1 table omitted
Continuing efforts are needed to improve surveillance for birth defects, which are the leading cause of infant mortality in the United States.1 Although state and local surveillance data indicate that approximately 3% of births are affected by any of 45 birth defects, no national estimates based on population-based birth defects surveillance have been available for specific types of birth defects other than neural tube defects (spina bifida and anencephaly). This report describes estimates of national prevalence and number of affected births in the United States each year during 1999-2001 for 18 selected major birth defects. The findings indicated that 10 of the 18 defects affected more than 1,000 infants each year in the United States. The conditions with the highest prevalence included orofacial clefts, which affect approximately 6,800 infants annually, and Down syndrome, which affects approximately 5,500 infants annually. Population-based national prevalence estimates of birth defects can help determine resource needs for basic and public health research and assist in planning for the health-care and educational needs of the U.S. population.
State birth defects surveillance systems traditionally have collected data on major structural birth defects and birth defects resulting from chromosomal abnormalities. Major structural birth defects are defined as conditions that (1) result from a malformation, deformation, or disruption in one or more parts of the body; (2) are present at birth; and (3) have a serious, adverse effect on health, development, or functional ability. Previous national estimates of birth defects prevalence were based on data from the Birth Defects Monitoring Program (BDMP); BDMP estimates used hospital discharge data to ascertain defects diagnosed at birth and were not population-based.2 The National Birth Defects Prevention Network (NBDPN) collects population-based surveillance data annually from 34 participating states for up to 45 major birth defects and publishes the prevalence of these defects by state.3 NBDPN data provide a means for comparing state-specific prevalence estimates for each defect by state, racial/ethnic population, type of surveillance, pregnancy outcomes included, and size of program; however, to date, national prevalence using these data has been estimated only for neural tube defects.4
To create the most reliable and valid estimates for national prevalence, this analysis included only states that had (1) active case-finding for all 18 included defects for 1999-2001 birth years and (2) data reported to NBDPN for all 3 years (1999-2001). NBDPN data from 11 states meeting these criteria (Alabama, Arkansas, California, Georgia, Hawaii, Iowa, Massachusetts, North Carolina, Oklahoma, Texas, and Utah) were used to calculate state-specific and average prevalence estimates (per 10,000 live births) and 95% confidence intervals for selected categories of major birth defects: eye defects, cardiovascular defects, orofacial defects, gastrointestinal defects, musculoskeletal defects, and chromosomal defects.
Pooled (i.e., average and unadjusted) prevalence estimates were calculated by summing the defect-specific counts across the 11 states and then dividing by the sum of the live births. The selected defects were chosen because they are recognizable at or shortly after birth and their ascertainment is less likely to be affected by regional differences in referral and clinical management practices than other types of defects. Neural tube defects were excluded because the national prevalence of these defects has been previously estimated using NBDPN data.4 The specific ascertainment methods and pregnancy outcomes included for each state are detailed in the program directory.3 All sites ascertain cases by abstracting medical records of potential cases identified from (1) discharge diagnosis indices from delivery and pediatric hospitals and (2) review of various hospital logs, including labor and delivery, neonatal intensive care unit, pathology, and surgery logs. States included in this report vary in the extent to which prenatally diagnosed birth defects are ascertained by the surveillance system; four states routinely visit prenatal diagnostic centers to ascertain cases, four states obtain some prenatal data from sources such as genetics laboratories, and three states do not include cases of <20 weeks' gestation that were ascertained from specialized sources for prenatal ascertainment.
Race/ethnicity-specific prevalence estimates were calculated for non-Hispanic white, non-Hispanic black, Hispanic, and “other” populations on the basis of pooled data from the 11 states. To estimate the national prevalence, the pooled race/ethnicity-specific estimates were applied to the racial/ethnic distribution of all live births in the United States during 1999-2001.5 Because of the strong association between maternal age and chromosomal defects, maternal age-specific prevalence estimates were calculated for Down syndrome, trisomy 13, and trisomy 18, and the national estimates for these defects were adjusted for maternal age (<35 years and ≥35 years) instead of race/ethnicity.6 Using the national prevalence estimates, the annual number of U.S. births affected by these selected major defects was calculated. By convention, infants with more than one defect were counted in each category in which they had a defect.
The average prevalence ranged from 0.82 per 10,000 live births for truncus arteriosus to 12.94 for Down syndrome. Most estimates clustered near the 11-state average estimate; however, variation was observed between states for each defect. For example, hypoplastic left heart syndrome ranged from 1.16 per 10,000 live births in the state with the lowest prevalence to 3.75 in the state with the highest prevalence; cleft palate ranged from 3.89 per 10,000 live births in the state with the lowest prevalence to 9.65 in the state with the highest prevalence. Variation might have occurred for several reasons, including (1) differences in surveillance ascertainment methods, (2) differences in maternal risk factors, such as smoking or nutrition during pregnancy, (3) differences in the racial/ethnic composition of the population for defects that vary by race/ethnicity, (4) differences between urban and rural settings in either environmental exposures or access to health care, and (5) variation attributed to chance. Adjusted for the racial/ethnic distribution (or maternal age for the three chromosomal defects) of live births in the United States during 1999-2001, the estimated national prevalence was highest for orofacial clefts combined (cleft lip with or without cleft palate and cleft palate only), followed by Down syndrome, rectal and large intestinal atresia/stenosis, individual heart defects (transposition of the great arteries, atrioventricular septal defect, and tetralogy of Fallot), and reduction defects of the upper limbs. For the defects selected for this analysis, the pooled estimates for the 11 states were similar to the national estimates, which were adjusted for race/ethnicity or maternal age.
MA Canfield, PhD, TA Ramadhani, PhD, Texas Dept of State Health Svcs. N Yuskiv, MPH, MJ Davidoff, MPH, JR Petrini, PhD, March of Dimes Birth Defects Foundation, White Plains, NY. CA Hobbs, MD, Arkansas Reproductive Health Monitoring System, Little Rock, Arkansas. RS Kirby, PhD, Dept of Maternal and Child Health, School of Public Health, Univ of Alabama at Birmingham, Birmingham, Alabama. PA Romitti, PhD, Iowa Registry for Congenital and Inherited Disorders, Iowa City, Iowa. JS Collins, PhD, Greenwood Genetic Center, Greenwood, South Carolina. O Devine, PhD, MA Honein, PhD, CT Mai, MPH, LD Edmonds, MSPH, A Correa, MD, National Center on Birth Defects and Developmental Disabilities, CDC.
This report documents substantial progress towards improved national prevalence estimates for 18 selected major birth defects using population-based active birth defects surveillance data. These estimates are important to (1) plan for health-care and education needs of the U.S. population, (2) identify increased occurrences of birth defects in specific geographic regions by making comparisons between local and national prevalence estimates, (3) serve as a reference point for assessment of state surveillance systems, (4) evaluate national public health interventions, such as folic acid fortification of cereal and grain products, (5) compare U.S. prevalence estimates with those of other countries, and (6) help determine the appropriate allocation of resources for basic and public health research. Although variation in the prevalence of defects between states has been documented previously,7 national estimates have been made only for the prevalence of neural tube defects.4 The population-based estimates in this report represent an improvement over the hospital-based estimates from BDMP2 because the source population for the denominator is defined on the basis of maternal residence at delivery. Typically, hospital-based estimates cannot determine the geographic region in which patients reside, nor can they indicate who would seek care for a major defect at a particular health-care facility.
These estimates help quantify the public health importance of these defects and can help improve the planning of services for affected children and their families (e.g., the need for specific clinical specialists or multi-specialty clinics). Children affected by certain birth defects could benefit from the availability of multi-specialty clinics to address the coordination of multiple needs and continuity of ongoing care in one setting. For example, a child with a cleft palate might require care from multiple specialists, including a plastic surgeon, an ear/nose/throat physician, a speech therapist, and an orthodontist. Because not all states have birth defects surveillance systems, the national prevalence estimates can be used by states to estimate prevalence in their states and to assess health services and special education programs. States can also use the national estimates to evaluate and improve their existing birth defects surveillance programs and ensure that affected children are referred to the appropriate services. As of November 2004, a total of 26 states with birth defects surveillance programs had systems in place to ensure that children identified by the program were referred to early intervention services.
NBDPN's efforts to improve the uniformity of case definitions and data collection and to allow a mechanism for pooling data make analyses such as these possible. The 11 states pooled for these estimates represent approximately 22% of all U.S. live births. The racial/ethnic distribution of the 11 states overrepresented Hispanic births in comparison with the total U.S. population (28% versus 20%); however, the pooled estimates were similar to the national estimates when adjusted for race/ethnicity. This similarity is attributable, in part, to the limited variation in the occurrence of most major structural birth defects by race/ethnicity.8
The findings in this report are subject to at least five limitations. First, no attempt has been made to determine the severity within each type of birth defect, which limits the utility of these data for health-care planning. For example, certain orofacial clefts, such as a small unilateral cleft lip, might only require a simple one-stage repair, whereas others might require repeated surgeries. Second, this analysis could not separate infants with isolated defects from those with more than one major defect or those with a recognized syndrome. Third, although the distribution of state-specific prevalence estimates was narrow, determining how much of the variation is attributable to ascertainment differences and how much is attributable to true differences in prevalence is not possible. The prevalence of some of the defects is influenced by the sources of prenatal diagnoses used by the surveillance systems, and this varies across the 11 states included in the analysis, with most of the systems relying primarily on hospital-based sources for case-finding. The previous national estimate for neural tube defects was stratified by the presence or absence of prenatal ascertainment because the estimates for these two defects are affected by the ascertainment of prenatally diagnosed cases. The estimated number of spina bifida and anencephaly cases each year in the United States was estimated at 1,640 and 1,380, respectively, using systems with prenatal ascertainment, compared with 1,340 and 840, respectively, using systems without prenatal ascertainment. This stratification was not performed for this analysis because none of the 18 selected defects are affected by the presence or absence of prenatal ascertainment to the same extent as neural tube defects. For this reason, the findings in this report represent a conservative estimate of the number of cases each year nationwide by using the data from all 11 states in the estimates for the 18 selected major defects; the estimates for several defects would be higher if based only on systems that include prenatal sources of ascertainment. Fourth, these national estimates represent minimum estimates for the impact of these defects, because even those surveillance systems with active case-finding do not achieve 100% ascertainment. Finally, some of the most common birth defects (e.g., muscular ventricular septal defects) are not included among the selected defects because identification of these defects depends on referral patterns and access to and use of diagnostic procedures such as high-resolution color echocardiography, which are likely to vary by region.9
Accurate national prevalence estimates of major birth defects are essential because birth defects are the leading cause of infant mortality and a major contributor to childhood morbidity.1,10 This report represents progress in estimating the number of children born affected by these 18 major birth defects each year in the United States; nonetheless, continued improvement can be achieved by (1) enhancing completeness of ascertainment, (2) increasing the consistency of methods among birth defects surveillance systems, and (3) expanding the number of major birth defects for which reliable and valid estimates of national prevalence can be made.
This report is based on data submitted to the Data Committee of the National Birth Defects Prevention Network by Alabama, Arkansas, California, CDC (Georgia), Hawaii, Iowa, Massachusetts, North Carolina, Oklahoma, Texas, and Utah. This report is also based on contributions by GM Shaw, California Birth Defects Monitoring Program, Berkeley, California; S Viner Brown, Rhode Island Birth Defects Surveillance Program; B McDowell, WT Budelier, Iowa Registry for Congenital and Inherited Disorders, Iowa City; RE Stevenson, MD, Greenwood Genetic Center, Greenwood, South Carolina.
Improved National Prevalence Estimates for 18 Selected Major Birth Defects—United States, 1999-2001. JAMA. 2006;295(6):618-620. doi:10.1001/jama.295.6.618