I commend Franken et al1 for an outstanding attempt to eliminate selection bias from this retrospective case series comparing laparoscopic with open hepatectomy for the management of mostly malignant disease of the liver. The authors performed 52 laparoscopic hepatic resections during a period of approximately 10 years. They then tried to find comparable open liver resections with similar diagnoses and similar complexity during the same period. An investigator blinded to outcomes of the operations was chosen to match the cases. These 2 groups were thus matched and compared for outcomes. This novel approach seeks to eliminate the problem of selection bias that we see in any retrospective study. If you examine the 2 cohorts, they seem to be well matched. The patients were the same with respect to age, sex, extent of liver resection, degree of cirrhosis, dates of operation, and technique of resection, almost as if they had been prospectively randomized.
Biehl TR. A Comparison of Laparoscopic vs Open HepatectomyGood Try, but We Still Have Selection Bias. JAMA Surg. 2014;149(9):947. doi:10.1001/jamasurg.2014.1032