|
|
The latest trend in quality management is called "Six Sigma" and you can expect to see applications in healthcare in the near future. Six Sigma represents the evolution of Total Quality Management with a more quantitative assessment of process performance and clearer goals for process improvement.
The principles of Six Sigma go back to Motorola's approach to TQM a decade ago. Motorola established a goal that 6 sigmas or standard deviations of process variation should fit within the tolerance limits for the process, hence the name Six Sigma. Many of the leaders of Six Sigma, such as Harry and Schroeder [1], originally worked for Motorola.
The power of Six Sigma comes from having a universal measure of process performance on the "sigma scale" to facilitate benchmarking across industries. The Six Sigma methodology can be applied anytime the outcome of a process can be measured. For many processes, poor outcomes can be counted as errors or defects, expressed as defects per million (DPM), then converted to a Sigma metric using a standard table available in any Six Sigma text [1]. At this time when outcomes are of great interest in healthcare, Six Sigma provides a general methodology to describe performance in quantitative terms that will be widely understood.
To illustrate this calculation, consider the well-known problem with Firestone tires on Ford SUVs. A poor outcome, or defect, can be defined as a tire blow-out that causes an accident. Using data available to the public, there have been 2000 accidents in vehicles equipped with 6,000,000 tires, therefore the defect rate is 333 DPM (2000/6,000,000). Using the DPM to Sigma conversion table [1], this figure corresponds to 4.9-sigma performance. You can also do this using the Sigma calculator on this website.
Probably only a few processes in healthcare perform as well as Firestone tire production! I'm sure you'll find that statement shocking, but error rates of 1% to 2% are commonly considered acceptable for healthcare processes. Those error rates correspond to 10,000 to 20,000 DPM or process performance of 3.8 to 3.6 sigma. We should be striving for error rates of 0.1% (4.6 sigma) to 0.01% (5.2 sigma) and ultimately 0.001% (5.8 sigma).
The first application of Sigma metrics to laboratory data was published in 2000 by Nevalainen et al [1], in what will someday be referred to as a landmark paper. Order accuracy was observed to have an error rate of 1.8%, or 18,000 DPM, which is 3.6 sigma performance. Hematology specimen acceptability showed a 0.38% error rate, or 3,800 DPM, which is 4.15 sigma performance. The best performance was observed for report errors, which were only 0.0477%, or 477 DPM, which is 4.80 sigma performance. The worst performance was TDM timing errors of 24.4%, or 244,000 DPM, which is 2.20 sigma performance.
For comparison or benchmarking purposes, Nevalainen cited the following figures. Airline baggage handling shows a 0.4% error rate, or 4000 DPM, which is 4.15 sigma performance. Airline safety (from the normal system of random causes, not assignable causes such as the recent terrorist hijackings) has a very low fatality rate of 0.43 deaths per million passenger miles, which is better than 6-sigma performance. And, as illustrated earlier, Firestone tire production is near the 5-sigma performance level.
Fortunately for laboratories, it is very easy to assess the performance of analytical testing processes on the sigma scale. The maximum tolerance limits can be taken from CLIA proficiency testing criteria; process variation and bias can be estimated from method validation experiments, peer-comparison data, proficiency testing results, and routine QC data. To calculate the Sigma metric, you can take the CLIA proficiency testing criterion, subtract the bias observed for your method, and divide by the SD or CV of your method.
For cholesterol, for example, the CLIA criterion is 10%. Given a method with a bias of 0% and a CV of 2%, the process would be characterized as having 5-sigma performance [(10-2)/2]. Given that the goal for "world class quality" is 6-sigma performance, it follows that the goal for imprecision for a cholesterol method should be 1.7% (10/6).
By comparison, the NCEP national guidelines for cholesterol methods specify a CV of 3.0% (or better) and a bias of 3% (or better), which could lead to a process whose performance is as poor as 2.33 sigma [(10-3)/3]. In any other industry, a process having less than 3-sigma performance would NOT even be considered for routine production or operation. With that one example, you can see how Six Sigma will impact our thinking and understanding of process performance and change our goals and specifications for analytical methods.
Would you like to explain to patients why healthcare processes are seldom as good as airline baggage handling and Firestone tire production? The "why" is that we ourselves have not properly understood how to assess the quality of our processes and to set goals for process improvement. Six Sigma will change that forever! Laboratories are fortunate because the concepts can be easily applied and will provide a much more quantitative methodology for managing the quality of laboratory services [3].
For more detailed discussion of Six Sigma concepts and their application for establishing performance specifications and QC procedures for laboratory tests, see the following resources on this website:
- Six Sigma quality management and desirable laboratory precision. http://www.westgard.com/essay35.htm
- Six Sigma quality management and requisite laboratory QC. http://www.westgard.com/essay36.htm
- Six Sigma quality design and control processes. http://www.westgard.com/lesson67.htm
- Harry M, Schroeder R. Six Sigma: The Breakthrough Management Strategy Revolutionizing the World's Top Corporations. New York:Currency, 2000.
- Nevalainen D, Berte L, Kraft C, Leigh E, Morgan T. Evaluating laboratory performance on quality indicators with the six sigma scale. Arch Pathol Lab Med 2000;124:516-519.
- Westgard JO. Six Sigma Quality Design and Control: Desirable precision and requisite QC for laboratory testing processes. Madison, WI:Westgard QC, Inc., 2001.
Six Sigma DPM Table
Six Sigma Basics: Process improvement, goals, and measurements
Six Sigma: Outcome Measurement of Process Performance
Six Sigma: Quality Design and Control Processes
Six Sigma: General attributes of the OPSpecs Design ToolSix Sigma: Medical Cutoffs as Tolerance Limits
Six Sigma: Quality Design and Control Applications
Six Sigma: Patient Data for Assessing Performance and Stability
Six Sigma: Newborn Screening
