Manufacturing-related quality statistics aren’t normally associated with biorepositories, but the world of statistical process improvement tools can be very useful for biobanking. One such case is the difference between two basic statistical metrics — Quality Yield and First Pass Yield. I'd like to shed some light on both metrics starting with their definition, then propose a recommendation of which to use in a biorepository setting, and conclude with a proposed performance target level.
First Pass Yield vs. Quality Yield
Quality Yield (QY), also known as Final Yield (FY)1 represents the acceptable pieces at the very end of a process divided by the original number of pieces fed into the process. In a manufacturing setting, this tells you the overall proportion of acceptable results; if the number of units of acceptable quality at the end of the process is the same as what can be expected given what is fed into the process, then you have achieved a 100% Quality Yield.1
However, QY only considers the final yield and does does not take into account such activities as diversion for re-work or re-testing. Nor does it tell you what processes involve defects: It is a high level determination of the percentage of good results relative to what you started with.1
Here is a simple five-process example:
- Process 1: 47 passed / 50 entered = .94 or 94.0%
- Process 2: 45 passed / 47 entered = .957 or 95.7%
- Process 3: 41 passed / 45 entered = .911 or 91.1%
- Process 4: 37 passed / 41 entered = .902 or 90.2%
- Process 5: 34 passed / 37 entered = .919 or 91.9%
- Quality Yield: 34 passed / 50 entered = .68 or 68%
In contrast, First Pass Yield (FPY), also known as First Time Yield or Throughput Yield, is calculated from the yields of each individual process without re-work (first pass). FPY is calculated by dividing the units entering the process, minus the defective units (regardless of whether they were discarded or fixed), by the total number of units entering the process. Many measures of productivity and efficiency fail to account for the cost of rework, and rework can consume a significant portion of value2 of the process in question.
Here is the FPY from our example, above:
- Process 1: (47 passed – 2 reworked = 45) / 50 entered = .90 or 90.0%
- Process 2: (45 passed – 1 reworked = 44) / 47 entered = .936 or 93.6%
- Process 3: (41 passed – 2 reworked = 39) / 45 entered = .867 or 86.7%
- Process 4: (37 passed – 0 reworked = 37) / 41 entered = .902 or 90.2%
- Process 5: (34 passed – 1 reworked = 33) / 37 entered = .892 or 89.2%
- First Pass Yield = (process 1 yield) x (process 2 yield) x (process 3 yield) x (process 4 yield) x (process 5 yield), or .90 x .936 x .867 x .902 x .892 = .588 or 58.8%
In a perfect world, all results of a process would be absolutely identical — zero variation! However, in the real world, variation is inevitable and is reflected statistically by the standard deviation (represented by the Greek letter sigma). The Six Sigma quality measure refers to a process in which six standard deviations fit between the target value and the customer’s specified limits. Six Sigma is widely regarded as a world class level of performance achieved by only a few companies.3
How Does This Apply to a Biorepository?
In a biorepository setting, there are several processes that can benefit from assessment of FPY, such as in- and out-bound material processing. In a typical biorepository, clients continually deposit and withdraw material, and these processes have well-defined steps associated with them.6 For example, the deposit of material may include the following steps:
1) Notification of incoming shipment
2) Arrival of shipment
3) Inspection of shipment
4) Unpacking of shipment
5) Data entry
6) Completion of Documentation
The receiving biorepository begins calculating FPY by aggregating the composite FPY for steps 2 to step 6.2 Step 1 is not included since this process is external to the biorepository.3
Can a Biorepository Achieve Six Sigma?
As mentioned earlier, the ideal FPY for any process is 100%, however, achieving even a Six Sigma performance level, which equates to 99.99966% or 3.4 defects per million opportunities, can only be accomplished with extremely efficient processes and highly dedicated internal and external participants. Biorepositories simply can’t control external participants, and thus Six Sigma is not a likely target. A better target would be a range from 3 to 6 sigma, or 93.33% to 99.9966% — the higher the better, of course. This can be accomplished through kaizen blitz on the components of each process to increase FPY.3,5 Happy calculating!
If you're interested in learning other ways to minimize your risk and optimize your processes, check out our eBook Protecting Your Protocol: Operating a Flawless Clinical Agent Repository!
1. Final Yield (FY). (April 2015). Retrieved from Six Sigma Material: http://www.six-sigma-material.com/Final-Yield.html
2. First Time Yield – FTY . (April 17, 2015). Retrieved from Six Sigma: http://www.isixsigma.com/dictionary/first-time-yield-fty/
3. Gill, A. (April 1, 2015). What is 6 Sigma – The Lean Process Guide. Retrieved from Lean Process: http://www.leanprocess.net/what-is-6-sigma/
4. Littlefield, M. (January 24, 2013). Manufacturing Metrics: First Pass Yield Benchmark Data. Retrieved from LNS Research: http://blog.lnsresearch.com/bid/170419/Manufacturing-Metrics-First-Pass-Yield-Benchmark-Data
5. Stegall, M. S. (August 1, 2013). Track the Elimination of Defects Using First-Pass Yield. Retrieved from M. S. Stegall & Associates, LLC : http://www.msstegall-consulting.com/blog-0/bid/307734/Track-the-Elimination-of-Defects-Using-First-Pass-Yield
6. Organization, W. H. (2007). Common Minimum Technical Standards And Protocols For Biological Resource Centres Dedicated to Cancer Research. Lyon Cedex, France: International Agency for Research on Cancer.