Objective. The use of intrapartum antibiotics to prevent early-onset group B streptococcal (EOGBS) infection has left pediatricians in a quandary about the appropriate evaluation and treatment of infants at risk for this infection. The aim of this study was to determine whether intrapartum antibiotic prophylaxis changed the constellation and timing of onset of clinical signs of group B streptococcal (GBS) infection in term infants.
Methodology. We conducted a retrospective chart review of infants who had EOGBS infection and were born in Southern California Kaiser Permanente Hospitals from 1988 through 1996. Objective criteria were used to ascertain maternal risk of infection, intrapartum antibiotic prophylaxis, and onset of clinical signs of infection.
Results. Three hundred nineteen infants with EOGBS sepsis, bacteremia, or clinically suspected infection were identified from a population of 277 912 live births. Of the 172 term infants with culture-positive infection who had clinical signs of infection, 95% exhibited them in the first 24 hours of life. All of the infants exposed to intrapartum antibiotics became ill within the first 24 hours of life.
Conclusions. Exposure to antibiotics during labor did not change the clinical spectrum of disease or the onset of clinical signs of infection within 24 hours of birth for term infants with EOGBS infection. A 48-hour stay is not required to monitor asymptomatic term infants exposed to intrapartum antibiotics for onset of GBS infection.
- group B streptococcal infection
- intrapartum antibiotics
- clinical signs
- maternal risk
- neonatal infections
- GBS =
- group B streptococcus(al) •
- EOGBS =
- early-onset group B streptococcus •
- CDC =
- Centers for Disease Control and Prevention •
- ICD-9 =
- International Classification of Diseases, Ninth Revision
In the 1970s, group B streptococcus (GBS) emerged as the most common bacterial pathogen in the neonatal period.1,2Despite advances in knowledge about the epidemiology, clinical manifestations, and vertical transmission from mother to infant, it continues to be the leading cause of neonatal infection.3,4 Attempts to decrease the incidence and severity of this potentially lethal infection have focused on preventing transmission from mother to infant during labor by administration of intrapartum antibiotics.5–10
The use of intrapartum antibiotics to prevent early-onset (≤6 days of age) group B streptococcus (EOGBS) infection has raised many questions about the appropriate evaluation and treatment of infants at risk for this infection.11–13 Algorithms for the evaluation and treatment of such infants have been suggested, but these approaches have not been validated by controlled trials.14,15 Given that 10% to 25% of pregnant women may be treated with antibiotics during labor,16 decisions about the evaluation and treatment of infants at risk for GBS infections affect a large number of newborns and may have substantial implications for hospital expenditures and resource allocation.12,13,17,18
We conducted a retrospective chart review of infants with EOGBS infection born from 1988 through 1996 in Southern California Kaiser Permanente Hospitals to evaluate the relationship between intrapartum antibiotic use and the clinical signs of infection and timing of onset of these signs in term infants with EOGBS infections. The risk-based strategy for the prevention of EOGBS infection was introduced in 1990 and was adopted officially as the clinical policy in 1994.
The primary aim of this study was to determine whether the constellation of infant signs and the timing of onset of these clinical signs changed with use of antibiotic prophylaxis during labor. In addition, we wanted to determine what proportion of women who delivered infants with EOGBS infection had risk factors as defined by the risk based approach of the Centers for Disease Control and Prevention (CDC),14 and to ascertain what proportion of women with risk factors received intrapartum antibiotics.
Approval to undertake this study was granted by the Kaiser Permanente-Southern California Institutional Review Board.
Potential cases were identified from 2 sources:International Classification of Diseases, Ninth Revision(ICD-9) codes in hospital discharge summaries and laboratory bacteriology data files. All infants born in a Southern California Kaiser Permanente hospital from January 1988 through December 1996 who were admitted to the hospital within the first 6 days of life and who had a discharge diagnosis of GBS infection, septicemia, or meningitis (ICD-9 codes 038.0, 320.2, 041.0, and 482.3) were included. Bacteriology files, available from 1990 through 1996, were reviewed to identify infants who had a positive culture for GBS in their blood or cerebral spinal fluid. After potential cases were identified, each infant's medical record was reviewed to ascertain whether the infant had a positive culture or a positive urine antigen test result for GBS or whether GBS infection was mentioned in the discharge diagnoses or discharge summary. If any of these criteria were met and the infant was 6 days old or younger at the time that the infection was diagnosed, the charts for the mother and infant were abstracted. Infants readmitted for treatment of GBS infection within the first 6 days of life had information from both hospitalizations abstracted separately.
A neonatologist (P.B.) reviewed all abstraction forms to verify that the infant had EOGBS infection. The infection was classified into 1 of 3 categories, based on the results of laboratory testing for GBS in different body fluids and clinical signs of infection exhibited by the infant. Infants who had a positive blood or cerebral spinal fluid culture result and had a clinical sign of infection were classified as septic. Infants with a positive blood culture result and no clinical signs were classified as bacteremic. Infants were classified as having clinically suspected GBS infection if there were clinical signs of infection in 2 or more categories and they had positive superficial culture results (urine, tracheal aspirate, or gastric aspirate), a positive bacterial antigen screen, or their mother had positive intrapartum culture results for GBS. Neonatal blood cultures were performed using the nonradiometric BACTEC (Becton Dickinson Co, Cockeysville, MD) blood culture system from 1988 through 1992 and the fluorometric BacT/Alert (Organon Teknika, Durham, NC) automated blood culture instrument from 1993 through 1996.
Maternal risk factors as defined by the CDC risk factor criteria14 were ascertained from information abstracted from the mother's chart. Duration of rupture of membranes was coded as <18 hours (no risk) or 18 hours or greater (risk). Highest maternal temperature during labor was coded as <38°C (no risk) or 38°C or greater (risk). Bacteriuria was identified by reviewing each chart for a urine culture result positive for GBS obtained at least 48 hours before delivery. Gestational age was based on the pediatrician's assessment in the infant's chart or taken from the obstetrics record if the pediatric assessment was not available. Preterm birth was defined as a birth before 37 weeks of completed gestation. Women with a previous infant who had had GBS infection were identified based on the prenatal record as well as the discharge summary of the mother and infant.
Information on antibiotics with antimicrobial activity against GBS (ampicillin, penicillin, cefazolin, clindamycin, and erythromycin) given to the mother during the 48 hours before delivery was recorded. Duration of antibiotic prophylaxis during the 48 hours before delivery was categorized as antibiotics given <4 hours before birth, at least 4 hours before birth, and no antibiotics given. Antibiotics were also categorized by number of doses. Two or more antibiotics administered simultaneously were counted as 1 dose. At least 4 hours between each dose of any drug were required for it to be considered a separate dose. Antibiotics given >48 hours before delivery or at the time of delivery were not classified as intrapartum prophylaxis.
Information on the presence or absence of newborn clinical signs that were indicative of sepsis was abstracted from the medical record of each infant. Clinical signs of infection were based on the 4 systems involved in the inflammatory response to infection: respiratory, cardiovascular, temperature control, and alterations in the white blood cell count.19 Respiratory signs included respiratory rate >60 breaths per minute, grunting, apnea, need for oxygen outside of delivery room, or need for mechanical ventilation.20Cardiovascular signs included heart rate >160 beats per minute at rest,21 poor perfusion, blood pressure <5th percentile for age, bicarbonate level <15 mEq/liter, pH <7.30, need for volume expansion, or need for vasopressor support. Temperature instability was defined as rectal temperature <36.5°C or >38.0°C. With 1 modification, the 7-point hematologic scoring system of Rodwell et al22 was used for this study. We modified the score to consider a leukocyte count to be abnormal if it was >25 000 regardless of when the blood was drawn, because we were unable to ascertain the time the blood was drawn. A hematologic score of ≥3 is considered abnormal.22 All of the blood tests used to calculate the score were drawn before the infants were treated with antibiotics.
Each category of clinical signs was comprised of several individual signs. Only 1 sign was needed in order for the category to be coded as present. Timing of onset of clinical signs was recorded with the exception of the abnormal hematologic score. Timing of onset of signs was categorized as beginning during the first 6 hours after birth, >6 hours to 24 hours, >24 hours to 48 hours, >48 hours after birth, and no clinical signs present. If the timing of onset of a particular sign could not be ascertained, we used the time of blood draw for sepsis work-up or time of admission to the special care nursery.
The date and time of the first dose of antibiotics administered to the infant were ascertained from the medical record. This was categorized as antibiotics given during the first 6 hours after birth, >6 hours to 24 hours, >24 hours to 48 hours, >48 hours after birth, and no antibiotics given.
SAS statistical software (SAS Institute, Cary, NC) was used for data analysis.23 Differences between categorical variables were tested using the χ2 statistic and differences between continuous variables were tested using the log-rank χ2 (χ) test statistic for nonparametric distributions. Table 1 presents information for infants with positive deep culture results, and Table 2 presents information on infants with clinically suspected infections.
Identification of Patients
From 1988 through 1996, 277 912 infants were born in Southern California Kaiser Permanente Hospitals. Hospital discharge data (ICD-9 codes) and bacteriology laboratory data identified 604 infants as potentially having EOGBS infection (Fig 1). Two hundred forty-six infants were excluded by preliminary review because they did not meet the eligibility criteria or because their medical records could not be located. The medical record abstractors abstracted the medical records of 358 mother/infant pairs that met eligibility requirements. An additional 39 infants were excluded after review by the neonatologist (P.B.) because she determined that these infants did not meet the case definition for EOGBS infection.
Of the 319 infants identified as having EOGBS infection, 212 had positive deep body fluid cultures (septic or bacteremic) and 107 were classified as having clinically suspected infection. The overall rate of EOGBS culture-positive infection was .8/1000 live births and 1.1/1000 live births when both infants with clinically suspected and culture-positive infections were considered. Five infants in this cohort died. Three were preterm and 2 were term. The mortality rate for infants with GBS-positive deep body fluid cultures was 2.3%. None of the infants with clinically suspected infection died.
Characteristics of the Study Sample
We excluded 49 preterm infants (which included 1 set of twins), 8 term infants with unknown maternal risk, and 7 term infants with unknown maternal intrapartum antibiotic information from the cohort of 319 mother–infant pairs to focus on 255 term infants and their mothers for whom we had complete risk factor and intrapartum antibiotic information. One hundred forty-four of these mothers (56%) had a risk factor for GBS infection and 81 mothers (56%) with a risk factor received intrapartum antibiotics.
One hundred fifty-six infants were septic (61%), 16 were bacteremic (6%), and 83 had clinically suspected infection (33%). Table 1presents information on the clinical characteristics of EOGBS infection for the 172 term infants with sepsis and bacteremia. Table 2 presents information on the 83 term infants with clinically suspected EOGBS infection. Forty-six of the infants with clinically suspected infections (55%) had either positive superficial cultures (urine, tracheal aspirate, or gastric aspirate) or were born to women who had positive intrapartum cultures for GBS. Thirty-seven of the clinically suspected cases (45%) had positive urine bacterial antigen test results for GBS as the criteria for inclusion as GBS disease.
Infants with positive culture results and infants with clinically suspected infections had similar frequency and severity of clinical signs of infection and received similar treatments (data not shown). Although there was no significant (P < .05) difference in treatment modalities required, infants with clinically suspected infection had a higher prevalence of cardiovascular signs (P = .04), a higher peak heart rate (P= .008), and a lower absolute neutrophil count (P = .04), compared with infants with culture-positive infection. Cardiovascular and respiratory signs predominated in infants with both culture-positive and clinically suspected infection.
Of the 81 infants exposed to intrapartum antibiotics, 33 (41%) had culture-positive infections and 48 (59%) had clinically suspected EOGBS infection. Of the women who received intrapartum antibiotics whose infants had culture-positive infections, 26 (79%) received 1 dose, 5 (15%) received >1 dose, and 2 (6%) had missing dosage information. Of the women whose infants had clinically suspected infection, 41 (85%) received 1 dose and 7 (15%) received >1 dose of intrapartum antibiotics.
Infants With Culture-Positive EOGBS Infection
Of the 172 term infants with culture-positive infection, 33 (19%) were born to mothers who were given intrapartum antibiotics, 48 (28%) were born to mothers with a risk factor who did not receive intrapartum antibiotics, and 91 (53%) were born to mothers without a risk factor. Of these 172 infants, 164 (95%) exhibited at least 1 clinical sign of infection and 8 (5%) had no clinical sign of infection (Table 1). Of the 164 infants with a clinical sign of infection, 156 (95%) presented in the first 24 hours of life and 8 (5%) presented after 24 hours of age. Clinical signs of infection appeared in the first 24 hours after birth in the majority of infants, regardless of obstetrical risk or intrapartum antibiotic prophylaxis. Timing of onset of clinical signs of infection varied by maternal risk and intrapartum antibiotic treatment (log-rank χ2 = 6.3; 2 df;P = .04). When we performed a pairwise comparison of the 3 groups, we found that the timing of onset of clinical signs of infection for infants born to mothers who had a risk factor and received intrapartum antibiotics was earlier than that for infants born to women who had no risk factors for infection and who did not receive intrapartum antibiotics (χ2 = 6.3; 1df; P = .01). Differences between the other 2 pairs were not significant.
Intrapartum antibiotics did not delay the initiation of antibiotic treatment for infants. The majority of infants were started on antibiotics in the first day of life, regardless of obstetrical risk or intrapartum antibiotic exposure. One hundred thirty-six of the 172 infants in the cohort (79%) were given antibiotics by 24 hours of age. Twenty-four infants (14%) received antibiotics between 24 and 48 hours and 12 infants (7%) received antibiotics after 48 hours. The distribution of the timing of first antibiotic dose was the same for the 3 groups of infants, regardless of maternal risk or intrapartum antibiotic exposure (χ2 = 3.5; 2 df;P = .18). Of the 36 infants who received antibiotics after 24 hours of life, 22 remained in the hospital until they were treated and 14 were readmitted for treatment. Of the 22 infants treated during the birth hospitalization, 14 exhibited clinical signs of infection during the first 24 hours, 4 developed signs after 24 hours, and 4 exhibited no clinical signs of infection.
Infants With Clinically Suspected EOGBS Infections
Of the 83 term infants with clinically suspected EOGBS infection, 48 (58%) were born to mothers who were given intrapartum antibiotics, 15 (18%) were born to mothers with a risk factor who did not receive intrapartum antibiotics, and 20 (24%) were born to mothers without a risk factor. All 83 infants (100%) presented with clinical signs of infection in the first 24 hours of life, with the majority (95%) presenting in the first 6 hours of life. Timing of onset of clinical signs of infection varied by maternal risk and intrapartum antibiotic treatment (log-rank χ2 = 18.5; 2 df;P = .0001). When we performed a pairwise comparison of the 3 groups, we found that infants born to mothers who had risk factors and received intrapartum antibiotics had onset of clinical signs of infection earlier than did infants born to women who were at risk for infection but who did not receive intrapartum antibiotics (χ2 = 6.8; 1 df; P = .01) and for infants whose mothers had no risk factors (χ2 = 18.4; 1 df; P= .0001).
The majority of infants with clinically suspected infection were started on antibiotics in the first day of life, regardless of obstetrical risk or intrapartum antibiotic exposure. Seventy-six of the 83 infants (93%) with clinically suspected infection were given antibiotics by 24 hours of age. Five infants (6%) received antibiotics between 24 and 48 hours, 1 infant (1%) received antibiotics after 48 hours, and 1 infant (1%) did not receive antibiotics. The distribution of the timing of first antibiotic dose differed by maternal risk and intrapartum antibiotic exposure (χ2 = 12.5; 2df; P = .002). When we performed a pairwise comparison of the 3 groups, infants born to mothers who had a risk factor and who received intrapartum antibiotics received postnatal antibiotics sooner than did infants whose mothers had a risk factor but did not receive intrapartum antibiotic treatment (χ2 = 6.0; 1 df; P = .1) and infants born to mothers without a risk factor (χ2 = 8.4; 1 df; P = .004).
Six infants with clinically suspected infection were treated with antibiotics after 24 hours of life. Five remained in the hospital until they were treated and 1 infant was readmitted for treatment. All of the infants exhibited a clinical sign of infection during the first 24 hours of life.
Infants Readmitted for Treatment
Fifteen term infants (6%) were readmitted to the hospital for treatment of EOGBS infection. Three infants had meningitis, 5 infants had EOGBS sepsis, 6 had bacteremia, and 1 had a clinically suspected infection. Eight of the 15 infants were born to women with a risk factor for EOGBS infection and none of the readmitted infants had been exposed to intrapartum antibiotics. Twelve of the 15 readmitted infants had some clinical signs of infection recorded during the first hospital admission. Nine of the infants were recalled for treatment after being identified by blood culture results (8 infants) or urine bacterial antigen testing (1 infant) obtained during the initial hospitalization, because of clinical signs of illness or maternal risk factors.
Number of Doses of Intrapartum Antibiotics
We analyzed the data by number of doses of intrapartum antibiotic treatment (data not shown). Only 5 of the 172 infants (3%) with positive culture results and 7 of the 83 infants (8%) with clinically suspected infections were born to mothers who received >1 dose of intrapartum antibiotics.
Our study presents data from a large cohort of term infants with EOGBS infections, where maternal risk factors and intrapartum antibiotic treatment are related to the infant's clinical illness and timing of onset of clinical signs of infection. Administration of antibiotics to the mother during labor did not change the clinical spectrum of neonatal illness and it did not delay the onset of clinical signs of infection or postnatal treatment of the infant. The great majority of infants in this study had clinical signs of infection during the first 24 hours after birth and received antibiotics by 24 hours of age.
Previous studies have documented that 88% to 94% of infants with EOGBS infection present in the first 24 hours after birth.3,7,24,25 Other studies have reported a similar early presentation of illness in infants with proven GBS sepsis whose mothers have received intrapartum antibiotic prophylaxis.26–28 Term infants with EOGBS sepsis characteristically have problems during the transition, with lower Apgar scores and cardiovascular and respiratory instability in the first hours after birth.3,24,25,29 Eighty percent of the infants with culture-positive infection and 95% of infants with clinically suspected infection in this study presented with signs of infection by 6 hours of age. The spectrum and onset of clinical signs of infection remain the same regardless of obstetrical risk factors or antibiotic treatment during labor. Close observation of all term infants during transition, regardless of obstetrical risk factors or intrapartum antibiotic prophylaxis, is of paramount importance in recognizing infants with neonatal sepsis. More than one half of the term infants with culture-proven infection were born to mothers with no obstetrical risk factors.
Our study suggests that clinical signs of infection are important markers for the presence of infection in the neonate. Only 8 infants of our cohort of 255 infants never had any clinical sign of infection. Early clinical signs of infection in neonates are nonspecific and insensitive.30 Nonspecific signs of instability during transition occur in ∼5% of infants.17 Persistence of abnormal signs and the presence of 2 or more signs increase the likelihood of infection.30 Although the scoring system used to categorize infants as having infection had the potential to label infants with mild transitional problems as exhibiting signs of infection, each case was reviewed by a neonatologist to verify that infants identified as having clinically suspected infection had significant clinical illness.
Six percent of our term infants with EOGBS infection had asymptomatic bacteremia, a finding comparable to other series that report 4% to 20% of EOGBS infection as bacteremia.24,26–28,31 Because as many as 9% of cord blood cultures obtained at delivery are positive for bacteria,32 transient bacteremia may be a relatively common event after delivery for neonates. Persistence of bacteremia in otherwise asymptomatic term infants has been documented.33Infants with bacteremia were identified because of maternal risk factors during labor, although occasionally they were identified when mothers became ill after delivery. More than one half of the infants in this series who were readmitted for treatment of EOGBS infection was identified by blood cultures drawn during the first hospitalization. At readmission, 1 of these infants had cleared the bacteremia, 6 continued to have bacteremia, 1 had developed meningitis, and 1 was not retested on readmission. Obtaining blood cultures from infants born to mothers with maternal risk factors who have not received adequate intrapartum antibiotic therapy may identify infants with asymptomatic bacteremia who are at risk for presenting later with sepsis, meningitis, or focal infection.
In this 9-year cohort, 57% of the women with a risk factor for infection received intrapartum antibiotics. The risk-based approach to preventing GBS infection was introduced 3 years into the study and fully implemented only during the last 3 years of the study. We chose this period because we wanted to compare infants born to mothers with risk factors who received intrapartum antibiotics with infants whose mothers had risk factors and did not receive intrapartum antibiotics to determine whether the use of antibiotics during labor changed the timing of onset of infection.
Although intrapartum antibiotic prophylaxis reduces the incidence of EOGBS infection, it does not totally prevent EOGBS disease. In this study, 19% of culture-positive infants were exposed to intrapartum antibiotics during labor, which is comparable to other studies.26–28,34,35 Although we did not identify a treatment duration or the number of doses at which prevention of neonatal infection is ensured, only 3% of infants with culture-proven infection received 2 or more doses of antibiotics during labor.
We included a group of infants with clinically suspected infection because of the concern that previous exposure to antibiotics may obscure the diagnosis of infection by depressing bacterial growth in blood cultures. Other studies have documented that administration of antibiotics before obtaining blood cultures will decrease the frequency of positive blood cultures in clinically ill infants.36,37A preponderance of infants with clinically suspected infection were exposed to antibiotic during labor. Although superficial cultures and bacterial antigen tests are not recommended as methods to diagnose infections in neonates,38,39 combining clinical signs of infection with these ancillary tests identified a group of infants with a diagnosis of presumptive infection who were colonized with GBS.40,41 The onset of clinical signs of infection and the spectrum of illness in this group were similar to infants with culture-proven infection and were not affected by exposure to antibiotics during labor. As the use of intrapartum antibiotics to prevent neonatal sepsis increases, tests that detect inflammatory responses to infection and are independent of growth of bacteria in sterile body fluids may become increasingly important adjuncts to the diagnosis of infections in neonates.42,43
The low incidence of presentation of EOGBS infections after 24 hours of age, regardless of obstetrical risk factors or intrapartum antibiotic treatment, suggests a need to reconsider the current recommendations of the CDC and American Academy of Pediatrics for observation of infants at risk for GBS infection for 48 hours in the hospital.14,15 Our study found that 26 of the 42 infants with EOGBS who received their first antibiotic treatment after 24 hours of age had maternal risk factors. If an 8% prematurity rate is applied to the cohort of 277 912 live-born infants, 255 680 term infants were born during the study. Assuming that 25% of the mothers of these infants had obstetrical risk factors,16 63 920 infants would be observed 1 extra day to identify an additional 26 infants whose mothers had maternal risk factors who will become ill. The attack rate on that day would be .04%, a rate less than that of patients without risk factors for EOGBS.44
The decision to observe term infants past 24 hours of age in the hospital for the onset of EOGBS infection should rest on the evaluation of the clinical status of the infant, not on the presence of maternal risk factors or intrapartum antibiotic treatment. The findings of this study suggest that, when adequate early neonatal observation and outpatient follow-up is ensured, a 48-hour stay is not required to monitor asymptomatic term infants exposed to intrapartum antibiotics for onset of EOGBS infection.
This research was funded through the Kaiser Permanente Community Service Funds.
We acknowledge the Southern California Kaiser Permanente Group B Strep Advisory Committee (Ann Vannier, MD; Patricia Schmidt, MD; Judy Collier, RNC, MSN; Victor Wong, MD; and Evan Steinberg, MD); the Northern California Kaiser Permanente Neonatal Infection Study Group; and Girma Wolde-Tsadik, PhD, for their technical advice and consultation.
We also thank Kathie Heller, Cecilia Hoxie, Zorah Zari, Julie Stern, and Noemi Nagaoka who collected the data for this study.
- Received February 9, 1999.
- Accepted December 8, 1999.
Reprint requests to (P.B.) Department of Pediatrics, Kaiser Permanente Southern California, San Diego Medical Center, 4647 Zion Ave, San Diego, CA. E-mail:
- Franciosi RA,
- Knostman JD,
- Zimmerman RA
- Baker CJ, Edward MS. Group B streptococcal infections. In: Remington JS, Kleing JO, eds. Infectious Diseases of the Fetus and Newborn Infant. 4th ed. Philadelphia, PA: WB Saunders Co; 1995:980–1054
- Schuchat A
- Easmon CSF,
- Hastings MJG,
- Deeley J,
- Bloxham B,
- Rivers RPA,
- Marwood R
- Garland SM, Fliegner JR. Group B streptococcus (GBS) and neonatal infections: the case for intrapartum chemoprophylaxis. Aust NZ J Obstet Gynaecol. 1991;31:2:119–122
- Department of Health and Human Services
- Committee on Infectious Diseases and Committee on Fetus and Newborn, American Academy of Pediatrics
- Mohle-Boetani JC,
- Lieu TA,
- Ray GT,
- Escobar G
- D'Harlingue AE, Durand DJ. Recognition, stabilization and transport of the high risk newborn. In: Klaus MH, Fanaroff AA, eds. Care of the High Risk Neonate. 4th ed. Philadelphia, PA: WB Saunders Co; 1993:72
- SAS Institute, Inc. The SAS System for Windows, Release 6.12. Cary, NC; 1989
- Moses LM,
- Heath PT,
- Wilkinson AR,
- Jeffery HE,
- Isaacs D
- McIntosh EDG,
- Jeffrey HE
- Buck C,
- Gallati H,
- Pohlandt F,
- Bartmann P
- Copyright © 2000 American Academy of Pediatrics