Are clinical laboratory proficiency tests as good as they can be?

Present proficiency test services that use the peer group mean and statistically derived ranges of acceptability are not serving us optimally and are even counterproductive in some respects. We recommend that the target value be determined by a widely accepted reference method and that acceptable ranges be based on criteria related to clinical need. This approach was adopted several years ago in Germany and has already eliminated the use of several unsatisfactory analytical methods. Because the transition would probably take many years, we propose an interim solution to allow instrument manufacturers and laboratorians to adapt to these changes. The current peer group means and acceptable ranges should be supplemented by reference method values and acceptable ranges, based on clinical need, so that manufacturers and laboratorians can judge their performance against these new criteria and make the necessary adjustment in instrumentation and methodology. These processes should be paralleled by efforts to produce proficiency test materials that will not exhibit the matrix problems of present-day preparations.