I posted this question a while ago and I'm sending a compilation of the replies to the whole list. "The latest College of American Pathology inspections require two levels of > cellular controls for CD4 and CD34 counts and also laboratory established > ranges. Since these controls tend to come in small lots and have limited > stability, how do you order and store controls (i.e. large lots to > maintain consistency or ?) and how do you establish ranges that are > statistically valid? First, several labs wanted to know the source of my question. It is based on new requirements listed in the Sept 2002 CAP Flow Cytometry Checklist. The question is: FLO.23800 These are the replies: To establish normal ranges, we run 40 normals and drop the lowest and highest value to determine the 95% confidence interval. A simple workaround for the bi-level control requirement would be to look at these analytes in the context of a single-platform determination. This would allow you to create your own multilevel controls by simply diluting the specimen. The percentage wouldn't change, but the absolute count would. Beckman-Coulter has a two levels of CD4, including a low product and also a CD34 product. Streck CD CHEX has products for 2 levels of CD34. One lab froze aliquots of PBSC harvests as their CD4 high control. The inclusion of the dual control for quantitative cellular assays is mandated by CLIA88 and the Federal Government, and is currently in place for automated hematology analyzers. So in that sense, single platform flow assays for lymphocyte subsets and CD34 Stem cells have simply been brought up to date. As far as the control reagents themselves are concerned, the flow labs can use a commercial vendor's dual controls, who we usually get 2-3 month supplies from (the stem cell are more like 1 to 1.5 months). Since the vendor determines a reference range, my reading of the guidelines would be that the individual laboratory would have to validate this range (not establish it). This is a lessor effort, and can be tailored to the actual amount of control material you have. For instance, we run the new control material and see if we fall into the vendor's established range. The frequency is typically "x1", particularly for low volume control lots. If the values fall into range, then we begin using them, but recalculate a new mean and +/- 2 sd range after 10 to 20 replicates over a like number of days (between run). ¯-------------- Best regards, Tony Bakke, PhD Oregon Health & Science U
This archive was generated by hypermail 2b29 : Sun Jan 05 2003 - 19:26:32 EST