Trends in the quality of highly cited surgical research over the past 20 years

Objective: To evaluate whether the methodologic quality of the most highly cited surgical clinical evidence has improved over the past 20 years. Summary Background Data: There have been increasing efforts to promote the practice of evidence-based medicine among surgeons, although it is unclear whether high-quality evidence is being used. Methods: A bibliometric review was performed among general surgery and medicine journals to identify the 50 most highly cited general surgery clinical research studies from 4 consecutive time periods (1985-1989, 1990-1994, 1995-1999, 2000-2004). Methodologic characteristics and overall data quality for all 200 highly cited studies were assessed by 3 independent reviewers using the validated GRADE scoring system and trends over time were analyzed. Results: Among 200 extracted articles, the quality of evidence was graded as very low in the majority of studies (48%); fewer studies contained data of low (13%), moderate (13%), or high (26%) quality. Study quality significantly improved over the time periods examined, with an increase in the proportion of studies graded as high quality from 12% in 1985 to 1989 to 40% in 2000 to 2004 (P <0.001). The improvement in data quality over time was associated with an increased proportion of randomized trials (26% vs. 48%; P <0.05) and increased reporting of adjusted, multivariate regression analyses (22% vs. 54%; P <0.001). Conclusions: There have been significant improvements in the quality of the most highly cited surgical evidence, characterized by more randomized trials and improved statistical methodology. These findings suggest that surgeons are increasingly recognizing and citing higher quality surgical evidence.