Abstract 19139: Impact of Medical Study Publication on Cholesterol Medication Prescriptions
Introduction: The challenges of translating research findings into practice have been well demonstrated, though some high impact studies do change care delivery. It is unknown whether such trials have spillover effects onto similar but unrelated medications.
Hypothesis: Cholesterol trial findings change prescriptions for both trial drugs and statins in the direction of the results, even if statins were not directly studied.
Methods: The Symphony Health Solutions’ PHAST Prescription Monthly database provided retail prescriptions from 2007 to 2013. We selected 5 cholesterol trials that were all published in the New England Journal of Medicine. The trials [main drug(s) studied, result] were: ENHANCE [simvastatin/ezetimibe, negative], JUPITER [rosuvastatin, positive], ARBITER 6 HALTS [niacin, positive; ezetimibe, negative], ACCORD LIPID [fenofibrate, negative], AIM HIGH [niacin, negative]. For each trial, we performed a difference-in-differences analysis for both the main drug(s) studied and statins using anti-hypertensive drugs as the control, comparing the 1 and 12 month prescription averages after to an average of the 12 months before the trial.
Results: The 12-month average study drug monthly prescriptions changed in the direction expected from the trial results for all (p<0.01), except niacin post ARBITER trial. After ENHANCE, simvastatin/ezetimbe changed -533,490 (95% CI -655,008, -411,976); after JUPITER, rosuvastatin changed 337,328 (278,722, 395,934); after ARBITER, niacin changed 1,335 (-17,625, 20,294) and ezetimibe changed -185,587 (-222,697, -148,147); after ACCORD, fenofibrate changed -263,522 (-316,805, -210,238); after AIM HIGH, niacin changed -154,460 (-182,760, -126,160). Over 50% of the year change occurred in the first month after each trial, except niacin post ARBITER. The trend of increasing statin prescriptions was unperturbed by the trial results.
Conclusions: Individual cholesterol trial drug prescriptions changed substantially and as expected after trial results, with the majority of the change occurring immediately after trial publication. There was no evidence for detrimental spillover effects on statin prescriptions. These results suggest that the uptake of these trial findings was appropriate.
Author Disclosures: W.B. Borden: None. H. Lin: None. T.F. Bishop: None. A.I. Mushlin: None. K. Simon: None.
- © 2014 by American Heart Association, Inc.