A LTHOUGH those close to the clinical laboratory have long recognized that laboratory errors can occur, the problem has not been openly discussed until recently. Walter Cronkite in a Columbia Broadcasting System program in 1965 focused lattention on the poor performance of certain mail-order laboratories and stimulated, in part, the introduction of bills in Congress to establish performance standards for clinical laboratories engaged in interstate commerce. In testifying before a Senate subcommittee, Dr. David J. Sencer, 'director, National Communicable Disease Center (NCDC), cited proficiency testing studies that demonstrated significant degrees of unsatisfactolry performance in various fields of clinical laboratory work (1). Although the results varied from labo,ratory to laboratory, he concluded that information indicates that erroineous results are obtained in more than 25 percent of all tests analyzed by these studies. As might be expected, this statement caused great concern and, at first, certain groups challenged 25 percent as being too high a percentage or maintained that it applied to laboratories other than their own. Others have maintained that this percentage is too conservative. Although s,cientists in clinical laboratories would like to be immune to error, there is no reason to expect human beings a,nd machines to obtain perfect results in a clinical laboratory when they do not elsewhere. Additional obijective evidence obtained since 1965 makes it unnecessary to belabor the point that medical laboratories can make errors or to debate the extent of the errors. All types of medical laboratori dependent, hoispital, and public hea re now generally recognized to, be subject For, and we can therefore proceed with the )f improving laboratory services. A number of significant programs and cooperative efforts have been started which should result in major improvements. Onlyua few can be considered in this discussion, but they illustrate what must be done to provide quality laboratory services for the American people.
Read full abstract