusion
Late 23-week infants have improved survival compared with early infants. Delaying delivery as little as 24-96 hours may improve survival for 23-week infants.
Survival, short- and long-term morbidity at the limits of viability are important components of neonatal and obstetric counseling. They also influence many obstetric management decisions, including the decision to intervene on fetal behalf and mode of delivery. These decisions are quite vexing, important, and controversial. Although these infants represent less than half of 1% of births in our institution, they account for a sizable percentage of perinatal deaths. Previous studies have noted that practitioners underestimate survival and overestimate morbidity. This in turn influences patient counseling and can affect obstetric decisions to intervene for fetal benefit and parental desire to initiate and continue care after delivery.
Rates of neonatal morbidity and mortality from preterm delivery has progressively improved over the years. However, this benefit has been shown mostly in infants born at 24-weeks’ gestation and greater. Data demonstrating improvement in outcomes of infants born in 23-weeks’ gestation are sparse and unconvincing despite newer therapies and treatments. Longitudinal studies performed worldwide have shown that decisions to intervene and resuscitate in week 23 are inconsistent and vary between and within countries including the United States, Japan, Israel, Sweden, Canada, and the United Kingdom. These decisions also impact antenatal management decisions such as use of steroids, tocolysis, and mode of delivery. In addition, there is weak evidence that these interventions improve the postnatal outcome in 23-week infants.
Many studies have examined outcome and survival in 23-week infants only in terms of complete weeks of gestational age and few have studied fractional portions within a given week. Moreover, previous research has focused on the value of each additional week of gestation, rather than on smaller increments of time, on survival, and neonatal intensive care unit (NICU) length of stay in pregnancies at older gestational ages. The benefit of these additional times in the 23-week population is not well studied.
Our primary aim is to examine infant survival of infants born at 23-weeks’ gestation and compare outcome between those born earlier in gestation vs later. We hypothesized that there is improved survival in infants born later (23 4/7 -23 6/7 weeks) vs earlier (23 0/7 -23 3/7 weeks) during the 23rd completed week of gestation.
Materials and Methods
Institutional review board approval was obtained for this study. We examined a cohort of 126 consecutive live born infants between 23 0/7 and 23 6/7 weeks’ gestation treated at the St. Louis University Perinatal Center, which includes Saint Mary’s Health Center and Cardinal Glennon Children’s Medical Center, over a 10-year period (2001 to 2010). Data was collected and recorded using criteria of the Vermont Oxford Network database (VON). VON describes that criteria for gestational age in weeks and days, being determined using a hierarchy of the best available obstetric parameters, including prenatal ultrasound. This takes precedence over neonatal estimation for gestational age. All infants in our cohort were dated with first or second trimester ultrasound in corroboration with maternal last menstrual period as appropriate. Definitions of antenatal steroid use, prenatal care, respiratory distress syndrome, chronic lung disease, sepsis, periventricular-intraventicular hemorrhage, retinopathy of prematurity, and necrotizing enterocolitis were as described by VON. Survival was defined as alive at first birthday or at hospital discharge, whichever event occurred first.
Infants born at 23 0/7 to 23 3/7 weeks were grouped into “early” and those at 23 4/7 to 23 6/7 weeks were “late.” Statistical comparisons between groups were made using χ 2 , t tests, and Kolmogorov-Smirnov as appropriate with regards to race, multiple gestation, antenatal steroid use, birthweight, Apgar score, inborn status, cesarean delivery, sex and neonatal complications, and survival. Analysis was performed with the entire population as well as with singleton gestations analyzed separately from twins. Multiple logistic regression was used to analyze those variables that influenced survival. Variables were chosen that were found to be significant in the univariate analysis, or had been noted to be significantly associated with survival in other studies.
To account for infants where a decision had been made not to resuscitate, a group called “comfort care” was created. This group was defined by the absence of oxygen, bag mask ventilation, surfactant administration, and endotracheal tube ventilation. Analysis was done both with “comfort care” infants included and excluded.
A P value of < .05 was used to denote statistical significance. All analyses were performed using SPSS version 18.0 for Windows (SPSS Inc, Chicago, IL).
Results
Prenatal and delivery characteristics for the entire cohort are described in Table 1 . The cohort was mostly African American with 72 infants born early and 54 late. About half of the cohort received antenatal steroids and most received surfactant. Table 2 shows the frequency and severity of neonatal morbidity and demonstrates almost universal morbidity in survivors.
| Characteristic | N = 126, n (%) |
|---|---|
| PRENATAL | |
| Race | |
| African American | 67 (53.2) |
| White | 58 (46) |
| Hispanic | 1 (0.8) |
| Location of birth | |
| Inborn | 103 (81.7) |
| Outborn | 23 (18.3) |
| Prenatal care | 120 (95.2) |
| Antenatal steroids | 66 (52.4) |
| Cesarean delivery | 26 (20.6) |
| Gestational age at delivery | |
| Early (23 0/7–23 3/7) | 72 (57.1) |
| Late (23 4/7–23 6/7) | 54 (42.9) |
| Multiple gestation | 28 (22.2) |
| DELIVERY | |
| Infant birthweight, g (median and IQR) | 558 (518.3–599.3) |
| Died in delivery room | 22 (17.5) |
| Left delivery room, died ≤12 h a | 12 (11.5) |
| Male infant | 71 (56.3) |
| 1-min Apgar score (median and IQR) | 2 (1–4) |
| 5-min Apgar score (median and IQR) | 4 (2–7) |
| Received surfactant in the delivery room | 69 (54.8) |
| Received surfactant outside delivery room | 102 (81) |
a Calculated for the 104 infants who left delivery room alive.
| Characteristic | N = 126, n (%) |
|---|---|
| Respiratory distress syndrome | 98 (94.2) |
| Early sepsis a | 4 (3.8) |
| Periventricular/intraventricular hemorrhage b | |
| Any | 57 (68.7) |
| Severe (grade 3 or 4) | 34 (41) |
| None | 26 (31.3) |
| Chronic lung disease c | 44 (95.7) |
| Necrotizing enterocolitis a | 11 (10.6) |
| Retinopathy of prematurity d | |
| Any | 47 (94) |
| Severe (grade 3 or 4) | 16 (32) |
| None | 3 (6) |
a Calculated for the 104 infants who left delivery room alive;
b Calculated from surviving 83 infants at time of diagnosis;
c Calculated from surviving 46 infants at time of diagnosis;
d Calculated from surviving 50 infants at time of diagnosis.
A comparison of the 2 groups is shown in Table 3 . No difference was seen between the groups in terms of mode of delivery, sex, or multiple gestations. The early group was less likely to have received antenatal steroids ( P = .016). A separate analysis of the cohort showed that 66 received antenatal steroids although 60 did not. Survival was 46% in the steroid group vs 30% in the no steroid group. This comparison did not reach statistical significance ( P = .11). This did not change when “comfort care” infants were excluded from the analysis (50% vs 35%; P = .1).
| Characteristic | Early (n = 72) (%) | Late (n = 54) (%) | P value |
|---|---|---|---|
| African American race | 41 (56.9) | 26 (48.1) | .388 |
| Mean birthweight, g | 546.92 | 595.98 | .016 |
| Comfort care only a | 12 (16.7) | 2 (3.7) | .045 |
| Inborn delivery | 56 (77.8) | 47 (87) | .183 |
| Delivery by cesarean section | 14 (19.5) | 12 (22.2) | .703 |
| Received antenatal steroids | 31 (43) | 35 (64.8) | .016 |
| Multiple gestation | 17 (23.6) | 11 (20.4) | .665 |
| 1-min Apgar score (median) | 2 | 2 | .848 |
| 5-min Apgar score (median) | 4 | 4 | .319 |
| Delivery room surfactant | 33 (45.8) | 36 (66.7) | .02 |
| Surfactant at any time | 54 (75) | 48 (88.9) | .049 |
| Died in delivery room | 16 (22.2) | 6 (11.1) | .104 |
| Died within 12 h | 9 (12.5) | 3 (5.6) | .137 |
| Male sex | 36 (50) | 35 (64.8) | .097 |
| Survival (2001-2005) b , c | 9 (23.7) | 17 (68) | < .001 |
| Survival (2006-2010) b , d | 9 (26.5) | 13 (44.8) | .128 |
| Overall survival b | 18 (25) | 30 (56) | < .001 |
a Comfort care defined as no oxygen, no bag mask, no endotracheal intubation, and no surfactant;
b Survival defined as alive at 1 y of age or hospital discharge;
c 2001-2005: early (n = 38) and late (n = 25);
There were a greater number of “comfort care” infants in the early group. All the “comfort care” infants died within 12 hours of delivery. Analyses on neonatal morbidities, including respiratory distress syndrome, necrotizing enterocolitis, and chronic lung disease did not show a significant difference between the early and late groups. Survival was improved in the late group ( P < .001). Among survivors, 2 were defined by reaching 1 year of age and all of the others were alive at discharge from the hospital.
Sixty-three infants were born in each half of the 10-year study period. There was no significant improvement ( P = .58) when overall survival in the first half of the decade (41%, n = 26) was compared with the second half (35%, n = 22). Table 4 shows similar analyses to Table 3 with “comfort care” infants excluded. There was no change in significance in any of the variables tested.
| Characteristic | Early (n = 60) (%) | Late (n = 52) (%) | P value |
|---|---|---|---|
| African American race | 30 (56.9) | 26 (48.1) | .839 |
| Mean birthweight, g | 557.53 | 599.87 | .049 |
| Received antenatal steroids | 25 (41.7) | 35 (67.3) | .007 |
| Multiple gestation | 15 (25) | 11 (21.2) | .631 |
| 1-min Apgar score (median) | 2.5 | 2 | .817 |
| 5-min Apgar score (median) | 4 | 5 | .463 |
| Died in delivery room | 6 (10) | 4 (7.6) | .669 |
| Died within 12 h | 7 (11.7) | 3 (5.8) | .255 |
| Male sex | 30 (50) | 34 (65.4) | .101 |
| Survival (2001-2005) b , c | 9 (26.5) c | 17 (68) c | .001 |
| Survival (2006-2010) b , d | 9 (34.6) d | 13 (48.1) d | .318 |
| Overall survival b | 18 (30) | 30 (57.7) | .003 |
a Comfort care defined as no oxygen, no bag mask, no endotracheal intubation, and no surfactant given at any time;
b Survival defined as 1 y of age or hospital discharge;
c 2001-2005: early (n = 34) and late (n = 25);
Table 5 shows a comparison of early vs late singleton gestations and the survival benefit for late vs early 23-week neonates is still seen. Table 6 shows early vs late twin gestations with the “comfort care” infants excluded. There is not a statistically significant survival benefit seen in the early vs late twin 23-week gestations.
| Characteristic | Early (n = 45) (%) | Late (n = 41) (%) | P value |
|---|---|---|---|
| African American race | 22 (48.9) | 19 (46.3) | .813 |
| Mean birthweight, g | 554.84 | 613.90 | .001 |
| Inborn delivery | 30 (66.7) | 35 (85.4) | .044 |
| Delivery by cesarean section | 11 (24.4) | 9 (22) | .785 |
| Received antenatal steroids | 16 (35.6) | 26 (63.4) | .01 |
| Delivery room surfactant | 22 (40) | 28 (68.3) | .068 |
| Surfactant at any time | 40 (88.9) | 38 (92.7) | .545 |
| Died in delivery room | 5 (11.1) | 3 (7.3) | .716 |
| Died within 12 h | 6 (13.3) | 2 (4.9) | .264 |
| Male sex | 23 (51.1) | 26 (63.4) | .25 |
| Survival (2001-2005) a , b | 6 (22.2) | 13 (56.1) | .002 |
| Survival (2006-2010) a , c | 6 (33.3) | 11 (50) | .289 |
| Overall survival a | 12 (26.7) | 24 (58.5) | .003 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
