Twenty percent of secondary publications of randomized controlled trials of drugs did not provide new results relative to the primary publication

Type of evidence supporting ACC/AHA and ESC clinical practice guidelines for acute coronary syndrome

AIM

The aim of clinical practice guidelines for ST elevation myocardial infarction (STEMI) and non-ST elevation acute coronary syndrome (NSTE-ACS) is to assist healthcare professionals in clinical decision-making. We evaluated the type of studies supporting these guidelines and their recommendations.

METHODS

All references and recommendations in the 2013 and 2014 ACC/AHA and 2017 and 2020 (ESC clinical guidelines for STEMI and NSTE-ACS were reviewed. References were classified into meta-analyses, randomised, non-randomised, and other types (e.g., position papers, reviews). Recommendations were classified according to class and their level of evidence (LOE).

RESULTS

We retrieved 2128 non-duplicated references: 8.4% were meta-analyses, 26.2% randomised studies, 44.7% non-randomised studies, and 20.7% 'other' papers. Meta-analyses were based on randomised data in 78% of cases and used individual-patient data in 20.2%. Compared to non-randomised studies, randomised studies were more frequently multicentre (85.5% vs. 65.5%) and international (58.2% vs. 28.5%). The type of studies supporting recommendations varied as per the LOE of the recommendation. For LOE-A recommendations, the breakdown of supporting recommendations was: 18.5% meta-analyses, 56.6% randomised studies, 16.6% non-randomised studies and 8.3% 'other' papers; for LOE-B this breakdown was 9%, 39.8%, 38.2%, and 12.9%; and for LOE-C; 4.6%, 19.3%, 30.3%, and 45.9%.

CONCLUSIONS

The references supporting the ACC/AHA and ESC guidelines on STEMI and NSTE-ACS consisted of non-randomised studies in ~ 45% of cases, with less than a third of the references consisting of meta-analyses and randomised studies. The type of studies supporting guideline recommendations varied widely by the LOE of the recommendation.

A systematic assessment of the characteristics of randomized controlled trials cited by acute coronary syndrome clinical practice guidelines

AIMS

The aim of this study was to describe the methodological features of the randomized controlled trials (RCTs) cited in American and European clinical practice guidelines (CPGs) for ST elevation myocardial infarction (STEMI) and non-ST elevation acute coronary syndrome (NSTE-ACS).

METHODS AND RESULTS

Out of 2128 non-duplicated references cited in the 2013 and 2014 American College of Cardiology/American Heart Association and 2017 and 2020 European Society of Cardiology CPGs for STEMI and NSTE-ACS, we extracted data for 407 RCTs (19.1% of total references). The majority were multicenter studies (81.8%), evaluated pharmacological interventions (63.1%), had a 2-arm (82.6%), and superiority (90.4%) design. Most RCTs (60.2%) had an active comparator, and 46.2% were funded by industry. The median observed sample size was 1001 patients (84.2% of RCTs achieved ≥80% of the intended sample size). Most RCTs had a single primary outcome (90.9%), which was a composite in just over half (51.9%). Among the RCTs testing for superiority, 44.0% reported a P-value of ≥0.05 for the primary outcome and 61.9% observed a risk reduction of >15%. The observed treatment effect was lower-than-expected in 67.6% of RCTs, with 34.4% having at least a 20% lower-than-expected treatment effect. The calculated post hoc statistical power was ≥80% for 33.9% of cited RCTs.

CONCLUSIONS

This analysis demonstrates that RCTs cited by CPGs can still have significant methodological issues and limitations, highlighting that a better understanding of the methodological aspects of RCTs is crucial in order to formulate recommendations relevant to clinical practice.

Influence of Clinical Trials of Acute Coronary Syndrome Beyond the Primary Hypothesis: A Systematic Review

Importance

Conducting a clinical trial involves significant risks, time, and resources. The return on investment for these trials, measured by advancing health care and contributions to the scientific literature, is often uncertain.

Objective

To assess the long-term effects of major clinical trials of acute coronary syndromes contemporary to the Assessment of Pexelizumab in Acute Myocardial Infarction (APEX-AMI) trial, which did not achieve its primary objective.

Evidence Review

The Cochrane Central Register of Controlled Trials database was screened for clinical trials of acute coronary syndromes (including unstable angina, ST-elevation myocardial infarction, and non-ST-elevation myocardial infarction) with more than 1000 participants and primary results published between January 1, 2005, and December 31, 2009, in Circulation, European Heart Journal, JAMA, Journal of the American College of Cardiology, The Lancet, and The New England Journal of Medicine. For identified trials, bibliographic information, citations, trial name, registration, inclusion diagnosis, intervention type, sample size, primary outcome result, sponsor information, and academic involvement were extracted. To identify secondary analyses, bibliographic information for citing articles, their citations, and their abstracts were extracted. Clinical practice guideline bibliographies for citations of trial publications were reviewed, and the class and level of evidence of resulting recommendations were extracted.

Findings

Of 784 records screened, 30 were primary publications of 25 clinical trials. Through December 31, 2018, these trials were cited a median of 497 times (interquartile range [IQR], 424-931 citations). Trials that did not achieve their primary objective had fewer primary citations (the number of times that each published journal article with the primary [main] results of a trial was cited) (median, 443 [IQR, 396-468] vs 868 [IQR, 645-1774] citations, P = .006). The frequency of secondary analyses peaked within 5 years of the primary trial at 643. Trials that did not achieve the primary objective had fewer secondary analyses (median, 15 [IQR, 5-31] vs 18 [IQR, 10-43] analyses, P = .44) that were not cited significantly less often (median, 484 [IQR, 191-1299] vs 1124 [IQR, 410-4283] citations, P = .16). All trials were cited by at least 1 clinical practice guideline.

Conclusions and Relevance

This review found that trials that achieved the primary objective were frequently cited. Secondary research activity did not differ by primary result, and the primary trials and secondary analyses contributed to clinical practice recommendations. These data show the long-term importance of clinical trials regardless of primary outcome result.