Background. Accurately estimating the prevalence of intellectual and developmental disabilities (ID/DD) in the United States remains a challenge, with reported rates varying widely (3%-17%) based on the definitions and data sources used. Administrative data, especially from Medicare and Medicaid claims, due to their broad population coverage and longitudinal structure, can be a valuable resource for identifying and monitoring the health of individuals with many conditions, including ID/DD. However, current tools like the Centers for Medicare & Medicaid Services’ Chronic Conditions Data Warehouse (CCW) condition flags for intellectual disabilities, autism spectrum disorders, and other developmental disabilities exclude many potentially relevant diagnoses, leading to substantial variation across federal and nonfederal research in International Classification of Diseases (ICD) diagnosis codes included in claims-based ID/DD definitions.
Objectives. In this brief, we present findings from a descriptive analysis of Medicare and Medicaid administrative data and discussions with a multidisciplinary technical expert panel (TEP) toward informing a future update to the CCW condition flags for ID/DD.
Methods. We first constructed a baseline ID/DD definition comprising six conditions that built on ICD-10 codes used in the 2024 CCW condition flags, supplemented by TEP input. TEP members were convened at a two-hour virtual meeting in November 2024 and asynchronously before and after the meeting. With the aid of prior research and TEP input, we also identified eight additional conditions and associated ICD-10 codes for consideration to expand the baseline definition. We then analyzed Medicare and Medicaid claims and encounter data from 2021–2022 for beneficiaries aged 2 years and older, continuously enrolled in either program, and residing in the United States (U.S.) to estimate (1) the prevalence of ID/DD in administrative data using the baseline definition, (2) the prevalence of the eight other conditions, and (3) the co-occurrence of the additional eight conditions with the baseline ID/DD definition. We also convened a multidisciplinary TEP, including clinicians. TEP members also offered important operational considerations for implementing such an expansion.
Results. Among the 125 million eligible Medicare and Medicaid beneficiaries aged 2 years and older, 2.9% (or approximately 3.6 million people) had at least one ID/DD diagnosis based on ICD-10 codes in the baseline definition. Fewer beneficiaries had an ID-only diagnosis (0.8%) than a DD-only diagnosis (1.7%), and a smaller subset (0.4%) of beneficiaries had a diagnosis of both ID and DD. The share of beneficiaries with an ID/DD diagnosis declined with age, from 6.4% among the population aged 2-18 years to 0.5% among the population aged 65 years and older. This decline may be due largely to a reduction in DD diagnoses, likely influenced by use of care, service eligibility, and diagnostic practices across the lifespan, as well as shorter life expectancy for some people with DD. Among the eight additional conditions not included in the baseline definition, attention-deficit/hyperactivity disorder (ADHD) had the highest overall prevalence across all age groups at 4.1%, particularly among children and adolescents (9.4%). Other nervous system disorders followed at 3.8% overall, which were most common in older adults (8.1%). Speech and language disorders had the third highest overall prevalence (2.1%), with a concentration among the 2-18 age group (7.5%). The remaining five conditions—learning disabilities, spina bifida, cerebral palsy, other chromosomal abnormalities, and other congenital anomalies—had overall prevalence rates below 0.5%. Co-occurrence analysis of the eight additional conditions with the baseline ID/DD definition revealed that individuals with cerebral palsy were most likely to also have an ID/DD diagnosis (53.2%), followed by those with other chromosomal abnormalities (47.9%) and speech and language disorders (37.6%). In contrast, ADHD and other nervous system disorders—despite their relatively high prevalence in administrative data—had substantially lower rates of co-occurrence with baseline ID/DD conditions (16.5% and 3.0%, respectively). Co-occurrence rates with baseline ID/DD conditions tended to decline with age for most other conditions, but some, such as learning disabilities and speech and language disorders, had their highest co-occurrence rates in the adult age group (19–64). Among children and adolescents, co-occurrence rates with baseline ID/DD conditions exceeded 20% for all eight additional conditions, with particularly high overlap for chromosomal abnormalities (72.0%), cerebral palsy (65.1%), and spina bifida (50.9%).
Conclusion. The lack of some relevant diagnoses in the CCW ID and DD condition flags, coupled with definitional inconsistencies in research, reduce the generalizability and comparability of research findings for individuals with ID/DD, potentially impeding translation of research into evidence-based policies and care delivery practices for this important population. This exploratory analysis highlights the potential for substantial variation in the size of the ID/DD population depending on which conditions are included in a future expansion of existing CCW condition flags. Our analysis revealed that several other conditions—such as ADHD, speech and language disorders, and cerebral palsy—are prevalent in claims data and often co-occur with ID/DD, particularly among children and adolescents. However, the extent of co-occurrence varied widely across conditions and age groups and may be influenced by diagnostic and billing practices, service eligibility, and lifespan factors. While this analysis offers an important foundation, it is exploratory in nature. Further work is needed to finalize the conditions for use in an expansion of the baseline ID/DD definition, whether they are the eight additional conditions used in this brief or include other conditions.