Impact of cisapride label changes on codispensing of contraindicated medications

Objective To quantify the impact of the first three label changes and 'Dear Doctor' letters as sole interventions, sent to prescribers and pharmacists, on the prescribing and dispensing of cisapride and contraindicated drugs. Methods Using a managed care claims database, a total of 38 757 patients with cisapride prescriptions (July 1993-December 1998) were selected. An interrupted time series analysis and an exponentially weighted moving average analysis (EWMA) were conducted to determine whether there were changes in the proportion of contraindicated codispensings following each intervention. Results 3.6% of cisapride dispensings overlapped with contraindicated drug dispensings. Beginning in 1994, the overlapped proportions declined slowly from 4.5% ending at 3.2% in 1998. The interrupted time series analysis showed that there was a statistically significant abrupt permanent decrease in the proportion of contraindicated dispensing following the June 1998 label change (coefficient - 1.0094, p = 0.045), meaning that the proportion of contraindicated dispensing dropped nearly one percentage point after the third intervention. The 1995 label changes did not have statistically significant effects. The EWMA analysis showed that the proportion of contraindicated dispensings was lower than 3 standard deviations below average during the period after the June 1998 intervention, with no significant effects following the two earlier interventions. Conclusions The June 1998 cisapride label change and accompanying 'Dear Doctor' letter had a downward impact on the proportion of dispensing of contraindicated drugs with cisapride. The other two label changes did not have independent significant impacts. Copyright (C) 2003 John Wiley Sons, Ltd.