A province-level risk factor analysis of fluoroquinolone consumption patterns in Canada (2000-06)

To assess potential risk factors among socioeconomic variables and the rate of influenza for the use of different fluoroquinolone antimicrobials in Canada, and to evaluate modelling fluoroquinolone-use data by two different outcome measures. Fluoroquinolone use was described monthly from 2000 to 2006 by two outcome measurements: defined daily doses and prescription counts. Multivariable linear and negative binomial models were produced with socioeconomic and influenza rate data. Significant socioeconomic predictors varied among the individual fluoroquinolone models, which may reflect the range of infections that are treated with fluoroquinolones. However, socioeconomic variables within the ciprofloxacin and levofloxacin models were similar, and indicated that use was highest in advantaged populations, depending on the measures being assessed. The rate of influenza was a significant predictor within models describing levofloxacin use and the defined daily dose model for ciprofloxacin use, after accounting for season. Influenza significantly interacted with the education variable in the levofloxacin defined daily dose model. Significant associations between levofloxacin use and influenza rates, after accounting for season, may suggest that levofloxacin was used to treat secondary bacterial infections or was prescribed inappropriately for seasonal viral respiratory tract infections. Yearly patterns of ciprofloxacin use show that prescribing practices changed; more ciprofloxacin prescriptions were dispensed towards the end of the study period, but for smaller doses or shorter treatment times. Associations with socioeconomic variables suggest that the fluoroquinolones ciprofloxacin and levofloxacin were more likely to be used in advantaged populations, probably due to the high cost of fluoroquinolone antimicrobials in comparison to the penicillin and macrolide groups.