Division of Cardiology, C.A.S.T., Azienda Ospedaliero-Universitaria Policlinico "G. Rodolico-San Marco", University of Catania, Via S. Sofia, 78, 95100, Catania, Italy.
Am J Cardiovasc Drugs. 2024 Sep;24(5):651-661. doi: 10.1007/s40256-024-00670-4. Epub 2024 Aug 1.
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.
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.
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.
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.
临床试验对于医学进步至关重要，其方法和监管途径因治疗干预的类型（即药物或器械）而有很大差异。本研究旨在确定药物或器械干预类型是否会影响心血管医学中随机试验的影响。
我们分析了 2015 年至 2021 年主要心脏病学会议上报告的最新突破性随机对照试验。主要终点是获得的总引文数。次要终点包括 1 年和 2 年的引文数、总引文数和 1 年引文数、研究实施和发表的几个指标。根据分布情况，适当使用检验比较连续或分类变量。为了调整潜在混杂因素的结果，使用了单变量和多变量回归模型。此外，还进行了敏感性分析，以探讨中性或阳性研究结果对药物与器械试验比较影响的影响，以及 2019 年冠状病毒病（COVID-19）大流行对主要终点的影响。
在 382 项合格的随机试验中，227 项（59.4%）为药物试验，155 项（40.6%）为器械试验。药物试验的总引文中位数明显高于器械研究（93[四分位距 {IQR} 48-137]比 82[IQR 39-192]；p = 0.025）。这一差异在 1 年和 2 年时是一致的，在总提及次数和 1 年提及次数上也是一致的。除了药物研究更常提前终止（8.8%比 1.9%；p = 0.006）外，所有研究实施和发表的指标均相似。在调整了多个潜在混杂因素后，引文和提及次数的差异不再具有统计学意义。然而，药物试验提前终止的可能性仍然更高（调整后的优势比=1.15；95%置信区间 1.03-1.28；p = 0.009）。阳性研究结果显著影响引文数量和试验提前终止的可能性。
药物试验通常提前终止，获得的引文和提及次数多于器械试验。然而，这些差异主要归因于治疗本身以外的因素。同时发表的研究往往会受到更多关注，阳性结果的药物试验比中性结果的药物试验被引用的频率更高。
