medwireNews: Researchers demonstrate a means of gauging the effects of drugs tested in randomised trials in real-world patients.
Adam Timmis (University College London, UK) and team sought to account for the “healthy trial participant” effect by comparing outcomes of patients with stable coronary heart disease (CHD) who participated in PEGASUS-TIMI-54 with those of 7238 similar patients from the CALIBER (ClinicAl research using LInked Bespoke studies and Electronic health Records) database.
They say their findings help to address the applicability of trial findings to general practice. “This is a question that commonly confronts clinicians, and the methods we present provide a novel means of addressing it”, they write in The BMJ.
The PEGASUS-TIMI-54 trial participants were themselves high risk, being older than 65 years, with renal impairment, two or more myocardial infarctions or diabetes. But the 23.1% of CALIBER patients who met the PEGASUS-TIMI-54 inclusion and exclusion criteria appeared even more so. They were an average of 12 years older and, over the course of 3 years, had double the rate of major adverse cardiovascular outcomes, at 9.0% versus 18.8%.
Likewise, the corresponding rates of fatal, severe or intracranial bleeding were 1.3% versus 3.0%.
When the researchers applied the 3-year relative reduction in ischaemic outcomes achieved in PEGASUS-TIMI-54 with ticagrelor 60 mg daily to these CALIBER patients, they estimated that the treatment would prevent 101 ischaemic events per 10,000 patients treated/year. This would be at the expense of an extra 75 bleeding events per 10,000 patients treated/year.
When applied to 5279 CALIBER patients who met the PEGASUS-TIMI-54 inclusion but not exclusion criteria, the corresponding ischaemic and bleeding rates were 116 and 75 per 10,000 patients treated/year. And when applied to all 7238 real-world patients, the corresponding rates were 92 and 58 per 10,000 patients treated/year.
Timmis et al say that guideline writers should attempt to extrapolate clinical trial efficacy findings to real-world populations. “The electronic health record resource we describe is scalable and offers considerable potential for answering these questions, thereby placing trial results in a clinical context”, they say.
The team also noted progressive declines in use of all classes of preventive medication in the real-world cohort, with rates generally decreasing from about 80–90% to 40–60% over 3 years, with clopidogrel use falling from about 80% to 10%.
This happened despite ischaemic event rates being “in steep ascent”, say Timmis and team, adding that this “represents a missed opportunity to protect these high risk patients against ischaemic events.”
medwireNews is an independent medical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2016