Lyme Test Problems

To the Editor.—  I suspect a methodological error in the recent article on serologic tests for Lyme disease by Luger and Krauss.¹ The authors report that one of the university laboratories detected antibodies to Borrelia burgdorferi in 18 (100%) of the 18 specimens, while all other reference laboratories (n = 8) detected B burgdorferi in significantly fewer specimens (44% to 78%).The error likely resides in Luger and Krauss' use of the same university laboratory for inclusion into the study as well as for use as a reference laboratory. I would expect close to a 100% correlation using the same technique in the same laboratory. The laboratory used to meet the entrance criteria should have been excluded from the studied reference laboratories. If Luger and Krauss had excluded the laboratory used to meet the entrance criteria, then the study should have commented on the techniques used by laboratories having a