Investigation of cross contamination in a Mycobacterium tuberculosis laboratory using IS6110 DNA fingerprinting.

SETTING: A laboratory for routine culturing of Mycobacterium tuberculosis.
OBJECTIVE: Investigation of an episode of laboratory cross contamination using IS6110 restriction fragment length polymorphism (RFLP) typing. Improvement of laboratory protocols to prevent contaminations in the future. To stress the importance of 'good laboratory practice', and interaction with clinicians about laboratory results.
DESIGN: Fingerprinting of mycobacterial isolates from 1) cultures suspected of being contaminated and 2) strains suspected of being the source of the cross-contamination.
RESULTS: RFLP typing results indicated that clinical samples were contaminated by strains which had been processed in species identification procedures one day earlier in the same safety cabinet. This cross contamination also resulted in exceptional RFLP typing results--mixed banding patterns. Three patients were treated on the basis of false-positive laboratory results. Because the laboratory results were confusing for the clinicians, the treatment of one true tuberculosis patient was severely delayed.
CONCLUSION: 'Good laboratory practice' is very important to prevent cross contamination. RFLP typing proved to be a useful tool to trace the source of contamination. Interaction with clinicians receiving doubtful results is of the utmost importance.