In a nested case-control analysis, subjects who died within 90 to 365 days of hospital discharge were compared with respect to ICS exposure before death (ie, the index date) to age-matched and gender-matched control subjects who had survived to the same point in time. It was thus possible for the same individual to be both a case patient and a control subject. We compared case patients and control subjects regarding the most recent receipt of ICS between hospital discharge and the index date. Exposure to ICSs was divided into the following five mutually exclusive groups: ICSs within 30 days; ICSs in 30 to 60 days; ICSs in 60 to 90 days; and ICSs in > 90 days prior to death or not at all. We repeated this analysis for deaths ascribed to COPD and to cardiovascular causes.
Cox proportional hazard models were used to analyze the cohort study. Conditional logistic regression was used to analyze the nested case-control study. Both hazard and odds ratios (ORs) were adjusted for the effects of age, sex, Charlson comorbidity score, physician visits in the year prior to hospital admission, and medication received following discharge from the hospital. The 95% confidence intervals (CIs) were calculated for hazard ratios (HRs) and ORs. The results were considered statistically significant if p < 0.05. For analysis, we used a statistical software package (SAS, version 8.2; SAS Institute; Cary, NC). The Ethics Board of the University of Manitoba and the Health Information Privacy Committee of Manitoba Health approved the study. buy avandia online
Of 5,491 people discharged from the hospital with a diagnosis of COPD from 1996 to 2000, 1,007 (18.0%) were 35 to 64 years old and 4,584 (82.0%) were older. Of subjects > 65 years old, 562 (12.3%) died within 90 days and were excluded from the analysis. The remaining 4,022 subjects were divided into the following two groups: 1,629 subjects (40.4%) who received ICSs within 90 days of discharge from the hospital; and 2,393 subjects (59.5%) who did not (Fig 1).
Figure 1. Flow chart of treatment of subjects > 65 years of age who were hospitalized for COPD.