Test errors are significantly more likely to occur at doctors' office laboratories than at larger settings such as independent labs and hospital sites, two new studies suggest.
Federal researchers said patients should not be alarmed: Most labs are performing satisfactorily. But an author of the second study, done in California, advises patients to avoid having routine lab procedures done in doctors' offices.
The studies, in the Feb. 11 Journal of the American Medical Association, come while federal legislation is pending that would ease government regulations on medical labs that took effect in 1994. The American Medical Association backs the proposed legislation, claiming the regulations put too many restrictions on doctors' office labs.
But the authors of two studies support the standards as they now exist. The regulations, which include occasional proficiency tests, have greatly improved the performance of such labs, they said.
Carlyn Collins, M.D., a pathology researcher at the federal Centers for Disease Control and Prevention and a co-author of the government study, said 90% of the 160,000 medical lab testing sites nationwide are outside hospitals and large, independent labs -- mostly in doctors' offices.
Staffers in doctors' office labs may have minimal medical training; in many cases a receptionist may perform routine tests, said Lee Hilborne, M.D., a University of California at Los Angeles pathologist and co-author of the California study.
The CDC study examined test results from 1994, the first full year that periodic proficiency testing was required of labs in doctors' offices and other "nontraditional" settings under the federal regulations. Results from 17,058 testing sites nationwide were divided into two groups: hospital and independent labs, and nontraditional sites.
The CDC researchers reviewed results of 30 common procedures, including blood-sugar evaluations, blood-cell counts and cholesterol tests. Errors were nearly three times more likely to occur in tests performed at the nontraditional sites.
Still, Collins stressed that 91% of the tests done at nontraditional sites were satisfactory, compared with 97% of those done at traditional labs.
The California researchers analyzed results of 11 common lab tests performed in 1996 at 1,100 sites. Errors were up to four times more likely at nontraditional labs, although they were less likely in doctors' offices where licensed medical technologists administered the tests.