Purpose — To describe the change in use of tamoxifen over time and across counties in Ontario.
Methods — Data from the Ontario Drug Benefit (ODB) plan, Census Canada, and the Ontario Cancer Registry (OCR) were combined and analysed to determine rates of tamoxifen use for females over 65 for each county and the province overall, by year. Rates were analyzed by repeated measures ANOVA to determine significance of changes over time. Consistency of tamoxifen use across counties was determined by the Spearman rank correlation coefficient, and overall variation between counties was described using three statistical techniques: Chi-square analysis, the extremal quotient (EQ), and the systematic component of variation (SCV).
Results — The number of one-month tamoxifen prescriptions per incident case of breast cancer rose significantly from 13.51 in 1985 to 20.54 in 1990 (p < 0.001) and to 34.06 in 1992 (p = 0.001). Viewed differently, the number of women over 65 receiving tamoxifen prescriptions per incident case of breast cancer changed from 1.91 in 1985 to 3.14 in 1990 to 4.54 in 1992. Statistically significant variation in the rate of tamoxifen prescribing was demonstrated between Ontario counties in all three years by Chi Squared analysis (p < 0.0001). Both the EQ and the SCV declined from 1985 to 1990, suggesting more uniform prescribing across the province. Little change in overall variation was seen between 1990 and 1992. All counties over time tended to prescribe generic preparations more often and shifted from 10 mg to 20 mg formulations.
Conclusions — The significant increase in the rate of tamoxifen use and trend towards more uniform prescribing across Ontario between 1985 and 1990 coincided with the publication of two important documents outlining the benefits of tamoxifen in early breast cancer. Despite this trend, variation in tamoxifen use between counties remains. There has been little change in uniformity of prescribing since 1990.