PEPFAR Quality Control and Method Validation Activity 9: Part 2

If our current mean shifts so that the 1.65 SD is at the Activity: Calculating ASEc Right Key Points Regarding Sigma 55 58 12 31 mmol/L SD assign 2 mmol/L 3 mmollL 3 mmol/L 1 mmol/L Target 16 mmol/L Target 56 mmol/L Tb s = 16 mmoI/L Target Value = mmol/L assign 16 mmol/L assign 15 mmol/L assign 16 mmollL Target Value 2 mmollL aEA = 15 mmol/L TEA = 2 mmoI/L TEa 22 mmol/L TEA = 10 mmoI/L TEa 22 ass Ign Target Value 12 mmollL Target Value 128 mmollL X assign 16 mmollL assign 16 mmollL Target 16 mmol/L X assign 15 mmol/L assign 16 mmoIlL Target 56 mmol/L X assign 16 mmoIlL 2 mmollL 1 mmol/L 2 mmol/L 3 mmoI/L 3 mmol L Target Value 1B mmol/L Sigma-metrics Rule Selection Sigma-metrics Rule Worksheet 1: Calculating SEC and Sigma assign 15 mmol/L TEA = 15 mmoI/L assign 15 mmdL aEA = 15 mmol/L 15 mmol/L SD aSgn 3 mmol/L 3 mmollL Target Value 12 mmol/L Target Value 18 mmol/L Target Value = 18 mmoI/L Target Value = 18 mmol/L Target Value = 12 36 18 15 SD assign Margin for Error What is 24 nsEc Critical Systematic Err,r Critical Systematic Error (LSEc) Critical Systematic Error Sigma-metric • Sigma-metric Critical Systematic Error (ASEc) assign 23 mmollL assign 23 mmol/L 20 mmol/L mmol/L 0bs= 23 mmoI/L k assign 23 mmol/L i assign 23 mmollL assign 16 mmollL 3 mmollL 3 mmol/L 2 mmollL Target Value 128 mmollL Target Value 28 mmollL Target Value 18 mmol/L Target Value = 12 Target Value = 28 mmollL Target Value 28 mmol/L Target Value 128 mmoI/L TEa 20 error limit, how much of our data measurements will SD asSign SD assign Result .3 SD -5 SD .4 SD +7 SD + 1 SD +4 SD +2 SD +2SD +6 SD +8 SD +5 SD + 2 SD vs. 79 36 SD Lower Limit Upper Limit What will you needo? What will need? What will you dop What will you do? If our current mean shifts so that the SO at the Activity: Calculating ASEc Directions: Use the following key numbers to complete the table below. Right Purpose Kev Points Regarding Sigma 50 Use the Sigma-metric Rule Selection Margin for Error = TEA -TE Margin for Error What What is the Critical Systematic Error (DSEc) Critical Systematic Error (LSEc) Sigma-metric Critical Systematic Error Result error limit, how much of our data measurements TE < Upper Limit Lower Limit exceed the TEA limit? Purpose What What do? What vou do? What vou vou do? +3 SD .4 SD I SD Taruet Value Target Value 21 24 20 Upper TEA Limit Lower TEA Limit • Relationship exists with SEC exceed the TEA limit? Taruet Value Target Value True Value True Guide • Relatvonsh:o exists With SEC • Relationship exists With SEC Guide Using simple math and working in 47 44 67 • Combines the Kev Quality Nun)bers Into • Combines the Kev Quality Numbers Into • Combines the Key Quality Numbers Into Calculate the ASEc and Sigma- 36 35 73 Indicates how far Upper TEA Limit Upper Limit Lower Limit Sigma= [(TEa-l Biasob l) SD factor 28 79 factor ' Combines the 4 Key Quality Numbers into • Combines the 4 Key Quality N -Imbers into Bias ' Indicates how far the +8 SD (SEC = Sigma-1 65 or - SEC 1 66) Target Value Sigma-metric Rule Selection Guide Sigma-metrics Rule Selection Guide Guide Mean SD 2SD Mean 16 mmoI/L 16 3 mmol/L 16 mmoW/L 2 mmollL 10 16 mmoWL 10 mmol/L 16 mmoI:/L Answer one statistic = Biasl)/SDl - Sigma Sigma= ASEc Bias0bsl)/SD0bs] — z factor •é - QC SD •6 QC t SD - 1 SD - 5 SD - QC SD pairs, -4 SD - t St) - do 2 SD - SD pairs ASEc metric using the four key nun-bets metric using the four key numbers 24 21 50 14 29 44 40 • Sigma-metric • S'gma-nnetr•c • Sigma-metac Sigma-metric • Sigma-metr:c = - 1.65 or Sigma = + 1.65) 73 79 6 sigma - any QC will do (just don't use 2 SD limits) • Measures the multiples of SD that fit Target Value +8 SD .5SD quality 0b s = 23 mmoI/L 0b s = 55 mmoI/L 0b s = 16 mmoI/L assign 15 mmol/L 55 mmoI/L Mean = 15 mmol/L Answer 6 oceßent - easy to control using SQC 6 - easv to control using SQC supplied in the table. Suggest a supplied in the table. 47 50 38 44 61 31 Bias 5 sigma - single-rules such as 1:3s or 1:2.5s • Indicates far the nnean can shift within limits Use the information supplied on within within T E. one statistic Limit BiasI)/SD] - 1.65 ASEc = BiasI)/SD] - 1.65 Lower mean can shift before results are from results Upper upper Upper TEA Limit ASEc = 20 • 6 sigma - any QC will do (just don't use 2 SD limits) 4 sigma - multirules S good control rule based on control rule based on the • Benchmarks the performance of the quality performance before quality performance before quality performance requirements 12 18 to exceeded I SD -ISD - SD -I SD +2SD -3 SD .2SD a SD .3 SD +7 SD • ISD - 6 SD < 4 sigma - multirules with look-back to previous < a o,gma - rnultirules with 38 Taruet Value Target Value = mmol/L Target Value Target Value = 12 Target Value 12 mmollL Target Value 128 mmoI/L Worksheet 1 to calculate the ASEc and 4 fair • Sigma-metric runs, the runs. increase the runs, Increase the ron runs. Increase 24 measurement procedure •n relationship to measurement procedure an relationship to 12 20 13.4% r data runs, increase the number of controls analyzed 61 ASEc = Sigma — 1.65 provided Sigma-metric Rule provided Sigma-metri + 1 SD .3 SD -3 SD .65 3 poor — must •nnprove n)ethod 3 poor — nnust •mprove n)ethod 3 poor — nnust •nnprove n)ethod ' Measures the multiples of SD that fit methods nwthods SD quality performance the target value Target Value 3 or less - look for better analytical methods • 5 sigma - single-rules such as 1:3s or 1:2.5s Sigma-metric. Sigma-metric or 1:2.5s the quality required Scale) the quality required (l e Scale) wrong test result a factor of 1 65) WHERE WHERE WE 38 35 Answer = Answer 24 T E.% = 20 mmol/L T Q 10 mmol/i- T Q: 10 mmQl/L 20 10 mmol/L T E. 10 T E. 10 mmol/L T E. 10 mmol/'L T E. = 10 mmol/'L SO 0b s = 16 mmoI/L SD0bs= 2 mmoI/L SD0bs= 3 mmoI/L SD0b s = 23 mmoI/L 2 mmollL performance Selection Guide. 20 18 67 0b s = 35 mmoi/L • SD .7 SD +2SD +8 SD .4 SD +6 SD -ISD TEA Z factor ot Gaussian • Can be used to select appropriate control TEA 0 6 excellent — easy to control using SQC 22 41 20 36 21 24 Can Identify and target laboratory resources o: Can Identify and target laboratory resources or Can Identifv and target laboratory resources 0' Q Close enough to true 'o be considered true Close enough to to be considered true ASEc = - 1.65 Use the Sigma-metric Rule Selection • Use the Sigma-metric Rule Selection Sigma-metrics Rule Selection Sigma 38 35 TE • Can be used to select appropraate control • Can be used to select appropriate control rules for a method that T E that Lab D • 4 sigma - multirules ' Indicates how far the mean can shift within TEA limits = Biasl)/SD] - 1.65 = Biasl)/SD] - requirements are + 1 SD poorly controlled measurement procedures 42 Lower Limit Lower TEA Limit Upper TEA Limit ASE-c Biasl)/SD] - 1 65 BiasI)/SDl - 1 65 Upper Upper Limit Lower • ISD +2SD I SD Or uses tector of 1 65 Dr: W'estgard use; a t factor of 1 65 rules for a method Guide provided next to the table to Target Value the target value 29 79 50 3.35 24 Post 0 5 good 148 12 145 150 150 148 13.4% Target Value the target value Target Value = 12 +1 SD What will you need? What will you needo? • < 4 sigma - multirules with look-back to previous suggest an appropriate control rule. ' Benchmarks the performance of the before quality performance requirements chmarkSa5e performance of the WHERE WE WHERE WE ARE 15 11 42 18 24 Target Value the target value Taruet Value Target Value = 12 50 -23SD +2 SD - 1 SD .5SD .2SD 12 10 22 52 + 1 SD 1 SD Target Value = 128 mmoI/L Target Value = 28 mmoI/L Target Value = 56 mmoI/L Targei 412 mmol/L Target Value = 12 mmoI/L 3 mmol/L et Value = mmoI/L • Worksheet 1: Calculating SEC • Worksheet 1: Calcul Target Value = 18 mmoI/L Verify your responses with the pair 0 4 fair runs, increase the number of controls analyzed 50 29 24 76 21 26 20 TEA How Far Can Your Mean Shift? QC Rules Based on the Sigma-metric SEc Sigma rong WANT TO measurement procedure in relationship to asurennent procedure in relationship to • Calculation Device ( • Calculation Device (e.g. phone, ' Indicates how far the .3 SD .7 SD located closest to you. Platelet SHIFT ASEc = Sigma - 1.65 ASEc = BiasI)/SD] - 1.65 OUR IBiasl)/SD1 -1 65 the target value Target Value Count 14 46 40 36 41 49 47 67 50 -3 SD • SD -36 SD .3 SD I SD .ISD -4 SD +4 SD -5 SD • ISD +7 SD SD calculator) 0 3 poor— must improve method -ISD SEc + 1.65 • 3 or less - look for better analytical methods 1.65 Resolve any discrepancies between BE TE 109/L) +4 SD -3 SD .3 SD .5 SD .4 SD -4 SD .2 SD the quality required (i.e. Sigma Scale) Dr. Westgard uses a z factor of 1.65 wrong test result with a z factor of 1.65) quality requirea (i.e. Sigma Scale)__ from ignificant WHERE WE ARE 16 14 44 29 35 38 22 .4 SD - 2 SD +4 SD -7 SD -Z SD Tb s = 16 mmoI/L TEA = 20 mmoI/L TEA = 10 mmoI/L TEA = 22 mmoI/L 10 mmol/L the air the pairs. SD -1.65 SD - 5 SD .5 SD -4 SD 1.65 SD TEA = 15 mmoI/L TEAs = 15 mmoI/L 0b s = 15 mmoI/L TEa 22 mmol/L assign 15 mmol/L aEA = 15 mmol/L - 2 SD • SD - 5 SD 1. performance Lab A 12 42 28 10 29 46 - 2 SD - 1 SD 1.65 SD -5 SD - 5 SD I SD +2SD +3SD -3 SD -3 so .3 SD Participate in Worksheet I's class ' Can be used to select appropriate control Can be used to select appropriate control cam shift before Z factor - defines the tail of from WHERE WE ARE Z factor - defines the tail of the Gaussian 67 +3 SD -3 SD +7 SD .3 SD -2 SD - 6 SD -36 SD +4 SD Il 1.65 SD - 2 SD - 1 SD After hange discussion. • Can identify and target laboratory resources on identify and target laboratory resources on Ciose enough to true to be considered true .ISD +2SD -ISD -3SD +2 SD distribution that exceeds TEA limit Upper Limit Lower TEA Limit 37 24 34 64 10 46 26 23 38 • Can be used to select appropriate contro rules for a method ' Can be used to select appropriate control -3 SD -4 SD .4 SD -8 SD -7 SD crossing the TEA limits WANT TO BE Upper TEA Limit Lower Limit Lower TEA Limit +4 SD -4 SD -3SD + SD .3 SD +3 SD -3SD +2 SD .3 SD -2SD -ISD -3 SD -2SD -3SD -4 SD -2SD I SD -3SD Upper TEA Limit Upper Limit Limit +7 SD -4 SD Lab D 38 11 37 33 poorly controlled measurement procedures -7 SD +8 SD + 1 SD .2 SD ASEc = BiasI)/SD] - 1.65 Upper TEA Limit Lower Limit Lower TEA Limit .5 SD .3 SD +7 SD .7 SD +2SD Upper Lower Dr. Westgard uses a z factor of 1.65 125 -36 SD rules for a method 21 21 From www.westgard.com/links-to-india-part-i-best-qc-practices-and-westgard-rules.htm. Last accessed on March 5, 2013. -7 SD Activity 9: How Far Can Your Mean Shift 12 18 15 29 20 31 22 24 14 21 23 16 26 13 10 46 17 36 32 19 50 44 TEA 11 1.35 TEA Limit 20 minutes 23 24 12 20 10 14 29 73 18