Abstracts

Rationale: Seizures can follow certain precursor conditions, and the prevalences of those conditions differ by race/ethnicity. With the racial/ethnic makeup of US nursing home (NH) residents changing with the US population. NH caregivers and administration should understand racial/ethnic differences in precursors and subsequent seizures due to the potentially severe consequences that seizures hold for elderly residents. We hypothesize that, following each of several precursor conditions, incident epilepsy/seizure disorder (epi/sz) rates will differ by race/ethnicity among NH elderly. Methods: We first examined Minimum Data Set (MDS) data for residents, age 65 years with race/ethnicity available, at admission (N=3,585,969) to any Medicare/Medicaid certified US NH during 2003-2005 for indication of epi/sz. Epi/sz was MDS-derived as: ICD-9 code 349 or 780.39, sz as a current condition, or epilepsy-related mental retardation/developmental delay. Those with no such indication, and with ?1 follow-up MDS (N=2,775,299), were then followed through 2007 or end of NH stay for new-onset epi/sz. Results: Of 2,775,299 NH admissions followed forward, 226,768 (8.2%) of admissions were Black non-Hispanic (B), 2,423,006 (87.3%) were White non-Hispanic (W), and 125,525 (4.5%) were Hispanic or other race (O). 63,818 (2.3%) had incident epi/sz, for a rate of 17.4 new cases per 1000 person-years (PY) of follow-up: 30.6/1000PY in B, 15.8/1000PY in W, 23.6/1000PY in O. Prevalence of stroke at admission was 26% (B), 17% (W), and 23% (O), with incident epi/sz among those with stroke of 48.5/1000PY (B), 28.8/1000PY (W), and 39.3/1000PY (O). Prevalence of hemiplegia at admission was 11% (B), 5% (W), and 10% (O), with incident epi/sz among those with hemiplegia of 56.9/1000PY (B), 39.3/1000PY (W), and 45.9/1000PY (O). Prevalence of hypertension at admission was 77% (B), 65% (W), and 70% (O), with incident epi/sz among those with hypertension of 31.0/1000PY (B), 15.8/1000PY (W), and 24.3/1000PY (O); similar differences were seen for those with diabetes, Alzheimer s disease, non-Alzheimer s dementia, and/or cancer at admission. All rates differed significantly by race/ethnicity, with the higher rates for O largely driven by higher Hispanic rates compared to Native American/Alaskan Native and to Asian/Pacific Islander. Incident epi/sz among those with none of these conditions at admission was 23.5/1000PY (B), 12.0/1000PY (W), and 14.7/1000PY (O). Conclusions: For most of the precursor conditions examined, highest prevalences were seen among B and lowest among W. As hypothesized, incidence rates for epi/sz differed by race/ethnicity for all precursor conditions, with highest rates among B, lowest rates among W, and O rates in between and largely attributable to Hispanic rates. This finding has implications for NH-implemented prevention of, and monitoring for, incident seizures. Ongoing research is investigating incident epi/sz and its association with precursor and other clinical conditions, NH staffing and quality of care, and NH neighborhood US Census characteristics. Study Supported by: CDC/ASPH #S3822