BACKGROUND: Recruitment via the Internet for observational studies is an efficient mode, particularly when large panels of willing respondents are readily available. Previous studies have shown that despite demographic differences between Internet and non-Internet users, the underlying characteristics and relationships among key variables of interest were not significantly different (Best et al., 2001; Roster et al., 2004). An ongoing multinational Internet cohort study of smokers (referred to as the ATTEMPT study) provided the opportunity to examine differences between ATTEMPT participant and national survey data.
OBJECTIVES: The objective of this sub-analysis was to compare demographic and smoking history characteristics of smokers in the ATTEMPT study with national survey data of United States (U.S.) smokers. identified. Within this study base, cases were defined as patients with a diagnosis of a hypersensitivity or allergic reaction. The date of the last event was the index date. Up to three controles were matched on GP practice and type of diabetes. All prescriptions for sulfonamide drugs before the index date were identified. Current drug use was defined as use of the drugs in a 14-day time window before the index date. Sulfonamide drugs were classified according to the presence of an N1-substituent and an arylamine. Sex, age, immune disease-state, and history of allergic events were tested as potential confounders and effect modifiers. Conditional logistic regression was used to calculate odds ratios (OR) and 95% confidence intervals (95%CI).
RESULTS: Current use of sulfonamides was higher among cases (40.2%) compared to controls (28.0%), yielding an adjusted OR of 2.22 (95%CI 1.96–2.53). Stratification according to the presence or absence of an N1-substituent or an arylamine showed that current use of N1(þ) N4(þ) sulfonamide drugs was associated with hypersensitivity reactions (OR 5.08; 95% 2.45–10.51). Current use of N1(þ)N4() and N()N4() sulfonamide drugs was also associated with allergic reactions, although not as strong as the association with N1(þ)N4(þ) sulfonamide drugs: adj. ORs 1.98 (95%CI 1.68–2.32) and 1.59 (95%CI: 1.40– 1.82). There was no clear evidence for effect modification.
CONCLUSIONS: In patients with diabetes mellitus, the risk of hypersensitivity reactions among current users of sulfonamide drugs was two times higher compared to non-users. Stratifcation according to the presence of an N1-substituent or arylamine showed that there was a risk difference, with the highest risk associated with the use of sulfonamide drugs with an N1- and N4-substituent.
