SUMMARY. Blunders which occurred over a 1 year period in the clinical chemistry departments of two health districts were recorded and categorized according to type and detection stage. A blunder was defined as an incident leading to an incorrect result/set of results either being reported or detected at the final checking-out stage in the laboratory. Of the total of 120 blunders-which is a blunder rate of less than 0'1 % of requests-53 (440/0) were detected at the final checking-out stage. Blunders detected after the report had left the laboratory were divided into those subsequently picked up by laboratory personnel (23); those detected by clinicians (19); and those by external quality assessment schemes (21). The types of blunder were fairly equally distributed between the booking-in (36), analysis (38), and reporting (35) stages of the laboratory process. A formal review of blunders detected in laboratories is a valuable aid to overall performance. Additional key phrases: quality assurance; errors; analysisDespite passing interest during each of the three last decades in 'blunders' occurring in clinical chemistry laboratories, the situation remains as Northam! described it in 1977, namely 'that although much effort and expense is being devoted to the assessment of analytical variation, little attention has been directed to the detection of laboratory blunders'.That seriously addressing the issue of laboratory blunders has not occurred is surprising, because the most recent study-reported a rate of O' 3% in a large laboratory. Although this figure represented a significant improvement over the rate of 2'3% found by McSwiney and Woodrow;' it does imply that at least two serious errors occur each working day in large laboratories.To investigate the current situation, the South East Thames Quality Assurance Group decided to monitor the number, type and detection stage of blunders which were picked-up over a 1 year period in each of two health districts.Correspondence: Mrs R Lapworth. METHODA blunder was defined as an incident leading to an incorrect result/set of results either being reported or detected at the final checking-out stage in the laboratory. Errors detected before or during analysis, or before the final validation stage were not included. The staff in each laboratory were informed that the study was being carried out and agreed to bring any blunders that occurred to our attention. Details of these, and at what stage the blunder was detected were recorded by ourselves or occasionally by other senior laboratory staff. The blunders were then categorized according to type and detection stage.Blunders which affected samples distributed by the external quality assessment schemes (EQAS) during the period of this study were included if the results were more than three standard deviations (SD) from the method mean. The blunder rate affecting EQAS samples was also calculated separately as they may give a better estimate of the true detection rate.?Users of the service were not specifically asked to bring blun...

We believe all errors, whether analytical or nonanalytical, are potentially serious. All laboratories would be well-served by continuously monitoring their blunder rate and we agree with the authors that this should be carried out on a permanent basis.Incidentally, following the report of our study, the then Hospital Board of our clinical biochemistry laboratory allowed us to purchase our first laboratory based computer, based on our suggestion that computerization should reduce the reported rate of non-chemical errors. ROBERT Authors' replyIn our study! we defined a blunder as an incident leading to an incorrect result/set of results being reported, rather than the number of results affected. This was because we were more interested in the types of error that were detected and if these differed between the two laboratories. The blunder rate in Laboratory X was 0'3070 when calculated by the method of Chambers et 01. 2 We postulated that blunder rates expressed in this way will be affected by the test: request ratio. In addition, a single blunder incident might lead to several hundred erroneous results which would make inter-laboratory comparisons difficult to interpret.For individual laboratories, the main objective is to maximize the detection of blunders. This requires knowledge of both the types of mistake and the stage at which they occur in the laboratory process, so thatsteps can be taken to minimize their occurrence... Letters 391Inter-laboratory comparisons will always be problematical; but we agree with Chambers et 01. that the most reliable exchange of information would be through EQAS. Blunder rates for individual laboratories could be calculated and compared if data from all the UK schemes were coordinated and reported by the scheme organizers.We are of the opinion that all laboratories should formally investigate blunders, either by an intensive search over a short period of time or by continuous monitoring. Ideally, both methods should be employed, and clinicians encouraged to contact the laboratory when they detect discrepancies.R LAPWORTH