Characteristics and Impact of Randomized Trials on Drugs or Devices in Cardiovascular Medicine

Abstract

Introduction

Clinical trials, essential for medical advancement, vary significantly in methodology and regulatory pathways depending on the type of therapeutic intervention (i.e., drugs or devices). This study aimed to determine whether the drug or device intervention types influence the impact of randomized trials in cardiovascular medicine.

Methods

We analyzed late-breaking randomized controlled trials presented at major cardiology conferences from 2015 to 2021. The primary endpoint was the total number of citations obtained. Secondary endpoints included the number of citations at 1 and 2 years, number of total and 1-year mentions, and several metrics of study conduct and publication. Statistical analysis included tests for comparisons of continuous or categorical variables, based on their distribution, as appropriate. To adjust the results for potential confounders, univariable and multivariable regression models were utilized. Additionally, sensitivity analyses were conducted to explore both the effect of neutral or positive study outcomes on the comparative impact of drug versus device trials and the impact of the coronavirus disease 2019 (COVID-19) pandemic on the primary endpoint.

Results

Of 382 eligible randomized trials, 227 (59.4%) were trials of drugs and 155 (40.6%) were trials of devices. Drug trials had a higher median number of total citations compared to device studies (93 [interquartile range {IQR} 48–137] vs. 82 [IQR 39–192]; p = 0.025). This difference was consistent at 1 and 2 years and was also observed in the number of total mentions and mentions at 1 year. All the metrics of study conduct and publication were similar, except for drug studies being more often stopped prematurely (8.8 vs. 1.9%; p = 0.006). After adjusting for multiple potential confounders, the difference in citations and mentions was no longer statistically significant. However, drug trials remained more likely to be stopped prematurely (adjusted odds ratio = 1.15; 95% confidence interval 1.03–1.28; p = 0.009). Positive study outcomes significantly influenced the number of citations and the likelihood of a trial being stopped prematurely.

Conclusions

Drug trials are often stopped early and receive more citations and mentions than device trials. However, these differences are mainly due to factors other than the treatment itself. Studies published simultaneously tend to get more attention, and drug trials with positive results are cited more often than those with neutral results.