Table 2 Poisson models measuring primary outcomes over the 6-month post-randomization period
Dependent variable | Offset (denominator) | Adjustment | Additional details |
|---|---|---|---|
Number of antibiotics | Log of patient visits in those aged 65 or greater | Log of the baseline prescription rate (1 year pre-intervention) as well as physician years since medical school and sex [23] | Robust standard errors will be used, accounting for the correlation of multiple physicians at the same office-location using an exchangeable correlation. |
Number of unnecessary antibiotics | Log of patient visits for presumed viral condition | ||
Number of broad-spectrum antibiotics | Log of total antibiotic prescriptions | ||
Number of prolonged antibiotics | Log of total antibiotic prescriptions |