RESUMEN
In order to compare the prevalence of antibiotic resistance in different geographical areas, it is necessary to ensure that agreement is achieved between laboratories on the assignment of strains to 'susceptible' and 'resistant' categories. An international quality assessment study, involving 15 laboratories in eight countries, was performed to investigate the standard of performance of the susceptibility testing of Haemophilus influenzae. One hundred and fifty strains of H. influenzae were distributed from the London Hospital Medical College (LHMC) to all laboratories who were asked to test the susceptibility of the strains to ampicillin, chloramphenicol, tetracycline, trimethoprim, cephalosporins and ciprofloxacin. Laboratories were also asked to provide the details of methodology to test the susceptibility. Significant discrepancy between the LHMC and the participating laboratories appeared in the detection of resistance to ampicillin (especially beta-lactamase-negative strains resistant to ampicillin) as well as the assignment of susceptibility and resistance to chloramphenicol, tetracycline and trimethoprim. Often these reflected the use of inappropriate breakpoints which led to erroneous assignment of susceptibility. Other variations including disc content, medium and supplement, inoculum as well as failure to measure zone sizes properly also led to some repeating anomalies.