Background and Purpose--Inhospital medical complications account for a considerable portion of deaths during the early stage of stroke. However, relatively few studies have examined their long-term effects on mortality in stroke patients.

Methods--We prospectively and consecutively collected data on 579 patients with acute ischemic stroke from November 1998 to February 2001. Mortality was confirmed using national death certificate data from 1999 to 2003.

Results--During admission, one or more medical complications requiring intervention developed in 160 of these 579 patients (27.6%). For these 160 subjects, the 30-day, 90-day, 1-year, 2-year, 3-year, and 4-year mortalities were 16.3, 29.4, 46.9, 55.6, 61.3, and 70.7%, whereas the mortality figures for those without such complications (n=419) were 1.4, 3.8, 8.8, 15.0, 19.1, and 22.4 (P<0.001 with log-rank test). To eliminate the short-term effects of these complications and thus reveal their long-term effects, we investigated differences in mortality versus the presence of inhospital complications at more than 30 days, 90 days, 1 year, 2 years, and 3 years after stroke, respectively. Cox’s proportional hazard regression analysis was applied at these times after stroke and showed that all hazard ratios of medical complications in terms of mortality were statistically larger than one, regardless of adjusting for effects of potential predictors on mortality.

Conclusions--Our study shows that stroke patient mortality is influenced by inhospital medical complications significantly up to the chronic stage. This finding suggests that the appropriate prevention and management of inhospital complications could improve short-term and long-term prognoses after stroke.