PEDIATRICS Vol. 119 No. 2 February 2007, pp. e384-e390 (doi:10.1542/peds.2006-0610)
ARTICLE |
Issues in Estimating the Prevalence of Fetal Alcohol Syndrome: Examination of 2 Counties in New York State
Congenital Malformations Registry, New York State Department of Health, Troy, New York
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
OBJECTIVE. Two demographically similar counties included in the New York Fetal Alcohol Syndrome Surveillance Network had very different prevalence rates. This study examined the components of the surveillance in an attempt to discover the reasons for this discrepancy.
METHODS. Erie County and Monroe County were the 2 most populous counties included in the New York Fetal Alcohol Syndrome Surveillance Network. Erie County includes Buffalo, the second largest city in New York State, and Monroe County includes Rochester, the third largest city. Multiple sources of ascertainment included birth defect surveillance systems, genetic clinics, and early intervention programs. The case definition was based on the Institute of Medicine criteria of an abnormality in each of the following 3 areas: facial features, central nervous system, and growth.
RESULTS. Children born in Erie County or Monroe County between 1995 and 1999 were included. The fetal alcohol syndrome prevalence rates in these 2 counties were 0.90 cases per 1000 births and 0.21 cases per 1000 births, respectively. The 2 counties were demographically similar and had similar rates of binge drinking among women of childbearing age. There was less participation in the surveillance system by sources in Monroe County. Erie County had a very active clinician with a specialized fetal alcohol syndrome clinic.
CONCLUSIONS. The participation of clinicians in one county, especially one with expertise in fetal alcohol syndrome, was the most likely explanation for the differences in prevalence rates between the counties.
Key Words: fetal alcohol syndrome • prevalence • birth defects • statistical estimation • data collections • facial features • diagnostic criteria
Abbreviations: FAS—fetal alcohol syndrome • FASSNet—Fetal Alcohol Syndrome Surveillance Network • IOM—Institute of Medicine • CNS—central nervous system
Wide variations in fetal alcohol syndrome (FAS) prevalence rates have been found in various studies. The results range from 
0.2 cases per 1000 newborns in passive systems to 1.0 case per 1000 births in some clinic-based studies1,2 and 3.0 cases per 1000 children in a study of first-grade children in one county in Washington State.3 Even higher prevalence rates have been found in studies of specific communities. A prevalence rate of 65.2 to 74.5 cases per 1000 was found for school-aged children in a wine-growing community in South Africa.4
There are many factors that may account for this variation. Disparities in FAS prevalence rates based on demographic features such as socioeconomic status and race/ethnicity are well known.2,5 Abel6 found that prevalence rates were generally 10 times higher for black individuals. Variation could result from different levels of alcohol intake or different drinking patterns, but differences in ascertainment methods or age groups included could also be contributing factors.2,5,7 FAS is best diagnosed in children 2 to 11 years of age; facial features may be difficult to distinguish in some newborns and become easier to recognize over time.4,8–12 Newborns may not be well examined for FAS.11,13 Therefore, prevalence rates may seem lower if only newborns or younger children are included.3 Other reasons for disparities include differences in case definitions, the expertise and willingness of clinicians to make the diagnosis, and diligence in searching for cases.8,9,13
The Institute of Medicine (IOM) report compared surveillance methods for FAS, pointing out that active surveillance systems find more cases and the diagnoses are "more likely to be valid and reliable."14 The report pointed out that the difficulties with active surveillance include the expense and, because of population selection, resulting prevalence rates may not be generalizable.14 Alaska performed multiple-source FAS surveillance and found that 65% of the cases were identified through active screening and referral programs to specialty diagnostic clinics, "[illustrating] the magnitude of under-reporting by all of the passive reporting data systems."15 Through active screening, Clarren et al3 found a prevalence rate of 3.1 cases per 1000 children, one of the higher prevalence rates found for a general population in the United States. May and Gossage2 demonstrated similarly that "active case ascertainment generally yields the highest number of cases and rates of FAS for a particular population."
The basic criteria for diagnosis of FAS were originally laid out by Jones and Smith16 and, with some modification, are still in use.14 In addition to maternal alcohol use, they include criteria in 3 areas, namely, central nervous system (CNS) abnormalities, such as microcephaly or developmental delays; growth retardation (prenatal and postnatal); and specific abnormal facial features, such as short palpebral fissures, flat philtrum, or thin upper lip. The facial features are the most specific factors in the diagnosis that differentiate FAS from other syndromes or conditions that result in growth retardation. However, each feature "represents a minor anomaly or variant of normal,"10 and features may be overlooked unless clinicians search specifically for them.
In 1997, the Centers for Disease Control and Prevention provided funding for 5 states (Alaska, Arizona, Colorado, New York, and Wisconsin) to develop the Fetal Alcohol Syndrome Surveillance Network (FASSNet). This project was described in detail elsewhere.17 Results from 4 of the states were published, and some variation in prevalence rates was demonstrated.18 In examining regional data from the New York FASSNet system, differences were noted in the prevalence of FAS in 2 of the largest counties in the surveillance region. In this article, possible reasons for the differences, including demographic factors, alcohol use, and sources for case ascertainment, are examined.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
FASSNet
The FASSNet project objectives were to enhance or to develop a multiple-source surveillance system, to generate population-based surveillance data, to establish relationships with diagnostic and service programs, to evaluate the completeness of the surveillance system method, and to implement provider training. Four of the participating sites (Alaska, Arizona, Colorado, and New York) used the same general methods, including a common case definition, a common data collection form, and multiple sources for identification of cases. These sources included hospitals, birth defect registries, genetic clinics, developmental clinics, and early intervention programs. No direct examinations of the children were performed for the surveillance. All information was abstracted from medical records with the standardized form.
The case definition for FASSNet surveillance was based on the IOM criteria,14 which were adapted for the project by a FASSNet committee of experts, including a dysmorphologist, a geneticist, and a developmental psychologist. The IOM criteria provide no specific parameters for each of the criteria, although there has been a recent effort to do so.19 The criteria were quantified by the FASSNet committee to be uniform and specific for use across sites. The FASSNet surveillance case definition criteria are described in Table 1.
|
To meet the criteria for growth retardation (ie, weight, height, or weight for height), a child's growth measures (at any age) needed to be
10th percentile for age. Standard growth curves for both intrauterine and postnatal growth were selected for uniform evaluation of growth.17 To meet the criteria for CNS abnormality, either structural (ie, head circumference) or functional (ie, mental retardation, developmental delay, or attention-deficit disorder) criteria were used. A standard head circumference growth curve was chosen for evaluation of both birth and postnatal head circumference. A child whose head circumference was 10th percentile for age met the CNS structural criteria for FAS. To measure functional delays (ie, developmental delay, mental retardation, and other intellectual deficits), a list of acceptable standardized tests was developed in consultation with developmental psychologists. Tests scores of 
1 SD below the mean for a child of similar age were considered to meet the criteria for CNS abnormality. Alternatively, a diagnosis of developmental delay, mental retardation, or attention-deficit disorder by a qualified clinician met the CNS criteria. More than 95% of the New York FASSNet case subjects who met the CNS criteria did so on the basis of a head circumference of 10th percentile. To meet the facial dysmorphic criteria, 3 facial anomalies (ie, short palpebral fissures, abnormal philtrum, and thin upper lip) were chosen as case definition criteria, on the basis of the characteristics that have been shown to discriminate best between children who have FAS and those who do not.20 Case status was determined electronically by using a computer algorithm to evaluate all of the data abstracted for the child. A case was classified as "definite" if the criteria in each of the 3 categories were met. A "probable" case classification met the criteria for facial features and either growth criteria or CNS criteria.
New York FASSNet
The surveillance area for the New York FASSNet consisted of 9 counties in western New York, which included both urban and rural areas. The 2 largest and most urban counties are Erie County and Monroe County. Erie County includes Buffalo, the second largest city in New York State, and Monroe County includes Rochester, the third largest city. Case ascertainment began in 1998 and was retrospective to birth year 1995. Birth certificates were used to define maternal residence at birth and other demographic variables. Birth years 1995 through 1999 were included, and definite and probable cases were combined and used to calculate the FASSNet prevalence. Maternal alcohol use was not used as a criterion, although >90% of the children confirmed as having FAS had a notation of maternal alcohol use in their medical records.
In the New York FASSNet, cases were ascertained from multiple sources, including genetic clinics, early intervention programs, hospital discharge data, birth defect surveillance systems, developmental clinics, vital records, and other clinics, including a special clinic for mothers with substance abuse problems and their children. Some sources were directly accessible by FASSNet staff members; these included hospital discharge data, birth certificate data, and data from 2 birth defect surveillance systems (the statewide Congenital Malformations Registry, which relies on hospital reporting, and a regional surveillance system in which staff members review records of children identified with birth defects). Other sources, such as genetic or developmental clinics and early intervention programs, required that potential cases be identified to FASSNet staff members. Although these sources were visited regularly by FASSNet staff members, the clinics varied in how they identified cases. Some had computerized records, whereas others kept lists as cases were identified.
Frequently, all of the information needed to meet the case definition was not available from the ascertainment source; therefore, a child would be evaluated with other sources, to obtain more information for confirmation. For each child, the FASSNet database tracked the initial source(s) of ascertainment and any follow-up sources. Therefore, it can be determined which sources identified children and which provided the specific information for the case definition items, such as facial features.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
FASSNet Prevalence Rates According to County
All prevalence rates presented were calculated per 1000 live births and were based on maternal residence at birth in Erie County or Monroe County. The prevalence rates in Erie County and Monroe County were 0.90 and 0.21 cases per 1000 live births, respectively. FAS prevalence rates according to race/ethnicity for 1995 to 1999 are presented in Table 2. The prevalence rate among white individuals in Monroe County was approximately one half of that in Erie County; the prevalence rate among black individuals in Monroe County was slightly more than 10% of that in Erie County. In Erie County, the prevalence rate among black individuals was 10 times that among white individuals; in Monroe County, the prevalence rate among black individuals was twice that among white individuals.
|
Demographic Factors
For the study period, Erie County had
18% more live births than Monroe County (58435 vs 47901). Maternal and infant characteristics are shown in Table 3. In general, the characteristics were very similar in the 2 counties; however, Monroe County had a slightly higher percentage of minority births.
|
The New York Behavioral Risk Factor Surveillance System survey provided county-level data on binge drinking (defined as
5 drinks at 1 occasion) for the years 1995, 1997, and 1999. Women in Monroe County reported somewhat higher levels of binge drinking than did women in Erie County (16.6% vs 12.3%). Rates for both counties were higher than reported rates of binge drinking for the state as a whole for the same time period, which ranged from 6.3% to 8.2%. The small numbers available did not allow more-detailed analysis.
Ascertainment Sources for New York FASSNet
To gain a better understanding of how children were identified and how the criteria for the case definition were met, the ascertainment sources for each county were examined. This evaluation included not only children confirmed as having FAS but also all cases ascertained by the FASSNet system, both confirmed and unconfirmed. The comparison of initial ascertainment sources according to county is presented in Table 4. The totals for the sources are more than the total number of children in each category because a case might have been ascertained independently by >1 source.
|
All Identified Children
Overall, as illustrated in Table 4, Erie County had more than twice as many initial ascertainment abstractions in the system as Monroe County (542 and 222, respectively) and twice as many children (420 and 208, respectively). Sources directly accessible by FASSNet staff members accounted for 89% of Monroe County ascertainments, compared with 66% of Erie County referrals. In Erie County, 75 children (18% of total children) were referred from genetic clinics, compared with 5 children (2.5% of total children) in Monroe County. The percentages of children identified by the system who were black were similar (67.9% in Erie County and 63.0% in Monroe County).
Children Confirmed as Having FAS
Comparisons of children confirmed as having FAS were more difficult, because the numbers in Monroe County were small. As shown in Table 4, 13% of the children from Erie County who were identified through FASSNet were confirmed as meeting the FAS case definition (53 of 420 children), whereas 5% from Monroe County were confirmed (10 of 208 children). In Erie County, many of the children confirmed as having FAS (36 of 53 children; 68%) were identified initially through genetic clinics, compared with only 20% (2 of 10 children) in Monroe County identified through that source. In both counties, approximately one half of the children referred from genetic clinics met the FAS case definition (in Erie County: 36 of 75 children; 48%; in Monroe County: 2 of 5 children; 40%). Children who were black accounted for 68% of confirmed cases in Erie County and 40% of confirmed cases in Monroe County.
Identification of Facial Features
Because the description of facial features, which depends on active clinical recognition, is the critical element in the surveillance case definition, the sources providing information on abnormal facial features used for case definition, regardless of FAS case status, were examined. More children in Erie County had documentation on facial features in their records, compared with Monroe County; genetic clinics were a major source of information (Table 5). Facial feature information found through hospital discharge sources originated frequently from inpatient genetic consultations.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
In the New York FASSNet, 2 neighboring urban counties, which are demographically similar, were found to have very different FAS prevalence rates. Given the slightly higher rates of minority births and binge drinking in Monroe County, it might be expected that the FAS prevalence rate would be similar to, if not higher than, that in Erie County. Instead, the prevalence rate was less than one third that found in Erie County. This discordance illustrates some of the problems and challenges of FAS surveillance.
Although the county-level Behavioral Risk Factor Surveillance System data on binge drinking were not specific to pregnant women, the data do demonstrate that reported binge drinking among women of childbearing age is somewhat higher in Monroe County. Heavy alcohol use before pregnancy is highly predictive of continued use in pregnancy.21 Even if a woman reduces her drinking at pregnancy recognition, the fetus can be affected very early in pregnancy, before the woman realizes she is pregnant.22,23 It is unlikely that the time of pregnancy recognition varied greatly between Monroe County and Erie County.
Overall, in Monroe County, fewer children were identified, a lower percentage of the children were confirmed as having FAS, and a lower percentage of black children were confirmed as having FAS. There was less "active" participation by Monroe County sources, especially by the clinical sources. Most cases were ascertained from sources accessed directly by FASSNet staff members, and few were ascertained from diagnostic clinics. Diagnostic clinics have been an important source of cases in other surveillance systems.2,15 In Erie County, genetic clinics were an important source of case ascertainment and information on facial features; in Monroe County, the numbers ascertained and confirmed through genetic clinics were lower. The lower percentage confirmed in Monroe County, compared with Erie County, may indicate difficulty in obtaining the needed documentation of facial features. The geneticists in Monroe County have expressed a willingness to make the diagnosis, but clinicians in that county may be more reluctant to refer patients to geneticists. Geneticists, through both clinic records and inpatient consultations, provided most of the facial feature information. The limited information on facial features made it more difficult for the FASSNet system to identify or to confirm potential cases in Monroe County. In addition, most of the FASSNet sources are biased toward the newborn period and younger children, but older children could be identified through genetic clinics. Thirty percent of New York FASSNet cases were diagnosed after 2 years of age, and the majority of them were ascertained through genetic clinics.
For a diagnosis of FAS, it is important that clinicians think about the possibility of FAS and how to identify the facial features. For surveillance purposes, the facial features need to be recorded in the medical record. FAS is approximately as common as Down syndrome (1.3 cases per 1000).24 However, a diagnosis of FAS is more complicated. There is no diagnostic test specific for FAS. A child may need to be monitored after birth, because FAS facial features may not be present at birth and growth and CNS deficiencies may not be detected until the child is older.4,8–12
It is important that clinicians understand that there are reasons to make a diagnosis other than improving FAS surveillance. There can be direct benefits to the child and the family. Although some clinicians do not want to "label" the child, they overlook an opportunity to intervene with the family. In addition, studies have shown that children who are diagnosed earlier fare better, because they may receive specific interventions.25 The study of first-grade children in Washington found that, of the children who screened positive for FAS, only 1 of 7 had been diagnosed previously and only one half were receiving special services.3 There may be many children with undiagnosed FAS in communities. Not all children with FAS are mentally retarded, but they have serious learning problems that may not be recognized; therefore, the children may not receive appropriate interventions and may experience failure in school. There needs to be an awareness that, although these children have problems, diagnosis allows for the use of programs that can help the children be more successful in life.26,27
Much of the success of FASSNet in Erie County seems to be attributed to the involvement of a knowledgeable clinician, a dysmorphologist. He has expertise in FAS and was involved actively with FASSNet, regularly reporting cases to the FASSNet staff. He documented facial features (both normal and abnormal) routinely in the record. He interacted with several of the FASSNet sources in Erie County and worked to educate health care providers in Erie County to be aware of the diagnosis, as well as the procedures for making the diagnosis.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The prevalence of FAS is highly variable among populations. It is frequently difficult to know what accounts for these differences. In this article, the components of a surveillance project were examined in detail, in an attempt to discover the reasons behind the large differences in FAS prevalence between 2 neighboring and demographically similar counties. The willingness and cooperation of clinicians in one county, especially one with expertise in FAS, represent the most likely explanation.
|
ACKNOWLEDGMENTS |
|---|
This project was supported by grants from the Centers for Disease Control and Prevention, Division of Birth Defects and Developmental Disabilities (cooperative agreements U50/CCU2145560 and U50/CCU223184).
We are very grateful for the valuable contributions of Christina Westfield, Sandi Gangell, and Dr Luther Robinson to the FASSNet project. We thank Pat Steen and April Austin for their editorial comments.
|
FOOTNOTES |
|---|
Accepted Aug 11, 2006.
Address correspondence to Charlotte M. Druschel, MD, MPH, Congenital Malformations Registry, New York State Department of Health, 547 River St, Room 200, Troy, NY 12180-2216. E-mail: cmd05{at}health.state.ny.us
Financial Disclosure: Dr Druschel is an employee of the New York State Department of Health. Ms Fox is an employee of Health Research Inc.
This study was presented in part at the annual Maternal and Child Health Epidemiology Conference; December 12, 2002; Clearwater Beach, FL.
This analysis was based on surveillance data collected by the Congenital Malformation Registry as part of the FASSNet project funded by the Centers for Disease Control and Prevention. The responsibility for the use and interpretation of these data is entirely that of the authors. The Centers for Disease Control and Prevention had no role in data analysis and interpretation or in the preparation, review, or approval of this manuscript.
Dr Druschel and Ms Fox had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
- Centers for Disease Control and Prevention. Fetal alcohol syndrome: United States, 1979–1992. MMWRs Morb Mortal Wkly Rep. 1993;42 :339 –341
- May PA, Gossage JP. Estimating the prevalence of fetal alcohol syndrome: a summary. Alcohol Res Health. 2001;25 :159 –167[Web of Science][Medline]
- Clarren SK, Randels SP, Sanderson M, Fineman RM. Screening for fetal alcohol syndrome in primary schools: a feasibility study. Teratology. 2001;63 :3 –10[CrossRef][Web of Science][Medline]
- Viljoen DL, Gossage JP, Brooke L, et al. Fetal alcohol syndrome epidemiology in a South African community: a second study of a very high prevalence area. J Stud Alcohol. 2005;66 :593 –604[Web of Science][Medline]
- Abel EL, Sokol RJ. A revised conservative estimate of the incidence of FAS and its economic impact. Alcohol Clin Exp Res. 1991;15 :514 –524[CrossRef][Web of Science][Medline]
- Abel EL. An update on incidence of FAS: FAS is not an equal opportunity birth defect. Neurotoxicol Teratol. 1995;17 :437 –443[CrossRef][Web of Science][Medline]
- Abel EL, Hannigan JH. Maternal risk factors in fetal alcohol syndrome: provocative and permissive influences. Neurotoxicol Teratol. 1995;17 :445 –462[CrossRef][Web of Science][Medline]
- Jones KL. Early recognition of prenatal alcohol effects: a pediatrician's responsibility. J Pediatr. 1999;135 :405 –406[CrossRef][Web of Science][Medline]
- Jones KL. From recognition to responsibility: Josef Warkany, David Smith, and the fetal alcohol syndrome in the 21st century. Birth Defects Res A Clin Mol Teratol. 2003;67 :13 –20[CrossRef][Web of Science][Medline]
- Aase JM. Clinical recognition of FAS: difficulties of detection and diagnosis. Alcohol Health Res World. 1994;18 :5 –9
- Little BB, Snell LM, Rosenfeld CR, Gilstrap LC III, Gant NF. Failure to recognize fetal alcohol syndrome in newborn infants.
Am J Dis Child. 1990;144
:1142
–1146
[Abstract/Free Full Text] - Stoler JM, Holmes LB. Recognition of facial features of fetal alcohol syndrome in the newborn. Am J Med Genet C Semin Med Genet. 2004;127 :21 –27
- Stoler JM, Holmes LB. Under-recognition of prenatal alcohol effects in infants of known alcohol-abusing women. J Pediatr. 1999;135 :430 –436[CrossRef][Web of Science][Medline]
- Stratton KR, Howe CJ, Battaglia FC, eds. Fetal Alcohol Syndrome: Diagnosis, Epidemiology, Prevention, and Treatment Washington, DC: National Academy Press; 1996
- Egeland GM, Perham-Hester KA, Gessner BD, Ingle D, Berner JE, Middaugh JP. Fetal alcohol syndrome in Alaska, 1977 through 1992: an administrative prevalence derived from multiple data sources.
Am J Public Health. 1998;88
:781
–786
[Abstract/Free Full Text] - Jones KL, Smith DW. Recognition of the fetal alcohol syndrome in early infancy. Lancet. 1973;2(7836) :999 –1001
- Hymbaugh K, Miller LA, Druschel CM, Podvin DW, Meaney FJ, Boyle CA. A multiple source methodology for the surveillance of fetal alcohol syndrome: the Fetal Alcohol Syndrome Surveillance Network (FASSNet). Teratology. 2002;66(suppl 1) :S41 –S49[CrossRef]
- Miller L, Tolliver R, Druschel C, et al. Fetal alcohol syndrome: Alaska, Arizona, Colorado, and New York, 1995–1997. MMWR Morb Mortal Wkly Rep. 2002;51 :433 –435[Medline]
- Hoyme HE, May PA, Kalberg WO, et al. A practical clinical approach to diagnosis of fetal alcohol spectrum disorders: clarification of the 1996 Institute of Medicine criteria.
Pediatrics. 2005;115
:39
–47
[Abstract/Free Full Text] - Astley SJ, Clarren SK. A fetal alcohol syndrome screening tool. Alcohol Clin Exp Res. 1995;19 :1565 –1571[CrossRef][Web of Science][Medline]
- Centers for Disease Control and Prevention. Alcohol use among women of childbearing age: United States 1991–1999. MMWR Morb Mortal Wkly Rep. 2002;51 :273 –276[Medline]
- Konovalov HV, Kovetsky NS, Bobryshe YV, Ashwell KW. Disorders of brain development in the progeny of mothers who used alcohol during pregnancy. Early Hum Dev. 1997;48 :153 –166[CrossRef][Web of Science][Medline]
- Centers for Disease Control and Prevention. Alcohol consumption among women who are pregnant or who might become pregnant: United States 2002. MMWR Morb Mortal Wkly Rep. 2004;53 :1178 –1181[Medline]
- Centers for Disease Control and Prevention. Improved national prevalence estimates for 18 selected major birth defects: United States, 1999–2001. MMWR Morb Mortal Wkly Rep. 2005;54 :1301 –1305
- Streissguth AP, Barr HM, Kogan J, Bookstein FL. Understanding the Occurrence of Secondary Disabilities in Clients With Fetal Alcohol Syndrome and Fetal Alcohol Effects Seattle, WA: University of Washington Press; 1996
- Streissguth AP, Barr HM, Kogan J, Bookstein FL. Primary and secondary disabilities in fetal alcohol syndrome. In: Streissguth AP, Kanter J, eds. The Challenges of Fetal Alcohol Syndrome Seattle, WA: University of Washington; 1999:25 –39
- Mattson SN, Schoenfeld AM, Riley EP. Teratogenic effects of alcohol on brain and behavior. Alcohol Res Health. 2001;25 :185 –191[Web of Science][Medline]
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
