PEPFAR Quality Control and Method Validation Activity 5: Part 4

Alerts us to cht 2 Type number sampks tes+ed 2 Type number sampks Type and number of t +es+ed Type number of GC sampks Type number sampks fes+ed Type number of GC stampks Type number sampks fesfed Type and number sampks fes+ed Type avid number of sampks •tested Type and number sampks fesfed number 3 Type and number of sompks tes+ed number 3) QC used QC used Lot changes number number Of sampks fesfed QC ased used +0 used +0 SD assigned X assigned 90 SD assigned - SD observed assigned Gaussian SO assigned Alerts us to ché control limits are set Alerts us to changes in Alerts us to ch. 3) QC used QC used Gaussian QC used +0 When Do You Create a New LJ Gaussian greater when QC used +0 QC ased +0 QC used — Controls (parallel t Assigned vs. Observed - What is the accuracy and p Wnherent Criteria Cape Clinic Laboratory control limits — Controls (parallel testing) — Controls (parallel te eval"a+e Gaussian accuracy and precision be greater when evalua+e Closer to the mean. evalua+e evalaa+e False Rejection q) The 09 periodic review abc Reject Analytical Need for Quality q) The review Wnherent 68.95-99 ID The of periodic review — Reagents We Wan Quality We Expect The 09 review Creatinine (umol/L) L-J Chart for Control XYZ (exp 30/11/XX) on Illustra Chemistry Analyzer (serial # 123) The of peviodic review of pevtcdic review q) The of peviod,c review q) The OF periodic revtew q) The review q) The OF periodic revtew control limits are The tradeoff Is that q) The periodic review for q) The of review 68-95-99 q) The of periodic review for •I) The of periodic review l) The of periodic review q) The of periodic review q) The 8 periodic review q) The OF periodic review Quality The of peviodlC review Quality 68-95.99 when Quality False Reject FalseReject False Re}ect System The of review for Quality Laborator The of periodic review Laboratory Laborator Cape Clinic Laboratory q) The of review Quality Laborator The review The pecic,diC revtew Goals closer to the mean. Laborator [Laborator rule Rule Laboratory — Reference solution Laboratoryx Laboratoryy • aborator Goals Goals •o:nrator We Are Goals Goals and rule The tradeoff is that To Be Goals here detec+inj and +rends suits de and +cencis — Reference solutiorv — Reference solutions and and +cends Normal Chart? defec+inj and you will also have a Gaussian we and +cends and *rends suits sults suits and and +rends sults and *rends and and - and *rends and and +rends Normal ts b) The achons ±a6en when b) The achons when resul+s sults suits suits b) b) The achons ±a16en when Calibration b) The achons when b) The acv. b) The achons ±aKen when b) data SD assigned 3 X assigned 90 SD assigned 37 X assigned 87 Assigned on b) The achons +Qhen when SD assigned 3 b) The achons X observed 90 SD observed 3 observed SD observed 3 b) Normal S +Q6en when resul+s when resul+s Distribution • Calibration Distribution Precision Pmprecision hen when resul+s when resul+s When Do You Create a New LJ when en when resul+s n when resul+s n when resu14s Assigned vs. Observed - What is the Creatinine (umol/L) L-J Chart for Control XYZ (exp 30/11/XX) on Illustra Chemistry Analyzer (serial # 123) Gaussian Analytical false exceed exceed exceed exceed exceed difference? oxc-eed Chart? Change Charta Sys tem Sys t em Sys tern — Expected (part of in False True Accept False Accept True Accepts Sym t em System Distribution Wider control Single Stable 68-95-99 SD assigned 6 X observed 90 XD assigned 90 X assigned 90 X assigned 87 SD assigned 2 SD assigned 2 SC assigned 2 SD assigned 3 Observed observed SD observed SD observed SD observed 3 X observed 90 schedule) limits are less. • Lot changes Analytical ro we A Re rule We ro we A Re wc Ane WC ARC WC ARC we ro ro De when wc System We us 1--1 chart we Wider cont WHERE WE DE ' Mean Nbnormal WHERE WC ro WHERE we ro BE WHERE we ro we ro : ro ro Be' WE ro ro ' ro ro Mean so WHERE WE ARE ERE WC WHERE WHERE WE ARE — Unexpected (a pro — Unexpected (a prob Distribution ro we A Re we sensitive to er Mode: Normal Quality Quality k I Quality } Quality ty wc ro WHERE we expecr ro ' WHERE we WHERE WHERE WE EXPECT we ro WHERE wHERe True Accept we A Re we we we Re we False Reject False Rejection limits are less 94.0 108.0 Mean Goals Goa's We us L-J charts and rule We us L-J charts Alerts us to changes in we RE wc Mean We ARC WHERE WHERE we ARC WHERE ARC we ARC we we A Re detection b we ARE violations to ale WHERE WHERE WC suits sults Suits •utts Median = Distribution — Controls (parallel testing) +3SD corrected) WHERE WE ARE 99.7% = WHERE we ARE sensitive to error WHERE violations to alert us to a WHERE BE' WHERE WE ARE WHERE WC ARE WHERE Normal Gaussian violations to ale Mean SD Mean with Bias detection but medically signifi Rejects Mean SD 4SD +3SD false Random Major component cha past 23SD rejections medically significant error. accuracy and precision precision SD observed SD observed observed SD observed Distribution Calibration assigned Gaussian assigned Gaussign Gaussian Gaussigned assign 198.0 93.0 94.0 +43SD +3SD medically signifi False False Median False Reject False Accept less — Reagents QUALITY — æents Mean 99.7% QUALITY QUALITY QUALITY QUA .1TY Change in the testing Mean limits less rejections limits are less Single manufacturer-directe CONTROL Population Population PATIENT CONTROL CONTROL Normal CON CONTROL False manufacturer-directed. manufacturer-directed QUALITY Single QUALITY 99.7% we false PATIENT Population CONTROL 202.5 QUALITY Accepts May or may not be based on Mathematically Vastribration Single SAMPLE ey essayes (QC) 99.7% (QC) .ssarjs (QC) (QC) Population — Reference solutions Population ssages 108.0 PATIENT 99.7% : ±3SD CONTROL PA 'IENT ATIENT have CONTROL escayes essayes ±3SD +3SD SAMPLE PATIENT CONTROL PATIENT co ±3SD co SAMPLE Mess ayes SAMPI Messages SA Look for SAMPLE Single SAMPLE SAMPLE . SAMP SAY F Population False (QC) (QC) (QC) Look Look o. Population AMPLE SAY p SAM p fact calculated using data Assigned on 202,0 202.0 SAMPLE abnormal SAMPLE Single Mode. + IS True Accepts I Assigned on True 201 s • Calibration SA Normal SAMPLE Look for The accucacy an Assigned on Population SA and the chart Assigned on 102.0 202.0 True Accepts True Accept Use the Median Mean Rejects Look for observed clustering essayes SD assigned 2 Change X assigned 90 • Values applied to the L-J Chart points (facts) the chart the chart z 200.5 Accept System Use the S assigned 3 ss Nbnormal observed Use the assigned The accuracy an 1898.0 assigned 1:2s rule as away WHERE E ARE — Expected (part of instrument maintenance WHERE away fr the chart WE ARC Observed Observed True Accept clustering clusteri 1:2s rule as Target WHERE we Re we ARE Accepts 1 TEa wHeRe we WHERE TEa SD observed • Sources often chosen other than .3SD 95% Answers WHERE WE ARE 1898.0 95% a scanning away fr False away WHeRe Population and a scanning ARe observed ce a scanning schedule) the mean the mean 108.0 984.0 +4SD device the IS 76.0 914.0 -ISD detect ± ISD 96.0 ± 2SD device device +hen False Rejection IS IS Rejects IS — WHERE we ARE currently observed performance in terms of accuracy and Total Error (TE) < Total ± ISD True Accepts ± ISD False Accept 810 .2SD Assigned On and and — Unexpected (a problem during testing is We us L-J charts and rule rule observed the chart = we ARE is ts SD observed 860 .2SD 76.0 810 .2SD • 2SD WHERE We IS 780 .2SD 7810 .2SD Observed IS include: precision of the method Allowable Error (TEA) we ts IS True Accept +hen Accepts when +hen Trule True +hen corrected) correctea) violations to alert us to a 198.0 -ISD Analytical 68% = e analy+icnl change +0 fie analyiical Analytical analy+ica\ v analy+ica I Chanee fro fie analy+ical • Past data points before a Specific to your method's cal he analy+ical 987.0 72.0 True Accept Tru Acc pts 108.0 96.0 is SD observed SD assigned 9 SD assigned assigned Sys+•em has occuvted. ± ISD medically significant error. Sys+em has occunted. , cuffed. has sys+em has occuvced. •curvced. Major component changed on the instrument de 875.0 96.0 .3SD +3SD -ISD change Change performance 95% = observed SD observed X observed 90 984.0 +3SD -ISD 75.0 +4SD e analy+ical e analyåical 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Analytical analyiica\ ± 2SD Mean Change in the testing process, usually Mean xe araly+ical 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 2829 30 • Cumulative values 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 .cuvced. SD observed SD assigned X assigned SD assigned 2 t. 99.7% assigned Gaussian Value 93 84 90 93 88 86 88 95 92 94 88 90 89 87 91 90 94 88 97 39 Value 93 u 90 93 88 86 88 95 92 94 88 90 89 87 91 90 94 88 97 observed manufacturer-directed. 90 91 95 90 85 91 94 89 91 85 89 • Package insert values 38 36 ±3SD 40 34 33 37 36