Background/aims: In liver surgery as well as other surgical fields, many randomized controlled trials (RCTs) have been performed to date. However, there has not been any detailed quality assessment of these RCTs. Therefore, this article evaluated the quality of RCTs involving liver surgery.
Methodology: A MEDLINE search was performed to identify prospective RCTs involving liver surgery published between January 1998 and December 2006. Eligible RCTs were evaluated by grading 5 items including sample size, calculation of sample size, randomization, intention to treat analysis, and loss to follow-up rate, allocating 0 to 2 points for each item. The total score as well as score for each item was compared between periods and countries.
Results: Seventy-three papers were eligible for review. Geographically, 40 RCTs were performed in Asia, 29 RCTs in Europe, and 4 RCTs in North America. Dividing by country, the most papers were published by Japan. China, France, and Germany followed in numerical order. The mean total score was 5.9 points. The mean total score in the second half period (6.3) was significantly higher than that in the first half period (5.3) with significant increases in the scores for calculation of sample size and randomization method. The mean total score in Japan was the lowest among the top three countries, with a significantly lower score for sample size.
Conclusions: RCTs involving liver surgery have been aggressively performed, especially in Asia and Europe. However, the quality of these RCTs is not so high. Improvement of internal validity with adequate study design is important to obtain reliable evidence.