PEPconnect

PEPFAR Quality Control and Method Validation Activity 7: Part 3

In this activity, participants are introduced to bias and true value. Using the information supplied by 3 key numbers, participants calculate the TE in units and percent. Additionally, participants explore the impact that bias and imprecision can have on TE and relate site-specific processes that affect TE to checklist items.

Note: Any reference to page numbers in the video may not correlate with supporting documents as documents may have been updated.

even expect l) Frequency Frequency l) Frequency o True Inherent Lb of periodic review for Analytecal 2) Type nu System Where of periodic review for The of periodic review q) The of periodic review q) The OF periodic review of periodic review The of petiodic review Nbnormal Quality System thange b) abnormal SystemJI q) The periodic revtew q) The of periodic revtew q) The of periodic revtew for 2) Type and numi Type and nu q) The of pevtodic review The OF periodic review q) The of periodic review System The OF pevtcdic revtew for Quality Criteria Quality The 09 periodic review Laboratory C Criteria We wan Imprecision True True — True 2) Type and nu Laboratory B Goals Laborator Laboratory Laboratory D Laboratory D Laboratory B Laboratory A Wnherent Laboratory C Laboratory A Goals defec+in Wnherent — True Criteria and *rends Goals n and *rends q and *rends and +rends closer to the mean. Need for We Wan Goals Distribution Laborator Laboratory and and +rends o and +rends and +cends and and *rends We Are We ARE Quality Criteria Value X Value Value; — True The of periodic review Wnherent Results and +cends Need Quality tons +ahen when r Quality Laboratory B b) The achons when Effective QC Results Need for b) The ac+tons when resul+: b) The achoos when b) The achons ±Qhen when b) The achons +Q6en when WHERE Criteria Laboratory B Criteria ons fQKen when Accuracy Results e achons when b) X Value Wnherent b) S when S when resu14 Wnherent Laborato Inherent S ±ahen when ons ±Q6en when resul+s Goals Goals ns ±QKen when resul+s Goals ns ±ahen when Laboratory D Laboratory A Criteria when LD The n when resul+s than Accuracy The tradeoff is that e Wan Where Quality and Need for We Are ns when Laborator Where Laboratory A Analytical Laboratory B Quality defec+inc 51-2148 We Wan Laboratory C q) The Pmprecision Need for We Expect q) The The that exceed Pmprecision Quality Results Effective QC We ARE ere Criteria Goals Accuracy Distribution b) System that n when ccuracy Bias defec+in l,) Frequency We Wan Precision We Are Laboratory A detecHnj frequency defec+inj Laboratory C WHERE WHERE WEP WHERE ARE WHERE WE ARE WHERE W WHERE WE AR WHERE ARE WEP t WHERE Goals Quality Frequency System Where Mean Goals exceed 0 Laboratory A True Pmprecision — True True To Be WHERE m WHERE we Where b) The ac+ior and Need for WHERE H) The b) The achons b) The Results 25 Results Mode: We Expect exceed We Are We ARE I Laborator b) The achons +aKen Mean QC We ARE imprecision Quality To TE RE Mean WHERE WE ro WHERE WE RE RE -O WHERE ARE ro WHERE WE ARE WHERE we ARE WHERE DO WE TO HERE ARE TE Where Vastriaution imprecision RE ro WHERE we RE WHERE 68-95-99 We Are — True True Value; WHERE To SE ARE WE Key Messaje Jeopotdy Criteria Results We Expect We Are 68-95-99 Pmprecision Precision Messaje Jeopardy Messqje Jeopardy Messqoe Jeopardy Messaoe Jeopardy I Bias I Abs Bias % Bias (SO) TE in % Standard Jeopardy exceed Need for Standard Trule WHERE we ARE Mean (X ) Accuracy WHERE WE ARE WHERE we ARE Precision Prec ision Value; RE Pule Goals We Wan (Target Value) WHERE WC ARE we ARE True Trule True Goals exceed Imprecision WHERE WE ro BE WHERE WE WARE WHERE WE EXPECT ro BE' WHERE WE WET TO WHERE WE A WHERE WE WANT WHERE WE ro BE WHERE WE ARE WHERE we WHERE WHERE WHERE WEPAR WHERE WEP WHERE WE mRE WHERE wc WHERE WE (abs bias) WHERE WE EXPECT ro BE Deviation (SO) Analytical QC Single Accuracy Deviation (SD) Bias Bias Value; Moan (k ) Mean, than True Absolute WHERE ARC WHERE WE ARE t WHERE Analyte -True Value Abs Bias + 1.65*SD (TE /True Value) (TE / True Value) (SD RE Abs Bias % + 1.65* Abs Bias % + 1.65*SD Quality -True Valuel 150 E WE ARE - WEARE RE 68.95-99 WHERE W TE FREWEARE detection but WHERE WE EXPECT TO BE WE ARE We ARE WHERE WE EXPECT TO BE': Systematic WHERE we ARE WHERE WE EXPECT ro BE Quality WHERE WE WHERE . WHERE WE EXPECT WHERE WE EXPECT WHERE Single System Random WHERE WE EXPECT ro WHERE WE ro BE WHERE WE EAPT ro WHERE WE EXPECT ro BE" Vastriaution WHERE WE ARE Analytical True W HERE WE ARE Standard WHERE WE WET TO WHERE WE EXPECT ro BE' Standard *100% RE detection WHERE WE WHERE Error (S E) Gaussian Bias Random 68-95.99 Single have fewer false % Total (SO) Goals (SO) 68-95-99 Gaussian Absolute Trule Population Absolute 145 150 145 - 150 Vastriaution Population detecti' WHERE WE RE WHERE A: value WHERE WE ARE WHERE WE WHERE WC RE A: Bias t WHERE RE RE Random Chavte Gaussian WHERE WE WARE WHERE WE WANT TO WHERE RE TE WHERE Systematic have Single Random Error (TE) Error (RE) Trule Messaje Jeopardy -hue value *cue value value X value X: value mean, SD value A: X Value Value V ariation Jeopardy vValue WHERE WE ARP SD value Vastriaution Value Value WHERE WE. Random Results Bias QUALITY Population RE WHERE Error (REA) WE Single Bias ICY V ariation RE SE (0 / 100% 0+ 1.65 *5 Random + 1.65 *3.3% Gaussian detection but ICY Value; CONTROL Vastrilation Error Error (RE) Viastriaution PATIENT WHERE DO WE WANT TO WHERE DO WE TO WHERE WE WENT TO WHERE WE EXPECT PATIENT WHERE WE WANT TO WHERE WE WARE WHERE WEP WHERE WE? WHERE WE WHERE WE ARE WHERE DO WE WANT True Value; Systematic Bias Analytical 605-99 inconsistent Inconsistent Messages QUALITY = 8.3 = 3.3% Look for Value y' A: mean, A: mean, SD than Value; the chart A: SD value A: *rue value *rue value A: mean A: value Sys tematic A: Value) SAMPLE mean, SD *rue value SD) value mean, SD *rue value X Value Value SD value Value; value SA mean True SE True SD rule Inconsistent System SAMPLE QUALITY abnormal essayes PATIENT ess ayes Messayes CONTROL QUALITY ± 2SD abnormal Pmprecision Error Capability Systematic •slave ICY Messages the IS We Are IS are care Analytical are from SAMPLE Capability % Total Systematic True Value; PATIENT *rue value CONTROL CONTROL QUALITY Error (SE) Messages •ustering Wue He *cue value ; PATIENT Wue from He *rue value Tessayes Single PT) PATIENT use the Gaussian from PATIENT (83.3 / 150) ssages -sages A: 5+ 1.65 *10 5+ 1.65 *5 3.3% + 1.65 *3.4% 145 150 145 - 150 150 Error SAMPLE Error V from Tbnotmal CONTROL worst case scenario (QC) 2SD Random my from True RE -E Error (RE) (QC) coefficient Vastriaution The accuracy SAMPLE clustering QUALITY 25 SAMPLE accuracy an case use scenario Random 1:2s rule as = 13.3 = 6.4% = 3.4% on SAMPLE Wnherent (QC) Coefficient ot Median Mean Value ? is/a€e • is/ace Is/ace . is/ace . is/ace an ±3SD 68% = ICY clustering True Trule — True True Consistent Single Look for Error PATIENT Total CONTRO RE CONTROL Use the a scanning SAMPLE Random ICY True Trule CONT Change Population — True Total co use the Change Normal ICY CONTRO chart es is QUALITY CONTROL PATIENT Total Error (TE) Error PATIENT Value Single Vastriaution X Value Laboratory B Laboratory C SAMPLE f:2s rule as The CONTROL (QC) X Value Value 68% = Change Single Laborato device Population Laboratory A Laboratory C Laboratory D % Total Single T) IRE (QC) b) WE ARE we ARE SE ARE o + 1.65*10 (10 WE ARE PATIENT an Look for : SamFk analysis hrs ses a scanning The accu : analysis has cre : quaff+ahve analysis has Population RE Use the We ARE (0 SAMPLE ± 2SD t ISD QUALITY SAMPLE •slate have Analyte WHERE we Use the Error Total abnormal and = 16.5 ? is(a€e ? is(ace ± 2SD ± ISD 1:2s rule as (QC) Look for WHERE device % Total we ARE CONTROL Total SAMPLE A: Total a scanning RE clustering (QC) 1:2s rule as abnormal Look for Assigned on 95% SAMPLE Use the IS 95% device the chan = the chart 99.7% Nbnormal ± 2SD away from the mean cumula+-ive cumula+ive device a scanning ± 2SD ±3SD clustering Lab B (5 / 100% WHERE DO WE TO WHERE DO WET TO WHERE WE A WHERE WE WANT TO WHERE WE (10 3.3% + 1.65 *3.4% 145 - 150 1450 - 150 1:2s rule as f:2s rule as the mean SAMPLE Look for away from mean Analytical 99.7% a scanning Error an from and WHERE WE = 21.5 = 6.7% = 6.9% -3.3% . is/ace 99.7% z the mean device Error •slate is(ave is/aqe is/ace device clustering ±3SD 99.7% =14.3 150 away from Total Look for ±3SD les es 145 150 Analytical on change fie analy+ical the mean abnormal change fro fie anqly+ical fro the analy+ical to fie analy+ical range fie analy+ical Analytical e analy+ica I Error (TAE) clustering TE RE on sys+em has occuvted. Systematic Sys+-em has occutted. has TE SE RE m has occuned. has occutfed. em has occuvced RE on ARE RE away from us occurted as occurced nas occurved. on on mean % Total the mean and - and and an RE +hen Error 26 25 24 on 24 25

  • total error
  • TE
  • bias
  • imprecision
  • largest variation
  • true value
  • target value
  • 3 key numbers
  • percent
  • SLIPTA
  • SLIPTA checklist
  • mean
  • SD
  • standard deviation
  • total variation
  • eliminate bias
  • reduce imprecision
  • absolute bias
  • cumulative mean