"The Myth and Reality of Disability Prevalence in Public Health: Measuring Disability for Epidemiology Studies"

Suzanne McDermott, PhD, Professor, University of South Carolina School of Medicine, Department of Family and Preventive Medicine, Columbia SC 29208


Disability is a human characteristic that has many dimensions.  These dimensions include medical, functional, civil rights/legal, and social perspectives.  The inclination for most scholars and public health professionals is to pick one of these perspectives and use it exclusively in program development and research.

The medical model of disability conceptualizes disability from an impairment standpoint and has been strongly associated with measurement of deficits and counting of specific problems. The model has relied on identification of medical diagnoses and/or ICD9 or ICD10 codes from medical records. Another medical model approach used in large population based studies is to ask individuals to endorse having a disability, often presented or read from a long list of medical diagnoses and conditions. In studies such as NHANES and NHIS self report, or family member report, of a disability is followed by a verification of the diagnosis in medical records. In other studies, such as the recently released CDC study of autism, the prevalence is estimated from parental report of having a child with an autism spectrum disorder diagnosis. The medical literature relies on the counting of diagnoses and codes to calculate incidence, prevalence, and to report on specific impairments and conditions. Although the use of data collected in the framework of the medical model is prominent in the medical and epidemiologic literature today, the disability community has an uncomfortable response to this model of disability identification. On the one hand many disability specific groups need these numbers to support their political action.  On the other hand advocates often feel the medical model is negative and dispassionate.  By and large this model has been discounted by proponents of the functional limitations and the social models, and they view it as passé.

For the past decade the dominant dimension used for the classification of disability has been the functional model. Using criteria that ask about limitations in function and use of assistive devices can exclude some people who have made adaptations or do not acknowledge differences in their functional ability. Individuals with mental retardation and other cognitive differences rarely acknowledge limitations in function and are usually missed in prevalence counts of disability.  And on the other hand, the inclusion of people with functional limitations in the definition of disability can over representing the magnitude of disability in the US.  The latest report from the Dana and Christopher Reeve foundation found that 1 in 50 Americans (1.9%) has paralysis, defined by the study as a central nervous system disorder resulting in difficulty or inability to move the upper or lower extremities.  According to this definition a person with a bracial plexis injury has paralysis.  Does this broad definition fit with the reality of the general perception of paralysis? Likewise, the Behavioral Risk Factor Surveillance System uses a two-question screener for disability (Are you limited in any way in any activities because of physical, mental, or emotional problems? Or Do you now have any health problem that requires you to use special equipment, such as a cane, a wheelchair, a special bed, or a special telephone?). This definition very likely captures people with obesity, advanced age, and chronic illness (such as COPD, asthma, diabetes, CHF). Some of these people do not have an impairment that results in disability.  And again, people with a cognitive impairment almost always respond negatively to both BRFSS questions.  As a consequence estimates of obesity and chronic conditions (diabetes, etc.) among people with disabilities, can be overestimated using a cross sectional dataset, when the condition might be causing the disability response.

The social model of disability focuses on environmental, economic, and political barriers to inclusion since the underlying belief is that disability is a social construct. The definition of environment used by proponents of the social model, include physical, social, economic, and political dimensions. Thus, the usual unit of measurement for proponents of the social model is the barriers in the environment, and only secondarily the individuals who are denied participation.  This perspective is useful to monitor compliance with the ADA and to measure and compare communities. It only allows us to study individuals from a limited perspective.

There have been attempts to provide an integrated model of disability in the literature.  The Nagi Model of the disabling process has been widely used. And the International Classification of Functioning, Disability and Health (ICF) received the endorsement of the international community through the WHO.  The ICF model is a complex classification that can provide a rich description of an individual.  However although this system of classification recognizes the interaction between the individual and the environment, and includes many factors, it does not meet the criteria of parsimony that is a core element of epidemiologic analyses.

The CDC and a number of other organizations endorsed the use of a two question screener to identify people with disabilities, so comparisons could be made about prevalence between states and to identify disparities in term so of health status and service utilization. These questions were crafted from the functional view of disability. The BRFSS is the national system that has become the cornerstone for these questions and the resultant multivariable analyses that have been used to explain health disparities for people with disabilities. It is time to step back and question whether this is the best we can do.  Can we really accept the notion that a Chartbook on disability which shows photographs of a person using a wheelchair contains information that attributes higher rates of obesity to people with disabilities or should we ask if people with obesity are responding affirmatively to one of the disability questions and distorting our notion of cause and effect.

This paper is intended to endorse a multi-dimensional approach that is parsimonious and useful for epidemiology.  The challenge is to identify a small set of questions that identify social, function, and medical issues that constitute a disability case definition. We need to be able to count people with disabilities to quantify service and support needs and to study the life course of people with specific disabilities. In addition, it is important to have more precise measures of disability so comparisons can be made to identify disparities and to measure changes over time.  We need to be careful so there isn’t confusion about cause and effect (attributing secondary conditions as outcomes of disability when they are in fact the cause of the functional limitation) and comparisons to people without disabilities are fair.  If we include people who experience disability due to obesity in the case definition of disability then we cannot say people with disability are more likely to be obese.  The creditability of disability epidemiology and the disability field is at stake.


A few questions, taken together, might improve the sensitivity and specificity of our case definition. 

  1. Do you use disability related equipment, such as a wheelchair, walker, TDD, augmentative device?
  2. Do you identify with or have you received special services as a member of a disability group?
  3. If yes to either question, what is your disability? 

If we asked these questions instead of the two BRFSS functional limitation core questions, would it be legitimate to exclude people who identified as very old, frail due to a progressive disease, or obese, from our definition of disability?


Biographic Sketch

Dr. Suzanne McDermott is a Professor and Research Director in the University of South Carolina, School of Medicine, Department of Family and Preventive Medicine, where she has been on the faculty since 1988.  Dr. McDermott conducts research related to causes of disability in children and the lives of people with disability.  She currently has over $2 million in active grant support from the NIH, CDC, and HRSA, and over 70 journal articles, and a book about disability and health for adults with developmental disabilities. Dr. McDermott teaches medical students and residents, works with state agencies to evaluate public health and disability programs, and has a daughter with mental retardation.