[Quality of a non-subsidised multicentre randomised trial in non-academic hospitals; an audit]

To determine whether a randomised clinical trial in general hospitals can be performed in a reliable way without financial support.

Descriptive.

The trial involved 1040 patients in whom an inguinal hernia had been corrected with a propylene mat in the period 1998-2003. They were randomised to either a single dose of prophylactic antibiotics or a placebo. The primary outcome measure was 'wound infection during the 3 months after the operation'. The audit concerned 98 patients, chosen at random, for whom 3 investigators from an independent bureau collected data from the patient charts and study forms during 3 clinic visits in the period 2000-2003. Whether various quality criteria had been fulfilled was assessed on the basis of these data.

In all 4 participating clinics, approval of the medical ethics assessment committees had been obtained. The trial protocol corresponded to the consent form. A proper informed consent was obtained from 63/98 patients (64%). The last (third) follow-up was missed in 23/98 patients (23%). During the trial, the patients who had missed this follow-up were contacted by telephone, which succeeded in 98%. The status and trial forms had been filled out completely and correctly with regard to the peroperative randomisation forms in 90% of the cases and with regard to the follow-up in 75% of the cases. The trial protocol was adhered to on almost every point.

The data collection in the studied, not financed trial was adequate, probably as a result of the relatively simple questions posed, the easily determinable outcome measure and the efforts expended above and beyond the protocol.