Methodology for Children and Adults with Pre-existing Conditions
The U.S. Department of Health and Human Services’ Office of the Assistant Secretary for Planning and Evaluation produced this analysis. The 2008 Medical Expenditure Panel Survey (MEPS) was used to identify individuals who would likely be denied due to a pre-existing condition if they were to apply for coverage in the individual market. A multi-pronged approach was used to identify conditions that would certainly or likely exclude individuals from being offered coverage. Our condition list was generated from two sources: eligibility guidelines from 19 pre-Affordable Care Act high-risk pools and underwriting guidelines from seven major insurance carriers.1 The MEPS was used to identify whether individuals had a medical visit for any of these conditions, experienced any disability days as a result of any of these conditions, and reported that they were bothered by any of these conditions in the past year. Additional questions regarding whether individuals had ever been diagnosed with a smaller set of conditions from these lists were used to further refine our measure.
Two estimates of the share of non-elderly individuals likely to be denied coverage in the non-group market are presented. The first includes only conditions that were identified using eligibility guidelines from high risk pools; and the second includes five additional common conditions (arthritis, asthma, high cholesterol, hypertension, and obesity) and a number of common mental health conditions that would result in an automatic decline, exclusion of the condition, or higher premiums according to the seven insurer guidelines examined. The first estimate includes conditions that will very likely cause an applicant to be denied coverage, and should be considered a lower bound estimate. The second estimate includes conditions that might result in a denial of coverage, but also might result in a ‘rate-up’ (that is, a higher premium) or a coverage rider (that is, a policy that excludes coverage for a pre-existing condition). The following describes in more detail the methodology.
First Measure: High-Risk Pool Definition of Pre-Existing Conditions
Individuals likely to be uninsurable were identified in the following manner. An approach developed by the Lewin Group was replicated and identified conditions reported by five or more of the 19 pre-Affordable Care Act state high risk pools as indicating automatic eligibility for enrollment in the pool.2 This list included the following conditions: alcohol and drug abuse, chemical dependency, acquired immune deficiency syndrome (AIDS), Alzheimer’s disease, angina pectoris, anorexia nervosa, aortic aneurysm, aplastic anemia, arteriosclerosis, artificial heart valve or heart valve replacement, ascites, brain tumor, cancer (excluding skin), cancer (metastatic), cardiomyopathy/primary cardiomyopathy, cerebral palsy/palsy, chronic obstructive pulmonary disease (COPD), chronic pancreatitis, cirrhosis of the liver, congestive heart failure, coronary artery disease, coronary insufficiency, coronary occlusion, Crohn’s disease, cystic fibrosis, dermatomyositis, diabetes, emphysema/pulmonary emphysema, Friedreichs’s disease/ataxia, hemophilia, active and chronic hepatitis, HIV positive, Hodgkin’s disease, hydrocephalus, intermittent claudication, kidney failure, kidney disease, and kidney disease with dialysis, lead poisoning with cerebral involvement, leukemia, Lou Gehrig’s Disease/amyotophic lateral sclerosis (ALS), lupus erythematosus, disseminate, and lupus, malignant tumors, major organ transplant, motor or sensory aphasia, multiple or disseminated sclerosis, muscular atrophy or dystrophy, myasthenia gravis, myocardial infarction, myotonia, paraplegia or quadriplegia, Parkinson’s disease, peripheral arteriosclerosis, polyarteritis, polycystic kidney, postero-lateral sclerosis, psychotic disorders, silicosis, splenic anemia, True Banti’s syndrome, Banti’s disease, rheumatoid arthritis, sickle cell anemia and disease, Stills disease, stroke, syringomyelia (spina bifida or myelomeningocele), tabes dorsailis, thalassemia (Cooley’s or Mediterranean anemia), ulcerative colitis and Wilson’s disease. 3
Individuals were also identified who reported that they had “ever been diagnosed” with the following conditions: coronary heart disease, myocardial infarction, other heart disease, angina pectoris, stroke, emphysema, cancer, and diabetes.
Individuals who were identified by one of these mechanisms were considered unlikely to be insurable in the private non-group market and are the basis of the first estimate. This estimate should be considered a lower bound, as there are potentially more conditions that insurers consider an automatic decline of coverage.
Second Measure: Insurers’ Definition of Pre-Existing Conditions
To construct the second measure, we included additional conditions that are likely to cause an applicant to be denied coverage, be ‘rated up’ (that is, charged a higher premium), or to be sold coverage with a rider that excludes coverage for one or more pre-existing conditions.
Individuals with five common conditions – arthritis, asthma, high cholesterol, or hypertension, and obesity (BMI > 35) were included in the second measure, as were individuals who had “ever been” diagnosed with arthritis, asthma, high cholesterol, or hypertension. These conditions were found to result in a denial, an exclusion of coverage for that condition, or a higher premium for individuals in all but one of the seven underwriting guidelines we examined.
In addition, individuals who were currently being treated for neurotic and related disorders, stress and adjustment disorders; conduct disorders; and emotional disturbances (including ADHD) were included in the second measure, as were individuals who had ever been diagnosed with ADHD. These types of mental health conditions were identified in the underwriting guidelines as conditions that would result in denial, waiting periods, condition exclusions or higher premiums. Information from ASPE-conducted interviews with insurance commissioners indicated that individuals in treatment for mental health conditions were generally denied coverage in the individual market. Given the conflicting evidence a conservative approach was used and these conditions were included in the second and not the first measure. Had these conditions been included in the first measure, the estimate of likely to be uninsurable individuals would have increased from 19.2 percent to 29 percent.
1 Underwriting guideline used include: Aetna Inc., Aetna Advantage Plans for Individuals and Families, Underwriting Guidelines., 2010, pp. 10-23. Blue Cross of California, California Agent Guide, Oct., 2010, pp 36-58.; Blue Shield of California, Application and Underwriting Process Guide. Aug., 2008, pp. 21-42. ; Health Net Inc., Field Underwriting and Enrollment Guidelines, Jan., 2006, pp. 11-20. Can be found on the web at: http://binsuredhere.com/health_net.htm; Vista Healthplan and its affiliate, VISTA South Florida, Individual Product Agent Guide. 9/2006. UnitedHealthcare, Golden Rule Insurance, Underwriting Guidelines, Aug., 2009, pp. 25-30. On web at: http://www.rbins.com/files/Goldenrule_Underwriting_Guidelines.pdf
HumanaOne Health and Life Products, Agent Eligibility and Underwriting Guide 2009-2010. On web at:. http://www.thecasongroup.com/forms/enrollment/Humana/H1%20Underwriting%20Guide%2010_2009.pdf
2 Families USA Foundation, “Health Reform: Help for American with Pre-Existing Conditions.” May 2010.
3 The Lewin Group analysis also included obesity (BMI > 35) on the list of conditions that would cause a denial. We did not find obesity on the condition lists used by state high risk pools, and therefore did not include it in the list of conditions that would almost certainly lead to a denial.