OBJECTIVE: To examine the association between epilepsy and gastrointestinal hemorrhage. PATIENTS AND METHODS: We conducted a nationwide retrospective cohort study by using data from Taiwan's National Health Insurance Research Database. Patients 20 years and older newly diagnosed as having epilepsy and nonepileptic adults were identified between January 1, 2000, and December 31, 2003, and were observed through December 31, 2008. Cox proportional hazards models were performed to calculate adjusted hazard ratios (HRs) and 95% CIs of gastrointestinal hemorrhage associated with epilepsy. RESULTS: Compared with the nonepileptic group (n=449,541), epileptic patients (n=1412) had a higher incidence of gastrointestinal hemorrhage (13.4 vs 2.9 per 1000 person-years), with an HR of 2.97 (95% CI, 2.49-3.53). The HRs of gastrointestinal hemorrhage for patients with generalized epilepsy, inpatient care, emergency care, and frequent outpatient visits for epilepsy were 3.50 (95% CI, 2.59-4.72), 3.96 (95% CI, 2.85-5.50), 4.35 (95% CI, 3.15-6.01), and 4.96 (95% CI, 3.97-6.21), respectively. Risks were significantly higher in epileptic patients with mental disorders (HR, 3.20; 95% CI, 2.55-4.01), aged 70 years and older (HR, 4.08; 95% CI, 2.89-5.77), and in the first year after epilepsy (HR, 4.81; 95%, CI, 3.14-7.34). CONCLUSION: Epilepsy is an independent determinant for gastrointestinal hemorrhage in a chronological and severity-dependent pattern. We urge the development of an adequate surveillance policy and strategy for the early prevention of gastrointestinal hemorrhage in epileptic patients.
OBJECTIVE: To examine the association between epilepsy and gastrointestinal hemorrhage. PATIENTS AND METHODS: We conducted a nationwide retrospective cohort study by using data from Taiwan's National Health Insurance Research Database. Patients 20 years and older newly diagnosed as having epilepsy and nonepileptic adults were identified between January 1, 2000, and December 31, 2003, and were observed through December 31, 2008. Cox proportional hazards models were performed to calculate adjusted hazard ratios (HRs) and 95% CIs of gastrointestinal hemorrhage associated with epilepsy. RESULTS: Compared with the nonepileptic group (n=449,541), epilepticpatients (n=1412) had a higher incidence of gastrointestinal hemorrhage (13.4 vs 2.9 per 1000 person-years), with an HR of 2.97 (95% CI, 2.49-3.53). The HRs of gastrointestinal hemorrhage for patients with generalized epilepsy, inpatient care, emergency care, and frequent outpatient visits for epilepsy were 3.50 (95% CI, 2.59-4.72), 3.96 (95% CI, 2.85-5.50), 4.35 (95% CI, 3.15-6.01), and 4.96 (95% CI, 3.97-6.21), respectively. Risks were significantly higher in epilepticpatients with mental disorders (HR, 3.20; 95% CI, 2.55-4.01), aged 70 years and older (HR, 4.08; 95% CI, 2.89-5.77), and in the first year after epilepsy (HR, 4.81; 95%, CI, 3.14-7.34). CONCLUSION:Epilepsy is an independent determinant for gastrointestinal hemorrhage in a chronological and severity-dependent pattern. We urge the development of an adequate surveillance policy and strategy for the early prevention of gastrointestinal hemorrhage in epilepticpatients.