The president of the European Association For Cardiothoracic Surgeons (EACTS) has called upon cardiothoracic surgeons to participate in more randomised controlled trials to elevate the quality of research carried out in the specialty. Peter Licht (Southern Danish University, Odense, Denmark) made the comments in his presidential address at the 34th EACTS annual meeting (EACTS 2020, 8–10 October, virtual).
“For years the level of evidence has been poor in many surgical specialties,” Licht commented, highlighting a 1996 editorial in The Lancet which compared surgical research to a “comic opera”, stating that “although surgeons have many questions, they provide few answers”.
In the article, explained Licht, Lancet editor Richard Horton compared the output of nine major surgical journals and found that the majority of published manuscripts were case series. “Only 7% of original research papers reported data from clinical trials,” commented Licht.
“Case series are without a doubt an important contribution to surgical research but Horton questioned whether we can trust this study method to yield valid results. The answer is no,” the EACTS president added.
“Although many landmark surgical innovations have been reported for the first time as a case report or small case series, these initial studies do not demonstrate clear evidence of efficacy. And, as we will be increasingly required to demonstrate and document the safety and effectiveness of our surgical treatments, the best way to provide this are randomised trials that are at the peak of the evidence tree.
“Randomised trials, however, are rare in cardiothoracic surgery and unfortunately, even when they are completed, many are by virtue of design, sample size and insufficient power, incapable of answering the questions that researchers aim to address.
“I looked at the methodology used in general thoracic papers as it appeared in the three highest ranking journals in cardiothoracic surgery. During the first three months issued from each journal 10 years after the publication of the Lancet editorial, only 5% of original thoracic papers reported data from a clinical trial. And, apart from one study that reported secondary outcomes from a previously reported large clinical trial most of them were small in numbers,” he said.
Over 20 years on from the publication of the article, and the picture has not improved vastly, Licht commented. “In this month’s issue of all three journals there was not a single clinical trial. So the majority of thoracic research continues to be non-randomised, [and] non-controlled, which Horton in his Lancet editorial was concerned about, because they are the weakest scientific evidence and may be regarded hopelessly invalid because of uncontrolled bias.”
However, Licht did acknowledge that randomised trials can pose challenging in a surgical field, due to difficulties in standardisation of approach. “Valid conclusions can only be achieved if the intervention of a study is standardised and that is rare in surgery because skills and expertise vary between surgeons. It is much easier when a drug is investigated, because the drug is the same every time, and blinding is trivial when you use placebo drugs, but almost impossible in surgery, where it may be unacceptable not to operate, or unethical to perform a sham operation.”
Furthermore, he called for more collaboration among healthcare centres and societies to advance surgical research. Licht said: “To retain our academic reputation thoracic surgeons in Europe should explore new ways to collaborate not just for educational opportunities but also for research.
“Our associations should explore ways to provide the necessary infrastructure to facilitate research networks among thoracic surgeons. Platforms where we can recruit more centres and where trialists, experts and epidemiologists can advise us how to improve our research.”