- Email: [email protected]
Predicting low apgar scores of infants weighing less than 1000 grams: The effect of corticosteroids
Predicting Low Apgar Scores of Infants Weighing Less Than 1000 Grams: The Effect of Corticosteroids MICHAEL 0. GARDNER, MD, ROBERT L. GOLDENBERG, MD, FRANCISCO L. GAUDIER, MD, MARY B. DUBARD, MA, KATHLEEN AND JOHN C. HAUTH, MD Objective: To evaluate maternal and neonatal factors that predict low Apgar scores in newborns weighing less than 1000 g. Methods: From a data set of all live-born infants who were delivered between 1979-1991 and who weighed 1000 g or less, we reviewed the records of 837 neonates born at 24-28 weeks’ gestation. Potential risk factors were evaluated for associations with a l-minute Apgar score of 3 or less and a 5-minute Apgar score of 6 or less. Analyses used ,$ test and multiple logistic regression. Results: The prevalence of l-minute Apgar scores of 3 or less decreased from 65.9% at 24 weeks to 38.2% at 28 weeks, and the prevalence of 5-minute Apgar scores of 6 or less decreased from 83.3% at 24 weeks to 51.2% at 28 weeks. As the birth weight increased from 500-599 g to 900-1000 g, l-minute Apgar scores of 3 or less decreased from 77.0% to 39.6%, and 5-minute Apgar scores of 6 or less decreased from 89.2% to 56.4%. Aside from gestational age and birth weight, corticosteroid use was the strongest predictor of Apgar scores above 3 at 1 minute and above 6 at 5 minutes. Male and nonvertex-presenting infants had an increased likelihood of low Apgar scores, as did infants with cord blood pH less than 7.05 or bicarbonate value less than 17 mEq/L. Conclusion: Neonates at very low gestational ages and birth weights are more likely than larger or more mature infants to have low Apgar scores. Males, nonvertexpresenting infants, and those who are acidotic at birth also have an increased prevalence of low scores. Infants born to mothers treated with antenatal corticosteroids are less likely to have low Apgar scores. This finding indicates that antenatal corticosteroids may benefit the newborn at birth,
From the Center for Obstetric Research, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, The University of Alabama at Birrninghnm, Biumingharn, Alabama. Supported in part by the Agency for Health Cure Policy and Research contract no. DHHS 282-92-0055. Presented nt the Annuul Clinical Meeting of The American College of Obstetricians and Gynecologists, Orlnndo, Florida, May S-22, 1994.
170
0029-7844/95/$9.50 0029-7844(94)00372-K
before
respiratory
(Obstet
Gynecol
distress
G. NELSON, MD,
syndrome
becomes
apparent.
1995;85:170-4)
Since its introduction over 40 years ago, the Apgar score has been used almost universally in the United States to assess the status of newborns.’ Many consider it a surrogate measure for asphyxia. However, because three of five components (respiratory effort, reflex irritability, and muscle tone) are related to developmental maturity, many have questioned if the Apgar score is reflective of neonatal well-being in the preterm infant.‘” Another study4 showed that Apgar scores do not correlate well with umbilical blood gas determinations in the preterm infant, and this has led some authors”,4 to conclude that low Apgar scores are not a specific indicator of asphyxia in the preterm newborn. A joint committee opinion of ACOG and the American Academy of Pediatrics stated that “the normal premature infant may thus receive a low score purely because of immaturity, with no evidence of anoxic insult or depression.“5 Recently, Gaudier et al (Gaudier FL, Goldenberg RL, Dubard M, Nelson KG, Hauth JC. Influence of acid-base status and Apgar scores on survival in 500 to 1,000 g infants [abstract]. Am J Obstet Gynecol 1994;170:387) found that for very low birth weight infants, the 5-minute Apgar score is a better predictor of neonatal survival than are umbilical cord gas values. Our study was performed to determine which maternal and fetal factors predict low Apgar scores in the preterm infant.
Materials and Methods The University of Alabama Extremely Low Birthweight Dataset has been described previously.6 We reviewed the records of all live-born infants who were delivered
Obstetrics & Gynecology
between 1979-1991 and who weighed between 500-1000 g. These neonates included 837 born between 24-28 weeks’ gestation. The best estimate of gestational age was determined by the date of the last menstrual period, physical examination, and sonographic data, if available. The data analyzed included the l- and 5-minute Apgar scores (which were nearly always assigned by a pediatric physician), the gestational age at delivery, birth weight, fetal presentation, and plurality. The maternal charts were reviewed for the presence of premature rupture of membranes (PROM), chorioamnionitis, and pre-delivery maternal treatment with corticosteroids. The mothers of 136 neonates were treated with corticosteroids. Cord blood gases, performed on 526 of the study neonates, were also abstracted for analysis. Twenty-four neonates had a pH less than 7.05, and 74 neonates had a bicarbonate value less than 17 mEq/L. None of the infants were given exogenous surfactant before assignment of Apgar scores. Because the data base included infants born over a 12-year period, we performed an analysis to preclude bias based on changes in obstetric and neonatal management over the years. The pattern of corticosteroid use did not change over the 12-year study period. Likewise, there was no significant change in the incidence of low Apgar scores from 1979 until 1991, when the data were stratified by gestational age or birth weight. Univariate analyses using 2 and MantelHaenszel tests were performed to calculate differences between groups based on gestational age or birth weight. Two multiple logistic regression analyses evaluated factors associated with a l-minute Apgar score of 3 or less and a 5-minute Apgar score of 6 or less. These scores were close to the median l- and 5-minute Apgar scores in our study population and have been used previously to define “low Apgar scores.“7’8 The first regression analysis used data from all infants born between 24-28 weeks’ gestation. Controlling for gestational age and birth weight, the analysis evaluated the effect on Apgar scores of fetal sex, plurality, position, chorioamnionitis, PROM, and antenatal maternal treatment with corticosteroids. The second regression analysis used data from 525 of the 837 maternal-infant pairs who had umbilical blood gas studies at delivery. Controlling for gestational age and birth weight again, a regression analysis was used to evaluate the effect on Apgar scores of fetal sex, position, chorioamnionitis, antenatal corticosteroids, and umbilical cord blood pH and bicarbonate levels. All data analyses were performed on SAS statistical software (SAS Institute, Inc., Cary, NC). Statistical significance was defined as P < .05.
VOL.
85, NO.
2, FEBRUARY
1995
Table
1. Apgar
Scores
No. of infants Median I-min Apgar score Median 5-min Apgar score 1-min Apgar score <3* 5-min Apgar score s6* * Distribution
significantly
Versus
Gestational
Age
24 weeks
25 weeks
26 weeks
27 weeks
28 weeks
123 2 5 65.9% 83.3%
185 2 5 63.9% 79.6%
222 3 6 56.7% 71.0%
185 5 7 36.5% 53.2%
122 4 7 38.2% 51.2%
different
(P < .OOl).
Rem1ts Table 1 summarizes the frequency of low Apgar scores and the median scores by week of gestational age from 24-28 weeks. The prevalence of low Apgar scores decreased with increasing gestational age. At 24 weeks, 83.3% of the neonates had low 5-minute Apgar scores, whereas only 51.2% of the 28-week neonates had low 5-minute Apgar scores. This trend was statistically significant (P < .OOl). Table 2 shows the median scores and the frequency of low Apgar scores stratified by birth weight. As with gestational age, the prevalence of low Apgar scores decreased with increasing birth weight. Of the infants with a birth weight between 500-599 g, 89.2% had a low 5-minute Apgar score; of those weighing between 900-1000 g, the prevalence of low 5-minute Apgar scores was 56.4%. This trend was also statistically significant (P < .OOl). Similar trends were seen using other cutoffs as a definition of a low Apgar score. Using regression techniques and controlling for gestational age and birth weight, we analyzed maternal and fetal factors associated with low Apgar scores. The odds ratios (ORs) and 95% confidence intervals of the dependent variables are summarized in Table 3. In this model, only male sex and nonvertex presentation were associated with an increased risk of both a low l- and 5-minute Apgar score, whereas corticosteroid use was associated with a decreased risk of low l- and 5-minute scores. In 525 infants, cord blood gas studies were available for review. We performed a second regression analysis, controlling for gestational age, birth weight, fetal sex,
Table
2. Apgar
Scores
No. of patients Median I-min Apgar score Median 5-min Apgar score I-min Apgar score ~3* 5-min Apgar score ~6* * Distribution
significantly
Gardner
et al
Versus
Birth
Weight
500 599 g
600699 g
700799 g
soo899 g
9001000 g
72 1.5 3 77.0% 89.2%
136 2 5 65.9% 80.4%
202 3 6 54.3% 69.6%
211 4 7 47.4% 62.9%’
216 4 7 39.6% 56.4%
different
(P < .OOl).
Apgar Scores in Preterm Infants
171
Table
3. Maternal Apgar
and Scores
Neonatal (n = 837)
Characteristics
l-minute score Characteristics Maternal PROM Preeclampsia Chorioamnionitis Cesarean delivery Antenatal corticosteroids Neonatal Race (black) Sex (male) Multiple birth Nonvertex presentation
Apgar 53
Versus
Low
5-minute score
Apgar 56
Odds ratio
95% Cl
Odds ratio
95% CI
1.2 0.6 1.4 1.7 0.4
0.9-l .8 0.4-l .o 0.9-2.1 1 .l-2.5 0.3-0.6
1.1 0.7 1.7 1.1 0.6
0.8-1.6 0.4-1.1 1.1-2.5 0.7-1.6 0.4-0.9
1.2 1.5 0.7 2.0
0.9-1.7 1.2-1.9 0.4-1.0 1.4-2.9
1.1 1.4 0.5 1.6
0.8 -1.5 1.0-1.9 0.3-0.8 1.1-2.3
PROM = premature rupture of membranes; Cl = confidence interval. Dependent variable: low l- or 5-minute Apgar scores. Independent variables in the logistic regression model include gestational age at delivery and birth weight.
PROM, and antenatal corticosteroid treatment. In addition, umbilical cord arterial pH, with a cutoff value of less than 7.05, and umbilical cord arterial bicarbonate, with a cutoff of less than 17 mEq/L, were added to the analysis as dependent variables (Table 4). In this analysis, a pH value less than 7.05 was associated with an OR of 6.5 for a low l-minute score and an OR of 4.9 for a low S-minute Apgar score. A bicarbonate value less than 17 mEq/L was associated with about a twofold increase in low l- and 5-minute scores, whereas a nonvertex presentation and amnionitis were also associated with an increase in low scores. In this multiple logistic regression model, prenatal corticosteroid position,
Table
4. Maternal,
Neonatal,
Characteristics
Characteristics Maternal PROM Chorioamnionitis Antenatal corticosteroids Neonatal Sex (male) Nonvertex presentation Umbilical artery pH ~7.05 Bicarbonate ~17 mEq/L PROM
= premature
and Umbilical
Versus
rupture
Low
Apgar
l-minute score
Apgar 53
Cord Gas Scores
(M = 525)
j-minute score
Apgar 56
Odds ratio
95%) Cl
Odds ratio
95% Cl
1.4 1.5 0.3
0.9-2.1 0.9-2.4 0.2-0.5
1.0 1.7 0.3
0.6-1.5 1.1-2.5 0.2-0.6
1.3 2.4 6.5 2.1
0.9-1.9 1.8-3.3 1.7-24.5 1.2-3.7
1.4 1.5 4.9 1.8
0.9-2.0 1.1-2.0 1.5-15.9 1 .O-3.2
of membranes;
Cl = confidence
inter-
val. Dependent variable: low I- or 5-minute Apgar scores. Independent variables in the logistic regression mode1 include gestational age at delivery and birth weight.
172
Gardner
et al
Apgar
Scores
in Preterm
infants
treatment was associated with an OR of 0.3 for both low l- and 5-minute Apgar scores.
Discussion The Apgar scoring system was designed to evaluate quickly the need for immediate newborn resuscitation.’ With time, low Apgar scores have become a surrogate marker for asphyxia.8,9 We demonstrated previously that low Apgar scores are very common in preterm infants, even those with normal cord pH values, and therefore were not an accurate indicator of asphyxia.4 Catlin et al* also found low Apgar scores in preterm neonates and showed that the degree of muscle tone, respiratory effort, and reflex irritability could be correlated with increasing gestational age. Low Apgar scores have not proven to be very good predictors of major developmental handicaps in the preterm infant.l”,l’ In spite of their limitations, the Apgar scores of preterm infants have been shown to be good predictors of neonatal mortality (Gaudier FL. Am J Obstet Gynecol 1994;170:387).12 Therefore, it is useful to review which factors may predict low Apgar scores in the preterm neonate. As expected, increasing gestational age and birth weight were each related inversely to the prevalence of low Apgar scores. Male infants, nonvertexpresenting infants, and those delivered by cesarean also had a higher risk of low Apgar scores. Severe cord blood acidosis, defined as a pH level less than 7.05, was associated with an OR of 4.9 for a low 5-minute Apgar score, whereas a low 5-minute Apgar score in infants with a bicarbonate value less than 17 mEq/L had an OR of 1.8 when compared to infants with higher bicarbonate values. Stark et al3 did not find a correlation between cord blood pH values and the incidence of 5-minute Apgar scores of 6 or less. However, in their study, none of the very low birth weight infants had a cord blood pH less than 7.05. Severe acidosis has been shown previously to correlate with an increased risk of low Apgar scores in both preterm and term infants.‘“-I6 Our findings agree with these studies. Perhaps the most notable finding in our study is the relation between maternal treatment with corticosteroids and the incidence of low l- and 5-minute Apgar scores in neonates less than 1000 g. When all other factors (including gestational age, birth weight, and cord blood gas values) are controlled, the odds of finding a low Apgar score in infants who received antenatal corticosteroid treatment was less than half that of the untreated infants. The reason for the association between corticosteroid use and higher Apgar scores is not clear. Because corticosteroid treatment was not randomized, selection bias may provide part of the explanation. However,
Obstetrics
& Gynecology
antenatal corticosteroid treatment has also been demonstrated to have a maturational effect on a large number of specific enzyme systems and organ functions, with some of these effects occurring within hours of maternal administration.17 These maturational effects may be apparent in the clinical status of very small infants at delivery, as reflected by higher Apgar scores. Antenatal corticosteroids are associated with maturation of fetal lungs, which results in a decreased incidence of respiratory distress syndrome. Corticosteroids are also associated with a reduction in the risk of intraventricular hemorrhage, which appears likely to be independent of the pulmonary effects, and a reduction in the risk of necrotizing enterocolitis.‘s,i9 There may also be an effect on cardiovascular system maturation as manifested by an infant’s ability to sustain a higher blood pressure.” However, the evidence that best confirms the positive effect of prenatal corticosteroid use on Apgar scores comes from randomized trials already published. A review of 14 English-language, randomized corticosteroid trials revealed nine that reported data describing the Apgar scores in the corticosteroid treatment and control groups.‘s,20-27 In eight of the nine trials, the l-minute Apgar scores were higher in the corticosteroid group, and in seven of the nine, the corticosteroid group had higher 5-minute Apgar scores. Although a typical meta-analysis cannot be performed because the Apgar data is presented in different formats, the likelihood that eight of nine studies would show results in the same direction by chance is 2%, and that seven of nine studies would show results in the same direction by chance is 9%. We conclude that our results from an observational data set are confirmed by published data available from randomized trials. Therefore, our findings of improved Apgar scores associated with corticosteroid use provide evidence for a positive effect of corticosteroids on infant development that precedes and perhaps contributes to the protective effect of this hormone on the development of respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, and, ultimately, neonatal death.‘9*2’
References 1. Apgar V. A proposal newborn infant. Cm-r
for a new method of evaluation Res Anesth Analg 1953;32:260-7.
of
the
2. Catlin EA, Carpenter MW, Brann BS, et al. The Apgar score revisited: Influence of gestational age. ] I’ediatr 1986;109:865-8. 3. Stark CF, Gibbs RS, Freedman WL. Comparison of umbilical artery pH and 5-minute Apgar score in the low birthweight and very birthweight infant. Am J Obstet Gynecol 1990;163:818-23. 4. Goldenberg RL, Huddleston JF, Nelson KG. Apgar scores umbilical arterial pH in preterm newborn infants. Am Gynecol 1984;149:651-4. 5. The American College of Obstetricians and Gynecologists
VOL.
85, NO.
2, FEBRUARY
1995
low and
J Obstet Com-
mittee Apgar
on Maternal and score. Committee
American 6. Gaudier birth
Fetal Medicine. Use and misuse of the Opinion no. 49. Washington, DC: The
College of Obstetricians and Gynecologists, 1986. FL, Goldenberg RL, Nelson KG, et al. Acid-base status
and subsequent
neurosensory
1,000 g infants. Am J Obstet 7. Cunningham FG, MacDonald LC III. Williams
in surviving
Gynecol 1994;170:48-53. PC, Gant NF, Leveno
Obstetrics.
19th ed. Norwalk,
ton & Lange, 1993:444-5. 8. Apgar V. The newborn (Apgar) Am
impairment
scoring
KJ, Gilstrap
Connecticut:
system.
Pediatr
AppleClin North
1966;13:645-50.
9. Paneth
N, Fox HE. The relationship
handicap: A survey of clinicians. 10. Nelson K, Ellenberg J. Apgar neurologic
disability.
Pediatrics
of Apgar Obstet scores
score
to neurologic
Gynecol 1983;61:547-50. as predictors of chronic
1981;68:36-44.
11. Behnke M, Eyler F, Carter R, Hardt N, Cruz A, Predictive value of Apgar scores for developmental premature
infants.
12. Jepson HA, limitations, 13. Hibbard and
at
500 to
Am J Perinatal
JU, Hibbard
MC,
Whalen
and morbidity
Obstet Gynecol 14. Goldaber KG,
Ml’.
Umbilical
in the very
1991;78:768-73. Gilstrap LG, Leveno
acidemia.
at term Am
as related
J Obstet
acidemia. Am J Obstet 17. Ballard RA. Antenatal
Evolution,
cord
birth
blood
weight
gas
infant.
JS. Pathologic
J, Brumfield
to the
Gynecol
16. Goodwin TM, Belai I, Hernandez ial complications in the term
low
KJ, Dax
acidemia. Obstet Gynecol 1991;78:1103-7. 15. Winkler CL, Hauth JC, Tucker JM, Owen tal complications
M. in
1989;6:18-21.
Talashek ML, Tichy AM. The Apgar score: and scoring guidelines. Birth 1991;18:83-92.
mortality
artery
Resnick outcome
fetal
CC. Neona-
degree
of umbilical
1991;164:637-41.
I’, Durand M, Paw RH. Asphyxnewborn with severe umbilical
Gynecol 1992;162:1506-12. glucocorticoid therapy:
Clinical
effects.
Monogr Endocrinol 1986;28:137-72. 18. Garite TJ, Rumney PJ, Briggs GG, et al. A randomized, placebocontrolled trial of betamethasone for the prevention of respiratory distress Gynecol 19. Crowley
syndrome at 24 to 28 weeks’ gestation. Am J Obstet 1992;166:646-51. P, Chalmers 1, Keirse MJNC. The effects of corticosteroid
administration dence from
before controlled
preterm delivery: trials. Br J Obstet
An overview of the eviGynaecol 1990;97:11-25.
20. Kari MA, Hallman M, Eronen M, et al. Prenatal dexamethasone treatment in conjunction with rescue therapy of human surfactant. A randomized placebo-controlled multicenter study. Pediatrics 1994;93:730-6. 21. Liggins GC, Howie RN. A controlled trial of antepartum glucocorticoid treatment for prevention of the respiratory distress syndrome in premature infants. Pediatrics 1972;50:515-25. 22. Taeusch HW, Frigoletto F, Kitzmiller J, et al. Risk of respiratory distress syndrome after prenatal dexamethasone treatment. Pediatrics 1979;63:64-72. 23. Doran TA, Swyer P, MacMurray B, et al. Results of a double-blind controlled study on the use of betamethasone in the prevention of respiratory distress syndrome. Am J Obstet Gynecol 1980;136:31320. 24. Schmidt PL, Sims ME, Strassner HT, Paul RH, Mueller E, McCart D. Effect of antepartum natal respiratory distress
glucocorticoid syndrome
and
administration upon neoperinatal infection. Am J
Obstet Gynecol 1984;148:178-86. 25. Iams JD, Talbert ML, Barrows H, Sachs L. Management prematurely ruptured membranes: A prospective comparison of observation versus J Obstet Gynecol 1985;151:32-8. 26. Morales WJ, Diebel ND, Lazar antenatal
dexamethasone
Gardner
et al
steroids
and timed
AJ, Zadrozny
on the prevention
of preterm randomized delivery.
D. The
of respiratory
Apgar Scores in Preterm
Infants
Am
effect
of
distress
173
syndrome in preterm gestations with premature rupture of membranes. Am J Obstet Gynecol 1986;154:591-5. 27. Gansu HR, Mullinger BM, Donnai P, Dash CH. Antenatal administration of betamethasone to prevent respiratory distress syndrome in preterm infants: Report of a UK multicenter trial. Br J Obstet Gynaecol 1989;96:401-10. Address
reprint
requests
to:
Robert L. Goldenberg, MD Tke University of Alabama at Bivmingkam Division of Maternal-Fetal Medicine Department of Obstetrics and Gynecology 619 South 20tk Street Bivmingkam, AL 35233-7333
174
Gardner
et al
Received May 23, 1994. Received in revised form October 3, 1994 Accepted October 18, 1994.
Apgar Scores in Preterm infants
Copyright 0 1995 Gynecologists.
by The
American
College
of Obstetricians
and
Obstetrics B Gynecology
From the Center for Obstetric Research, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, The University of Alabama at Birrninghnm, Biumingharn, Alabama. Supported in part by the Agency for Health Cure Policy and Research contract no. DHHS 282-92-0055. Presented nt the Annuul Clinical Meeting of The American College of Obstetricians and Gynecologists, Orlnndo, Florida, May S-22, 1994.
170
0029-7844/95/$9.50 0029-7844(94)00372-K
before
respiratory
(Obstet
Gynecol
distress
G. NELSON, MD,
syndrome
becomes
apparent.
1995;85:170-4)
Since its introduction over 40 years ago, the Apgar score has been used almost universally in the United States to assess the status of newborns.’ Many consider it a surrogate measure for asphyxia. However, because three of five components (respiratory effort, reflex irritability, and muscle tone) are related to developmental maturity, many have questioned if the Apgar score is reflective of neonatal well-being in the preterm infant.‘” Another study4 showed that Apgar scores do not correlate well with umbilical blood gas determinations in the preterm infant, and this has led some authors”,4 to conclude that low Apgar scores are not a specific indicator of asphyxia in the preterm newborn. A joint committee opinion of ACOG and the American Academy of Pediatrics stated that “the normal premature infant may thus receive a low score purely because of immaturity, with no evidence of anoxic insult or depression.“5 Recently, Gaudier et al (Gaudier FL, Goldenberg RL, Dubard M, Nelson KG, Hauth JC. Influence of acid-base status and Apgar scores on survival in 500 to 1,000 g infants [abstract]. Am J Obstet Gynecol 1994;170:387) found that for very low birth weight infants, the 5-minute Apgar score is a better predictor of neonatal survival than are umbilical cord gas values. Our study was performed to determine which maternal and fetal factors predict low Apgar scores in the preterm infant.
Materials and Methods The University of Alabama Extremely Low Birthweight Dataset has been described previously.6 We reviewed the records of all live-born infants who were delivered
Obstetrics & Gynecology
between 1979-1991 and who weighed between 500-1000 g. These neonates included 837 born between 24-28 weeks’ gestation. The best estimate of gestational age was determined by the date of the last menstrual period, physical examination, and sonographic data, if available. The data analyzed included the l- and 5-minute Apgar scores (which were nearly always assigned by a pediatric physician), the gestational age at delivery, birth weight, fetal presentation, and plurality. The maternal charts were reviewed for the presence of premature rupture of membranes (PROM), chorioamnionitis, and pre-delivery maternal treatment with corticosteroids. The mothers of 136 neonates were treated with corticosteroids. Cord blood gases, performed on 526 of the study neonates, were also abstracted for analysis. Twenty-four neonates had a pH less than 7.05, and 74 neonates had a bicarbonate value less than 17 mEq/L. None of the infants were given exogenous surfactant before assignment of Apgar scores. Because the data base included infants born over a 12-year period, we performed an analysis to preclude bias based on changes in obstetric and neonatal management over the years. The pattern of corticosteroid use did not change over the 12-year study period. Likewise, there was no significant change in the incidence of low Apgar scores from 1979 until 1991, when the data were stratified by gestational age or birth weight. Univariate analyses using 2 and MantelHaenszel tests were performed to calculate differences between groups based on gestational age or birth weight. Two multiple logistic regression analyses evaluated factors associated with a l-minute Apgar score of 3 or less and a 5-minute Apgar score of 6 or less. These scores were close to the median l- and 5-minute Apgar scores in our study population and have been used previously to define “low Apgar scores.“7’8 The first regression analysis used data from all infants born between 24-28 weeks’ gestation. Controlling for gestational age and birth weight, the analysis evaluated the effect on Apgar scores of fetal sex, plurality, position, chorioamnionitis, PROM, and antenatal maternal treatment with corticosteroids. The second regression analysis used data from 525 of the 837 maternal-infant pairs who had umbilical blood gas studies at delivery. Controlling for gestational age and birth weight again, a regression analysis was used to evaluate the effect on Apgar scores of fetal sex, position, chorioamnionitis, antenatal corticosteroids, and umbilical cord blood pH and bicarbonate levels. All data analyses were performed on SAS statistical software (SAS Institute, Inc., Cary, NC). Statistical significance was defined as P < .05.
VOL.
85, NO.
2, FEBRUARY
1995
Table
1. Apgar
Scores
No. of infants Median I-min Apgar score Median 5-min Apgar score 1-min Apgar score <3* 5-min Apgar score s6* * Distribution
significantly
Versus
Gestational
Age
24 weeks
25 weeks
26 weeks
27 weeks
28 weeks
123 2 5 65.9% 83.3%
185 2 5 63.9% 79.6%
222 3 6 56.7% 71.0%
185 5 7 36.5% 53.2%
122 4 7 38.2% 51.2%
different
(P < .OOl).
Rem1ts Table 1 summarizes the frequency of low Apgar scores and the median scores by week of gestational age from 24-28 weeks. The prevalence of low Apgar scores decreased with increasing gestational age. At 24 weeks, 83.3% of the neonates had low 5-minute Apgar scores, whereas only 51.2% of the 28-week neonates had low 5-minute Apgar scores. This trend was statistically significant (P < .OOl). Table 2 shows the median scores and the frequency of low Apgar scores stratified by birth weight. As with gestational age, the prevalence of low Apgar scores decreased with increasing birth weight. Of the infants with a birth weight between 500-599 g, 89.2% had a low 5-minute Apgar score; of those weighing between 900-1000 g, the prevalence of low 5-minute Apgar scores was 56.4%. This trend was also statistically significant (P < .OOl). Similar trends were seen using other cutoffs as a definition of a low Apgar score. Using regression techniques and controlling for gestational age and birth weight, we analyzed maternal and fetal factors associated with low Apgar scores. The odds ratios (ORs) and 95% confidence intervals of the dependent variables are summarized in Table 3. In this model, only male sex and nonvertex presentation were associated with an increased risk of both a low l- and 5-minute Apgar score, whereas corticosteroid use was associated with a decreased risk of low l- and 5-minute scores. In 525 infants, cord blood gas studies were available for review. We performed a second regression analysis, controlling for gestational age, birth weight, fetal sex,
Table
2. Apgar
Scores
No. of patients Median I-min Apgar score Median 5-min Apgar score I-min Apgar score ~3* 5-min Apgar score ~6* * Distribution
significantly
Gardner
et al
Versus
Birth
Weight
500 599 g
600699 g
700799 g
soo899 g
9001000 g
72 1.5 3 77.0% 89.2%
136 2 5 65.9% 80.4%
202 3 6 54.3% 69.6%
211 4 7 47.4% 62.9%’
216 4 7 39.6% 56.4%
different
(P < .OOl).
Apgar Scores in Preterm Infants
171
Table
3. Maternal Apgar
and Scores
Neonatal (n = 837)
Characteristics
l-minute score Characteristics Maternal PROM Preeclampsia Chorioamnionitis Cesarean delivery Antenatal corticosteroids Neonatal Race (black) Sex (male) Multiple birth Nonvertex presentation
Apgar 53
Versus
Low
5-minute score
Apgar 56
Odds ratio
95% Cl
Odds ratio
95% CI
1.2 0.6 1.4 1.7 0.4
0.9-l .8 0.4-l .o 0.9-2.1 1 .l-2.5 0.3-0.6
1.1 0.7 1.7 1.1 0.6
0.8-1.6 0.4-1.1 1.1-2.5 0.7-1.6 0.4-0.9
1.2 1.5 0.7 2.0
0.9-1.7 1.2-1.9 0.4-1.0 1.4-2.9
1.1 1.4 0.5 1.6
0.8 -1.5 1.0-1.9 0.3-0.8 1.1-2.3
PROM = premature rupture of membranes; Cl = confidence interval. Dependent variable: low l- or 5-minute Apgar scores. Independent variables in the logistic regression model include gestational age at delivery and birth weight.
PROM, and antenatal corticosteroid treatment. In addition, umbilical cord arterial pH, with a cutoff value of less than 7.05, and umbilical cord arterial bicarbonate, with a cutoff of less than 17 mEq/L, were added to the analysis as dependent variables (Table 4). In this analysis, a pH value less than 7.05 was associated with an OR of 6.5 for a low l-minute score and an OR of 4.9 for a low S-minute Apgar score. A bicarbonate value less than 17 mEq/L was associated with about a twofold increase in low l- and 5-minute scores, whereas a nonvertex presentation and amnionitis were also associated with an increase in low scores. In this multiple logistic regression model, prenatal corticosteroid position,
Table
4. Maternal,
Neonatal,
Characteristics
Characteristics Maternal PROM Chorioamnionitis Antenatal corticosteroids Neonatal Sex (male) Nonvertex presentation Umbilical artery pH ~7.05 Bicarbonate ~17 mEq/L PROM
= premature
and Umbilical
Versus
rupture
Low
Apgar
l-minute score
Apgar 53
Cord Gas Scores
(M = 525)
j-minute score
Apgar 56
Odds ratio
95%) Cl
Odds ratio
95% Cl
1.4 1.5 0.3
0.9-2.1 0.9-2.4 0.2-0.5
1.0 1.7 0.3
0.6-1.5 1.1-2.5 0.2-0.6
1.3 2.4 6.5 2.1
0.9-1.9 1.8-3.3 1.7-24.5 1.2-3.7
1.4 1.5 4.9 1.8
0.9-2.0 1.1-2.0 1.5-15.9 1 .O-3.2
of membranes;
Cl = confidence
inter-
val. Dependent variable: low I- or 5-minute Apgar scores. Independent variables in the logistic regression mode1 include gestational age at delivery and birth weight.
172
Gardner
et al
Apgar
Scores
in Preterm
infants
treatment was associated with an OR of 0.3 for both low l- and 5-minute Apgar scores.
Discussion The Apgar scoring system was designed to evaluate quickly the need for immediate newborn resuscitation.’ With time, low Apgar scores have become a surrogate marker for asphyxia.8,9 We demonstrated previously that low Apgar scores are very common in preterm infants, even those with normal cord pH values, and therefore were not an accurate indicator of asphyxia.4 Catlin et al* also found low Apgar scores in preterm neonates and showed that the degree of muscle tone, respiratory effort, and reflex irritability could be correlated with increasing gestational age. Low Apgar scores have not proven to be very good predictors of major developmental handicaps in the preterm infant.l”,l’ In spite of their limitations, the Apgar scores of preterm infants have been shown to be good predictors of neonatal mortality (Gaudier FL. Am J Obstet Gynecol 1994;170:387).12 Therefore, it is useful to review which factors may predict low Apgar scores in the preterm neonate. As expected, increasing gestational age and birth weight were each related inversely to the prevalence of low Apgar scores. Male infants, nonvertexpresenting infants, and those delivered by cesarean also had a higher risk of low Apgar scores. Severe cord blood acidosis, defined as a pH level less than 7.05, was associated with an OR of 4.9 for a low 5-minute Apgar score, whereas a low 5-minute Apgar score in infants with a bicarbonate value less than 17 mEq/L had an OR of 1.8 when compared to infants with higher bicarbonate values. Stark et al3 did not find a correlation between cord blood pH values and the incidence of 5-minute Apgar scores of 6 or less. However, in their study, none of the very low birth weight infants had a cord blood pH less than 7.05. Severe acidosis has been shown previously to correlate with an increased risk of low Apgar scores in both preterm and term infants.‘“-I6 Our findings agree with these studies. Perhaps the most notable finding in our study is the relation between maternal treatment with corticosteroids and the incidence of low l- and 5-minute Apgar scores in neonates less than 1000 g. When all other factors (including gestational age, birth weight, and cord blood gas values) are controlled, the odds of finding a low Apgar score in infants who received antenatal corticosteroid treatment was less than half that of the untreated infants. The reason for the association between corticosteroid use and higher Apgar scores is not clear. Because corticosteroid treatment was not randomized, selection bias may provide part of the explanation. However,
Obstetrics
& Gynecology
antenatal corticosteroid treatment has also been demonstrated to have a maturational effect on a large number of specific enzyme systems and organ functions, with some of these effects occurring within hours of maternal administration.17 These maturational effects may be apparent in the clinical status of very small infants at delivery, as reflected by higher Apgar scores. Antenatal corticosteroids are associated with maturation of fetal lungs, which results in a decreased incidence of respiratory distress syndrome. Corticosteroids are also associated with a reduction in the risk of intraventricular hemorrhage, which appears likely to be independent of the pulmonary effects, and a reduction in the risk of necrotizing enterocolitis.‘s,i9 There may also be an effect on cardiovascular system maturation as manifested by an infant’s ability to sustain a higher blood pressure.” However, the evidence that best confirms the positive effect of prenatal corticosteroid use on Apgar scores comes from randomized trials already published. A review of 14 English-language, randomized corticosteroid trials revealed nine that reported data describing the Apgar scores in the corticosteroid treatment and control groups.‘s,20-27 In eight of the nine trials, the l-minute Apgar scores were higher in the corticosteroid group, and in seven of the nine, the corticosteroid group had higher 5-minute Apgar scores. Although a typical meta-analysis cannot be performed because the Apgar data is presented in different formats, the likelihood that eight of nine studies would show results in the same direction by chance is 2%, and that seven of nine studies would show results in the same direction by chance is 9%. We conclude that our results from an observational data set are confirmed by published data available from randomized trials. Therefore, our findings of improved Apgar scores associated with corticosteroid use provide evidence for a positive effect of corticosteroids on infant development that precedes and perhaps contributes to the protective effect of this hormone on the development of respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, and, ultimately, neonatal death.‘9*2’
References 1. Apgar V. A proposal newborn infant. Cm-r
for a new method of evaluation Res Anesth Analg 1953;32:260-7.
of
the
2. Catlin EA, Carpenter MW, Brann BS, et al. The Apgar score revisited: Influence of gestational age. ] I’ediatr 1986;109:865-8. 3. Stark CF, Gibbs RS, Freedman WL. Comparison of umbilical artery pH and 5-minute Apgar score in the low birthweight and very birthweight infant. Am J Obstet Gynecol 1990;163:818-23. 4. Goldenberg RL, Huddleston JF, Nelson KG. Apgar scores umbilical arterial pH in preterm newborn infants. Am Gynecol 1984;149:651-4. 5. The American College of Obstetricians and Gynecologists
VOL.
85, NO.
2, FEBRUARY
1995
low and
J Obstet Com-
mittee Apgar
on Maternal and score. Committee
American 6. Gaudier birth
Fetal Medicine. Use and misuse of the Opinion no. 49. Washington, DC: The
College of Obstetricians and Gynecologists, 1986. FL, Goldenberg RL, Nelson KG, et al. Acid-base status
and subsequent
neurosensory
1,000 g infants. Am J Obstet 7. Cunningham FG, MacDonald LC III. Williams
in surviving
Gynecol 1994;170:48-53. PC, Gant NF, Leveno
Obstetrics.
19th ed. Norwalk,
ton & Lange, 1993:444-5. 8. Apgar V. The newborn (Apgar) Am
impairment
scoring
KJ, Gilstrap
Connecticut:
system.
Pediatr
AppleClin North
1966;13:645-50.
9. Paneth
N, Fox HE. The relationship
handicap: A survey of clinicians. 10. Nelson K, Ellenberg J. Apgar neurologic
disability.
Pediatrics
of Apgar Obstet scores
score
to neurologic
Gynecol 1983;61:547-50. as predictors of chronic
1981;68:36-44.
11. Behnke M, Eyler F, Carter R, Hardt N, Cruz A, Predictive value of Apgar scores for developmental premature
infants.
12. Jepson HA, limitations, 13. Hibbard and
at
500 to
Am J Perinatal
JU, Hibbard
MC,
Whalen
and morbidity
Obstet Gynecol 14. Goldaber KG,
Ml’.
Umbilical
in the very
1991;78:768-73. Gilstrap LG, Leveno
acidemia.
at term Am
as related
J Obstet
acidemia. Am J Obstet 17. Ballard RA. Antenatal
Evolution,
cord
birth
blood
weight
gas
infant.
JS. Pathologic
J, Brumfield
to the
Gynecol
16. Goodwin TM, Belai I, Hernandez ial complications in the term
low
KJ, Dax
acidemia. Obstet Gynecol 1991;78:1103-7. 15. Winkler CL, Hauth JC, Tucker JM, Owen tal complications
M. in
1989;6:18-21.
Talashek ML, Tichy AM. The Apgar score: and scoring guidelines. Birth 1991;18:83-92.
mortality
artery
Resnick outcome
fetal
CC. Neona-
degree
of umbilical
1991;164:637-41.
I’, Durand M, Paw RH. Asphyxnewborn with severe umbilical
Gynecol 1992;162:1506-12. glucocorticoid therapy:
Clinical
effects.
Monogr Endocrinol 1986;28:137-72. 18. Garite TJ, Rumney PJ, Briggs GG, et al. A randomized, placebocontrolled trial of betamethasone for the prevention of respiratory distress Gynecol 19. Crowley
syndrome at 24 to 28 weeks’ gestation. Am J Obstet 1992;166:646-51. P, Chalmers 1, Keirse MJNC. The effects of corticosteroid
administration dence from
before controlled
preterm delivery: trials. Br J Obstet
An overview of the eviGynaecol 1990;97:11-25.
20. Kari MA, Hallman M, Eronen M, et al. Prenatal dexamethasone treatment in conjunction with rescue therapy of human surfactant. A randomized placebo-controlled multicenter study. Pediatrics 1994;93:730-6. 21. Liggins GC, Howie RN. A controlled trial of antepartum glucocorticoid treatment for prevention of the respiratory distress syndrome in premature infants. Pediatrics 1972;50:515-25. 22. Taeusch HW, Frigoletto F, Kitzmiller J, et al. Risk of respiratory distress syndrome after prenatal dexamethasone treatment. Pediatrics 1979;63:64-72. 23. Doran TA, Swyer P, MacMurray B, et al. Results of a double-blind controlled study on the use of betamethasone in the prevention of respiratory distress syndrome. Am J Obstet Gynecol 1980;136:31320. 24. Schmidt PL, Sims ME, Strassner HT, Paul RH, Mueller E, McCart D. Effect of antepartum natal respiratory distress
glucocorticoid syndrome
and
administration upon neoperinatal infection. Am J
Obstet Gynecol 1984;148:178-86. 25. Iams JD, Talbert ML, Barrows H, Sachs L. Management prematurely ruptured membranes: A prospective comparison of observation versus J Obstet Gynecol 1985;151:32-8. 26. Morales WJ, Diebel ND, Lazar antenatal
dexamethasone
Gardner
et al
steroids
and timed
AJ, Zadrozny
on the prevention
of preterm randomized delivery.
D. The
of respiratory
Apgar Scores in Preterm
Infants
Am
effect
of
distress
173
syndrome in preterm gestations with premature rupture of membranes. Am J Obstet Gynecol 1986;154:591-5. 27. Gansu HR, Mullinger BM, Donnai P, Dash CH. Antenatal administration of betamethasone to prevent respiratory distress syndrome in preterm infants: Report of a UK multicenter trial. Br J Obstet Gynaecol 1989;96:401-10. Address
reprint
requests
to:
Robert L. Goldenberg, MD Tke University of Alabama at Bivmingkam Division of Maternal-Fetal Medicine Department of Obstetrics and Gynecology 619 South 20tk Street Bivmingkam, AL 35233-7333
174
Gardner
et al
Received May 23, 1994. Received in revised form October 3, 1994 Accepted October 18, 1994.
Apgar Scores in Preterm infants
Copyright 0 1995 Gynecologists.
by The
American
College
of Obstetricians
and
Obstetrics B Gynecology