Box 1.

Problems of performing randomised trials in surgery

• Structural, cultural, and psychological resistance exists to the use of randomisation
• The inherent variability of surgery requires precise definition of interventions and close monitoring of quality
• Surgical learning curves cause difficulty in timing and performing randomised trials of new techniques
• Comparisons of surgical and non-surgical treatments with greatly different risks causes difficulties with patients' equipoise
• Rare conditions and urgent and life threatening situations cause difficulties with recruitment, consent, and randomisation