Clinical trials are usually beneficial to patients and hospitals.

We examined the cost of antibiotics used prophylactically in the first 25 patients in a controlled clinical trial comparing different antibiotic regimens in the prophylaxis of sepsis. We compared this with the costs of prophylaxis in 25 similar patients immediately before the trial started and found that the costs in the pretrial patients were significantly greater than in the trial patients. The trial is therefore beneficial to the hospital. We also examined the dispensing errors in the same patients. We found none in the trial patients but there were 77 out of 627 doses prescribed in the pretrial patients. We concluded that the abolition of dispensing errors brought about by the trial must be beneficial to the patients. They were also benefitted by the close monitoring which reveals complications at an early stage, and the improved case notes. The Hawthorne effect on medical and nursing staff is the most likely cause of these phenomena.