The Public Health Dimensions of the Epilepsies: Epidemiologic Research and Surveillance of the Epilepsies A Systems-level Perspective David J. Thurman, MD, MPH Centers for Disease Control and Prevention dthurman@cdc.gov 770.488.6090
Definitions Epidemiologic research study of the distribution and determinants of health-related states in populations Public health surveillance The ongoing systematic collection and analysis of health data necessary for public health programs Routine practice of public health Epidemiology is the basic science of public health.
Purposes of Surveillance and Epidemiologic Studies Burden of illness Incidence & Prevalence Severity disability, mortality, quality of life Cost Risk factors / Etiology Health Disparities Access to appropriate care Trends Evaluation of public policy pnterventions
Health Data Resources for Epidemiology & Surveillance Established administrative datasets Coded hospital discharge data, vital records, etc. Established population surveys National health interview survey, BRFSS, others Retrospectively collected clinical data Hospital records, etc. Prospectively collected clinical data Data collected specifically for study More detailed, complete, and accurate Cost + ++ +++ ++++
Limitations of Health Data Resources Established Administrative Datasets Limited to information on original records (e.g., hospital/ed charts, death certificates) Limited diagnostic info in ICD epilepsy codes Few categories Can t distinguish incident from prevalent cases Errors in coding Common misclassification: epilepsy (ICD-9 345.x) vs. seizures (780.3) Difficult linkage across inpatient & outpatient data sets Not possible with CDC/NCHS National Health Care Surveys
Limitations of Health Data Resources Established population surveys Self- or proxy-reported data Diagnostic misclassifications Usually exclude institutionalized, military, and homeless populations Lower response rates may introduce bias Non-clinician interviewers Little clinical detail
Limitations of Health Data Resources Clinical Data Retrospectively collected clinical data (extracted from hospital/other clinical records) Diagnostic & other clinical details vary Requires skilled abstractors Prospectively collected clinical data (Data collected specifically for study) Methods to find/screen potential cases difficult Barriers to participation (incl. HIPAA, IRBs) may limit full case ascertainment or bias samples.
Practical Utility of Health Data Population Studies of Epilepsy Established administrative datasets At local level, can use to screen for possible prevalent cases Can use national data for temporal trends in healthcare use Established general population surveys Unsuited for epilepsy subtypes, etiology, or incidence Useful for prevalence, quality of life & healthcare access Retrospectively collected clinical data Useful for confirming possible (pre-screened) cases Prospectively collected clinical data Clinically detailed, diagnostically accurate studies of incidence & risk factors; also cohort studies of outcomes
The CDC Epilepsy Program: Supported Surveillance Activities Epilepsy questions in Behavioral Risk Factor Surveillance System (BRFSS) and the National Health Interview Survey (NHIS) Address prevalence, seizure frequency, access to specialty care, quality of life. National Health Care Surveys (CDC/NCHS) Trends in health care utilization inpatient, emergency, & outpatient. Reference: http://www.cdc.gov/epilepsy/research_projects.htm
The CDC Epilepsy Program: Supported Epidemiologic Studies Population-based studies of epilepsy prevalence or incidence (e.g., New York City, SC, Houston TX, southern AZ, southwestern KS, District of Columbia) Focus on health disparities, patterns of health care, and predictors of outcomes Studies of national epilepsy incidence using Medicare data sets Focus on all seniors Focus on nursing home Reference: http://www.cdc.gov/epilepsy/research_projects.htm
Future Needs Further defining the public health burden of epilepsy through population-based epidemiologic studies Updated cost estimates Comorbidities Disability and quality of life Health disparities Early Predictors of Outcomes Identifying new onset cases needing early referral for specialized care Epilepsy mortality surveillance Including SUDEP and injury-related mortality
Suggested Reading The CDC Epilepsy Program research website: http://www.cdc.gov/epilepsy/research_projects.htm Hirtz D, Thurman DJ, Gwinn-Hardy K et al. How common are the "common" neurologic disorders? Neurology 2007; 68(5):326-37. Kobau R, Zahran H, Thurman DJ, et al. Epilepsy Surveillance Among Adults 19 States, Behavioral Risk Factor Surveillance System, 2005. MMWR Morbidity and Mortality Weekly Report 2008; 57 (SS- 6):1-24.
End
Appendix (Optional Slides for Q & A)
Prevalence of Epilepsy: Selected U.S. Studies, 1978-2005 Copiah County, MS (1978) 7.1 / 1000 a Rochester, MN (1980) 7.1 / 1000 b N. Manhattan, New York, NY (2005) 5.0/1000 c Navajo Nation (2002) 10.2/1000 d All rates age-adjusted to U.S. 2000 population a. Haerer AF et al. Epilepsia 1986; 27(1):66-75. b. Hauser WA et al. Epilepsia 1991; 32(4):429-45. c. Kelvin AF et al. Epilepsy Research 2007; 77(2-3): 141-50. d. Parko K, Thurman DJ. Epilepsia 2009: 50(10): 2180 5.
Prevalence of Epilepsy by Age Composite of Selected U.S. Studies, 1978-2005 Sources: Begley, CE, et al. (2000), Epilepsia 41(3):342-51; Haerer, AF, et al. (1986), Epilepsia 27(1):66-75; Hauser, et al. (1991), Epilepsia 32(4): 429-45; Kelvin, EA, et al. (2007), Epilepsy Research 77(2-3):141-50; Murphy, et al. (1995), Epilepsia 36(9): 866-72.
Summary Estimates of Epilepsy Prevalence in the United States 6.8 per 1,000 (data from previous slide) 7.1 per 1,000 (Hirtz et al., 2007) Age-adjusted to U.S. 2000 population
Incidence of Epilepsy by Age Composite of 12 Studies in Developed Countries, 1988-2005
Estimates of Epilepsy Incidence United States & other Developed Countries Age Group Median Interquartile 0-4 93 70-108 5-19 37 26-44 20-64 29 24-36 65+ 77 52-109 Overall 41 32-52 Estimated annual incidence of epilepsy per 100,000 population, age-adjusted to U.S. 2000 population. Based on age-specific estimates from 14 studies in developed countries published 1988-2008.
Lifetime Risk of Epilepsy by Age Based on actuarial methods using median age-specific estimates from 12 studies in developed countries published 1988-2008.
Source: Olafsson E, Hauser WA, et al. Incidence of epilepsy in rural Iceland: a population-based study. Epilepsia 1996;37(10):951-5. Etiology of all unprovoked seizures Idiopathic 13.8% Vascular 9.0% Degenerative 6.6% Infection 1.4% Neoplasm 5.8% Trauma 4.6% Other 0.6% MR/CP 4.2% Genetic 0.4% Cryp togenic 53.7%
Access to Care Among People with Epilepsy Recent Data from the Behavioral Risk Factor Surveillance System Reference: Kobau R et al. Epilepsy Surveillance Among Adults 19 States, Behavioral Risk Factor Surveillance System, 2005. MMWR; 57 (SS-6):1-24.
BRFSS Epilepsy Survey BRFSS is state-based, random-digit-dialed telephone survey Surveys civilian, noninstitutionalized population aged >18 years Core questions used in 50 states Epilepsy questions added in some states in 2005 Screening question in 19 states 3 to 4 follow-up questions in 13 states
BRFSS - State-added Questions Have you ever been told by a doctor that you have a seizure disorder or epilepsy? Are you currently taking any medicine to control your seizure disorder or epilepsy? How many seizures of any type have you had in the last three months? In the past year have you seen a neurologist or epilepsy specialist for your epilepsy or seizure disorder?
Categories of Epilepsy Severity Lifetime epilepsy Active epilepsy ever told they had epilepsy AND currently taking AED or had seizure(s) in last 3 months Active epilepsy strata: with recent seizures without recent seizures.
Epilepsy Prevalence by Race BRFSS 2005 All Surveyed 8.4 / 1000 African Americans 8.9 / 1000 Hispanics 6.6 / 1000 Whites 8.7 / 1000 *Weighted estimate from 13 states.
Comparing BRFSS Reported Seizure Remission with Current Optimum "Optimum" Response of Seizures to AED Therapy BRFSS Reported Recent Seizures Not Remitting 30% Remit 1st Rx 50% Remit Revised Rx 20% Recent Sz 44% No Recent Sz 56% A higher proportion of PWE in BRFSS survey report recent seizures than expected w/ optimum treatment (44% vs. <30%).
A Third of People with Epilepsy AND Recent Seizures Have Not Seen a Neurologist in Past Year No Recent Sz Recent Sz 56% 44% Neuro Visit Last Year No Neuro 29% Visit 15% Source: Behavioral Risk Factor Surveillance System, 13 States, 2005 MMWR 2008; 57(SS-6)
Access to Medical Care by Level of Epilepsy Severity 25% 20% 15% 10% 5% 0% (-) Sz (+) Sz Couldn't Afford MD No Insurance