MADDSP uses a multiple source methodology for active population-based surveillance of five DDs including VI, HL, ASD, CP, and ID in five counties (Clayton, Cobb, DeKalb, Fulton, and Gwinnett) of metropolitan Atlanta, Georgia. Children are identified from record review at nine public school systems and selected private and public health sources that treat, diagnose, and/or serve children with DDs. Multiple records for a given child are compiled into one composite record which is then reviewed by a team of clinician reviewers to determine final case status. Additional methodological details have previously been published.²⁴ This project was approved by the Institutional Review Board at the Centers for Disease Control and Prevention (CDC).

We used MADDSP data from five surveillance years: 2000, 2002, 2004, 2006, 2008, hereafter referred to as ‘2000–2008’.

A case is defined as a child aged 8 years at any time during the surveillance year of interest, whose parent or legal guardian(s) resided in the five-county metropolitan Atlanta during the respective surveillance year, and who meets criteria for one or more of five DDs.

VI is defined as a measured visual acuity of 20/70 or worse in the better eye with correction. In the absence of a measured visual acuity, VI is defined based on a source record that includes a functional description, by a qualified physician or vision professional, of visual acuity of 20/70 or worse (e.g., light perception only) or a statement by a qualified physician or vision professional that the child had low vision or blindness. VI severity was defined as: a) ‘low vision’ with a visual acuity 20/70 to better than 20/200; b) ‘legal blindness’, with a visual acuity of 20/200 or worse, and c) cerebral visual impairment.

HL is defined as a measured, bilateral, pure-tone hearing loss at frequencies of 500, 1,000, and 2000 Hz averaging ≥40 dB, unaided, in the better ear. Severity of HL was defined based on the measurement in the better ear: moderate (hearing loss of 40–64 dB), severe (hearing loss of 65–84 dB), and profound (hearing loss of ≥85 dB).²⁵

ASD is defined by the documentation of behaviors (as described on a comprehensive evaluation by a qualified professional) that are consistent with the diagnostic criteria listed in the Diagnostic and Statistical Manual-IV, Text Revision (DSM-IV-TR) for any of the following conditions: Autistic Disorder, Pervasive Developmental Disorder-Not Otherwise Specified (PDD-NOS, including Atypical Autism), or Asperger Disorder.^19,26

ID is defined as an intelligence quotient (IQ) of ≤70 on the most recently administered psychometric test. ID severity is defined based on the International Classification of Diseases, Ninth Edition, Clinical Modification²⁷: mild (IQ = 50–70), moderate (IQ = 35–49), severe (IQ = 20–34), and profound (IQ <20).

CP is defined by documentation of a CP diagnosis or description of physical findings consistent with CP noted in clinical evaluation(s) at age 2 or older by a qualified professional.^24,28 Children with CP acquired after birth (post-neonatal CP) are included. CP subtypes are categorized as spastic (monoplegia, hemiplegia, diplegia, quadriplegia, or triplegia), non-spastic (dyskinetic, ataxic, hypotonic, dyskinetic-ataxic) and mixed CP or CP NOS (spastic-ataxic, spastic-dyskinetic, spastic athetoid, mixed CP or CP NOS).

Information on epilepsy and Down syndrome were abstracted from MADDSP records, if documented in the record by a medical doctor, a psychologist, or an educator.

Information on congenital malformations, syndromes, and chromosomal abnormalities was obtained by linkage with the Metropolitan Atlanta Congenital Defects Program (MACDP),²⁹ a population-based birth defects surveillance program of children born to mothers who resided within the five-county region of metropolitan Atlanta at the time of their birth. Diagnoses of congenital malformations were confirmed and classified by MACDP clinicians. MACDP is described elsewhere.³⁰

Information on birth, maternal and paternal characteristics were obtained from linkage with birth certificate data of children born in metropolitan Atlanta. Birth certificate data were not obtainable for children born outside of the 5 county metropolitan Atlanta who migrated into metropolitan Atlanta between birth and age 8 years (35%). We conducted a sub-analysis on selected characteristics of children with VI or HL by whether they were non-migrants (born and residing in metropolitan Atlanta at age 8) or in-migrants (born outside of metropolitan Atlanta and moved in by age 8) (data not shown). We found no difference between non-migrant and in-migrant children with HL by sex, severity of HL or presence of co-occurring DD. Differences were found between non-migrant and in-migrant children with VI by sex, race and ethnicity, and presence of co-occurring DDs, but not by severity of VI, which we believe may be the strongest indicator for selective migration. Child’s race/ethnicity was obtained primarily from abstracted MADDSP records and supplemented with information from the child’s birth certificate if not available from the abstracted data.

The prevalence for VI and HL was estimated per 1000 8-year-olds residing in the metropolitan Atlanta area during surveillance years 2000–2008. Denominator data were obtained from the CDC National Center for Health Statistics postcensal 2009 vintage population estimates for 2000–2008 surveillance years yielding a total of 230,973 8-year-olds in metropolitan Atlanta.³¹ Average annual period prevalence estimates, henceforth referred to as ‘prevalence’, were calculated for 2000–2008 overall by summation of the numerator data divided by the summation of denominator data across all surveillance years. The 95% confidence intervals (CIs) were calculated using the Poisson approximation to the binomial distribution. Poisson regression was used to test for linear trends in the VI and HL prevalence estimates from 2000 to 2008. Chi square or Fisher exact test of significance was used to examine differences in prevalence by sex and race/ethnicity as well as selected birth and parental characteristics, severity, co-occurring DDs between children with VI or HL with and without ASD.

Children were considered to have a previously documented ASD classification if they received a diagnosis of Autistic Disorder, PDD-NOS, Asperger Disorder, or ASD that was documented in an abstracted evaluation or by an ICD-9 billing code at any time from birth through the end of the year when they reached age 8 years, or if they received special education services under an autism eligibility. Ages at earliest known ASD diagnosis and earliest evaluation were calculated by subtracting the earliest date or age of known diagnosis or evaluation from the child’s date of birth. An earlier historical ASD diagnosis was taken if specified in a comprehensive evaluation. Between group differences in the median age of ASD evaluation were examined using the Mann–Whitney test. All statistical analyses were conducted using SAS (version 9.2; SAS Institute, Cary, NC).

The prevalence of VI during 2000–2008 was estimated as 1.2 per 1000 8-year-olds (95% CI = 1.1–1.4). Prevalence estimates across 2000–2008 were stable (p = 0.12) (data not shown). VI prevalence was higher among Hispanic children compared with White non-Hispanic children (p = 0.05), but did not differ significantly between boys and girls or between White and Black non-Hispanic children (Table 2). Sixty three percent of children with VI had at least another DD (including ASD, CP, ID, and/or HL). The frequency of ASD among children with VI was 7.2% (95% CI: 4.6–11.2); comprised of children with VI and ASD only (1.4%), and those with VI, ASD and other DDs (5.8%). These were lower than the co-occurrence of VI and ID (52.5%) and VI and CP (41.0%), but the overall frequency was slightly higher than VI and HL overall (5.8%) (Table 3). Among other conditions examined, 19.1% of children with VI had congenital malformations, Down syndrome, or a chromosomal abnormality, and 29.1% had epilepsy.

The prevalence of VI and co-occurring ASD was estimated as 0.09 per 1000 8-year-olds (95% CI = 0.06–0.13) and there was no significant linear trend in the prevalence estimates of this co-occurrence over the study period (p = 0.44). No significant differences were noted in the prevalence of VI with ASD by sex and race/ethnicity (Table 2). Children with VI and ASD were significantly more likely to be low birth weight (p = 0.01) and/or a preterm birth (p = 0.02) (Table 4) and had a significantly higher frequency of co-occurring ID (p = 0.04), compared to children with VI without ASD (Table 3). Three out of 20 children with VI and ASD had a congenital malformation, syndrome, or chromosomal abnormality.

Approximately 70.0% of children with VI and ASD had a previously documented ASD classification compared to 81.5% of children with ASD without VI. There was no difference in the median age of earliest evaluation between children with VI and ASD compared with children with ASD without VI (Table 5). Yet, the median age of earliest known ASD diagnosis in children with VI was significantly later (79 months) compared with children with ASD without VI (56 months, p = 0.02).

The prevalence of HL during 2000–2008 was estimated as 1.3 per 1000 8-year-olds (95% CI = 1.2–1.5). The annual HL prevalence estimates during the surveillance period did not vary significantly (p = 0.28) (data not shown). Overall, HL prevalence did not differ significantly by sex or race/ethnicity. Seventy seven percent of children with HL had co-occurring ASD, CP, ID or VI. The frequency of ASD among children with HL was 5.8% (95% CI = 3.7%–9.3%); including those with HL and ASD only (1.6%) and HL, ASD and other DDs (4.2%). The frequency of other co-occurring DDs among children with HL was highest for ID (23.7%) followed by CP (10.7%) and VI (5.2%). One half of children with HL had moderate loss, 23.7% were in the severe range and 26.9% had profound HL (Table 3). Among the other medical conditions examined, 12.3% of children with HL had a major congenital malformations syndrome, or chromosomal abnormality, 7.5% had epilepsy, and 2.3% had Down syndrome.

The prevalence of HL co-occurring with ASD was estimated as 0.08 per 1000 8-year-olds (95% CI = 0.05–0.12), with no significant linear trend in the prevalence over the surveillance period (p = 0.60). The prevalence of HL with co-occurring ASD was significantly higher among boys compared to girls; however, this was based on small samples (Table 2). Children with HL and ASD had a significantly different sex distribution compared to children with HL without ASD (p = 0.01) (Table 3). The majority of children with HL and ASD had moderate HL (61.1%). A greater proportion of children with HL and co-occurring ASD had co-occurring ID and co-occurring CP compared to children with HL without ASD. No significant differences were found between children with HL with and without ASD according to the other birth and parental characteristics examined (Table 4).

The proportion of children with a previously documented ASD classification was similar for children with ASD and HL and those with ASD without HL (77.8% and 81.4%, respectively). Comparison of median ages of first evaluation showed that children with ASD and HL appeared to have been evaluated significantly earlier than those with ASD without HL (40 months vs. 50 months) (p = 0.01) (Table 5). Nevertheless, the median age of earliest known ASD diagnosis among children with HL (57 months) was similar to the age of ASD diagnosis in children with ASD without HL (56 months).