Abstract
A susceptibility testing trial that formed part of the United Kingdom national external quality assessment scheme has been described previously. Results from this first trial showed an association between error rates and particular methods and practices. Changes in methods were recommended where appropriate. A second trial and survey of methods has shown reluctance to change methods and confirmed in most cases that high error rates were associated with the same methods and practices indicated by the first trial. Recommendations on disc content, method of methicillin testing, preparation of inoculum, use of controls and use of lysed blood for sulphonamide testing based on the results from these two trials are restated to encourage laboratories to review their methods critically. A statistical analysis of the results showed significant differences in performance among laboratories, and laboratories whose performance was significantly below the mean were identified. Poor performance was associated with the use of unsatisfactory methods. In view of the critical importance of susceptibility testing in patient care it is intended to use the results of susceptibility testing in the assessment of the performance of laboratories participating in the UK national external quality assessment scheme.