PEPFAR Quality Control and Method Validation Activity 12: Part 1

Are v 'Stribution — ' • 'noution orma/ nor Sh7tts oaussian Normal ' •urmal small; Shifts 'Ml-lab (Sep —all instru oents kr 20fg:2s Gaussian 3s Is LS ) I he or he that contral limits are set Analyt evaluave evalua+e Analytical eval„a+e evaluate 'Analytic a Iype and number or IAnalYt AnalY1 ,LAnalytica• *Analytical 'Analytical' ) t-cequt Systent Norma/ smaller Distribution We Expect rhe that of past abnorma of g:2s to tradeoff tradeoff t." tradeoff th't- that thate Nbnormal tradeoff i. that 3) QC used 3) QC used +0 3) QC ased -ho 3) QC 3) QC +0 Anal: Normal ter smaller Shifts Control limits are set QC ased be greater when systc Smaller Shifts Value; smaller Shifts 5 the tradeoff higher false rejection vrator Th. tradeoff The The tradeoff i'at used +0 Allowable The tradeoff is that The i. that The is that The trodeoH that is control •able Goals The tradeoff i. that The tradeoff that the is that The trade." that i. that Type number Of GC ±es+ed Inter-laboratory Comparison The tradeoff Is that Th. that The tradeoff i' that be Greater when Median The tradeoff is that- Tho tradeoff that tradeoff that The tradeoff The tradeoff i. that I the tradeoff is that The trade-off that Is th•t is that to control limits are set QC + o QC System System Shift$j Shift* - The Is that Th. The trade" that The that the is to the QC used QC used +0 40 The tr.-de-off that closer to to l) Frequency of QC e Mode: Criteria to mean. System you *ill at" have closer to the Type and number of sampks fes+ed to the Shifts Shitts Gaussianb tradeoff you - *Analytical's shitts- man you will also a you also detect larger you will also you will have is Th. •r (TEA) system Allowable measurement m stribution you will have a 1. you wilt ha" to the mean. Definitions of nt you also have a Analytical Totmal Quality rou will also ha" a you will also hm a you have 2) Type number Of QC samples +es+ed false miection System @nalyticar* Nbnormal Cape Clinic Laboratory Moder M Oder Caldufatine you you wiu you also ha. a 2 Type number of sampks tes+ed also have a closet to the q)e of periodic review The Frequency of review you will hay. a you wul haw a jou also have a you also haw Type number sampks +es+ed Type number Of GC +es+ed Normal you will aimo have a you will have a. will you also • you also haw a have a you ha" will also haw Wd' QC ase.d +0 •Analytical' JAnaIyticaI will have you will also have a you will also have will have a will also have a you will haw you also have a you Will will also have The also a also ha. tradeoH Is that closer to the mean. closer to tho mean. 5 of S? All-lab Group — instruments or methods used by you will at" have a you also hay. a you Wii' also have a you have J you will you also a also J y•ou will also have a have The tradeoff that q) The of periodic review for The that thant H) The penodic revtew you a you wilt have a Th. that also haw a tradeoff is that have Peer Group — comprised of other laboratories that use the you have you hay. a you have a you will a you • rou also have a you also IAnaIyticaI -Analytical defec+inj and *rends also haw is that Coo Distribution defecfinj and *rends you have a- also have a higher '*coon used 40 Distribution Analytical Cope i *Analytic'l : Analytic; We Wan l) Frequency (IC even+s Frequency of : Analytica Error (TEA) Omits tall. false eva\aave Error Detection Sys c is ion have a Ana.ys Analytic; 'Anal) AOowable Activity: Calculating SDI and CVI Activity: Calculating SDI •nd CVI also a • Ana I! Ana h Anal) Anal! TEA smallJr sh higher h"' evalua+e Anal: hW reiec Job Aid 1: Com arison of EQA Pro rams Worksheet 1: Calculatin SDI and CVI Job Aid 1 : Com arison of EQA Pro rams smaller Shifts higher false have of or false rejection Value Shifts than The review E TEA higher rejection higher false rejection Widher you have a also have a The of review q) The of periodic review able Gaussian - higher talke System reiecu« also haw closer to the me-an. Error System Imprecision M odea The tradeoff is that Norma/ Need for you haw 'election be greater when smatjet shifts shifts Error (TEA) higher false M Ode: Mode • smaller shifts 'mallet sh Distribution . < TEA < TEA Gaussian Job Aid 2: Inter retin Com arison Re orts Job Aid 2: Inter retin Comparison Re orts 2) Type and number M0de• be 5 of (QC and *rends -l smaller 'Analytical External Quality Assessment be greater of o. Modes smaller shitts nsscssmollt 5 of 5 of a- Mode. Analytical System ISO 15189:2012 - 5.6.3.2 Alternative approaches be greater when Mode• L-J Chart for XYZ Chemisty Analyzer rate Control E < TEA Error ' elec r ate rate 3) QC Smaller shifts ' ate t ate rate Type number Of tes+-ed Results QC used QC Where Laboratory Smaller shiv control are control you will also have a 2) Type M odea limits 14) The of review Systerm: Systemr Mode: Analytical q.ts Goals q) The OF peviodlC review Remove l) frequency of eve l) frequency OF (AC eve I frequency OF TE < TEA ISD Population Pmprecision Normal q) The periodic review used System Yystem syst. Mode: Normat control set Median • to t". Sys SYS Coefficient of Variation Index b) The achons ±Q6en when sys o Aro b) The achons +Qhen when resul+s To false r election -Analytical Analytical peciodic review for Smaller shifts: Distribution Laboratoryss Fffective QC The OF peciodlC review of peviodlC review of review Analyt Allowable closer to the I) Frequency o; (AC Analytical The of peviodic revtew I Frequency OF Distribution The OF periodic review for Frequency +es± control limits are set control limits an set describing thi Definitions of qualitative terms Anitions of qua Quality rate control limits are set lox Mode- false rejection describing Goals used laboratory. closer the mean. to the The of periodic review Laborator Ana Median Quality Frequency of pevtod.c Tevtew Laboratories. Median = Goals Goats to evalua+e 3) used Th. tradeoff that higher false re false Total Error Analyte: Glucose Control Material: Levell Control Levell Control Material: Level 1 closer to QC Units: nunoVL Units: wunoVL higher false reiection Analyte: defec±inj and +rends control limits smaller sh small;r Smaller defec+inj and *rends closer to the mean. I SD Population Consensus Based Metrics Standard l) : Analytical Normal l) Frequency you win a When starting your review, begin by asking yourself the following questions: An; 1 SD Population Wo Aro and *rends Type number of system smaller System pertcd,c revtew Laboratory participating laboratories to measure an analyte Internal QC— a single laboratory analyzing controls and making immediate internal decisions on the acceptance or rejection of analytical runs. IQC provides Err-or detection I. Calculate the CVI for each control. Th. that q) The of periodic review l) Frequency (AC +e• ased +0 Modee Error True Value; Frequency o;C characterist q) The of pevtodjc review same or similar instruments and methods. cot (C defecåinj and *rends defec±inj and Median = Median you also have a the is that higher false to number Frequency -fest Laboratoryy Mode— Mode: q) The of review defecåinj and +rends defec+inj and +cends Distribution Mode. Median System Mode = Programs 68-95.99 you a system Quality Mode- and +cends Modes M0de• Mode• - refection The of revtew for J Type number Mode: rate Normal Results evalaa+e ...9 Sing 10 System- System q) The of periodic review for System (SO) True 1) Type q) The Frequency OF periodic review q) The OF review System b Goals UAnalyticai Iloweable M Odea We Are System cedian Laboratory be greater when q) The Frequency of review The of peniodlC review Iowab'e TEA ISO 15189:2012 - 5.6.2.2 Quality control materials Totmal be when Laboratory. Cri Criteria ider control trd The of periodic review for Laborator Distribution M ode: Omits The tradeoff is that evo The tradeoff is that 0/0 Accuracy Mode* 01 Laboratory Type aod number acho 'Wable Median Quality . < TEA TE < Mode• Where- i der cont able Analytical Laboratory lox Wider q) The of pevtodic review The of revtew < TEA Goals Widher q) The OF periodic review for Results Median Wider The of periodic review The OF review for q) The of peviodic review The of review The of penodlC review Outliers control I'mits Whenever an interlaboratory comparison is not available, Meoian TEA te the Distribution Lot#: 3-123 Lot #: Lot 3-123 3-123 Exp Date: 1914120M assign 3.48 SD aSgn 19/4/20XX assign 3.48 SD aSgn 19/4/20*X assign 3.48 SD asSgn number 0.05 Results Type and number Median = Platelet Count (cells x 109/L) Where' - Laborator Normal 2) Type number Type anid numbec M Ode. Median • Median z Wider cont •ontrol Wider control What mean value should my laboratory obtain for this control? What mean value should my laboratory obtair a good estimate of the imprecision of the method by reviewing SD and %CV. Wider contr Wider control cont Wider cont Error SD Population Medan Chan shifts Precision e Wa Quality used used +0 Trueness Mean control are limits are Wider con Wider contro x assig cont Normal Wider cont"' Wider eon ontro ID OF periodic review q) The of control limits are Wider con q) The review defecfinj and +rends Where- Where - Where"' greater *hen are Wider co Widher you doa QC ase b) a Towable ATE) We Expect Need for What Wii" stable limits are less w.i!l Smaller use 70 Median Look for 5) The ac+ions when b) The achons +Q6en when resul+s b) The ad-ions when resql+s b) The achons when .2 Results you will also have a con pevtodtc review 3) tal Error al Error Goals I Result Purpose Puruose re Criteria Cri b) The achons +aKen when resul+s you will also have a Goals Mode: Median • ro Be Quality Wider re less b) The achons ±Qhen when we Are Wider con f'•.2s We Are is Goals e achons ±Q6en when resul+s The achons ±Q6en when laboratory. The Goals Wider c The of pevto,djc review The achons ±ahen when resul+s The achons ±a6en when and *rends Laboratory Resules Wnherent Where Where- that Wider CO less con limits System Lab. exceed OQCQep+Q61e Wider co less Goals limits are less cor limits arc less achons e. no and +rends Laboratory. Labora 68% Population I limits arc less (Laborator Quality TE < TEA Where I higher false rejection Laboratory Mean Where- Wnherent Where closet to the mean. Be *'def co ts are less Wider con e achons ±Q16en when Median to the Need achons +Q6en when resul+s Laborator less .2 Laborator ua it) Lab' cean Median = cCan chons when resul+s The OF revtew For QC limits less Quality Distribution We Expect Wo Expect We Ex Expect wo Expect Laboratory Laboratorf Rate derec± Laboratory. Need for Look for Need tor Need 'or Quality The Accuracy; limits are Wider less defecfinj and *rends defecåinj and +rends defec\inj and *rends defec±inj and +rends The revtew Stable Results e Wa limits are exceed itive to error Random C rite". Wider limits less oals limits are tess higher false reiection Labc I SD Population limits are less defec+lnj and (ATE) cean TE < TEA Distribution e Wan fo error Goals and +rends < TEA higher false ua tat Error be greater when Wider less sensitive to err sensitive to c Mean Where Need ive to Sensitive to defec+inj and A-rends Where- From: June 1, 20XX From: 20XX Through: ou h: Target V alue 6.90% To Be Juw30, June 30, ive to er We Are we Are Results 20U sensi We Ex pect The tradeoff is that The that tradeoff that Schemes sensitive to error' limits limits are less Analytical Analytical Analytical Are iLaboratory Normal Control to error High Control Wider control Laborator Goals Low Control to c Wider What SD or CV do most laboratories achieve with this method? What SD or CV do most laboratories achieve Ate) Sysmex XE- Sys limits are less sensitive to error, tive to e achons ±ahen when Results 10 Median Single sensitive to e sensitive to er sensitive to e Re sults System Results The laboratory shall use quality control materials that react to the ac OC sensitive to crror sensitive to sensitive to error itive to CC the e toe Single 5 o_ Total as Wider co less to e Re Precision sensitive to c uali are less limits less TO Be Goals limits defecfinj and defec+inj and +rends Result± Sing le teed b) Wider con Laborator def and +rends d±ec+inj and *rends defecflnj and +cends defec+lnj and *rends Gaussian Wider con limits def cåinj and *rends All-lab CV External QC— a program that involves many laboratories analyzing the same samples (control materials or PT samples). This allows individual laboratories Activity: Calculating SDI and CVI Activity, Calculating SDI and CVI Activity; Calculating SDI and CVI and *rends Laboratory. Laborator Goals System Jess q) SD Population to ons ±QKen when resul+s and ua ity sensitive to itive to err sensitive to DS +Q6en when resul+s and *rends Res (ATE) •am that involves samples (controi materia[ ore True Value; limits are The achons +Q6en when External Quality Assessment d and +rends Laborator limits describing the performance cribing the Laboratoryes. Ss +Q6Ken when resul+s Goals and *rends Analytical ahKen when resul+s detect larwr Goats Accuracy Single Wider cont def c+in and *rends ns when S when Accuracy of Laborator Accuracy, Accuracy; We Are Need sensitive to Definitions of qualitative terms +Qhen when resul+s able you also haw a you will also have a you will have a S +QKhen when resul+s Accuracy of measu Laboratory. Mode: you also Wider control de+ec.+in and de. and achons +ahen ' and The rev, (ATE) SD Coefficient ot Varif Coefficient ot Coefficient of Coefficient ot Varial ction but Coefficient ot Variation Index + ±Q6en when resul+s Quality evalua+e Rest Mean letection but {Q hen when resul+s def 3 wab'e Trueness r acy 68-95.99 aut ±Q6en when resu14-s +Q6en when resul+s ±Q6en when ±Q6en wh Laboratory Standard the the laboratory shall develop other approaches and limits are less detection but b) The achons ±ahen when b) The achons +Q14en when resul+s b) The achons +Qhen when q) The peood,c rev System System Allowable detection but b) The achons +Q6en when and +cends ua Re Population Wo Expect The tradeoff that Results b) The achons +Q6en when resul+s 68% Deviation (SD) detect Allowable System e Wa Wider control less Precision Pmprecision precision Sys tom Using simple math and working in Using simple math working Using simple math and workjng in, detection but • ±Q6en when resul+s The r•evaew The rev 2SD Analytical ess We Ex CCD to but Re b) The achons resul+s Total Using Calculate 1ÄnaIyti to (Target Value) OC - sensitive to Results 2012 Laboratory do? Population Results Wider ISO SD . s 62 review Tbe limits Where— l) Frequency e l) l) Frequency l) Frequency • false e achons +Qhen when resul+s achons +ahen when Chart Become familiar with how to 210Become familiar with how to chons ±Qhen when nhons ±Qhen when resul+s also 'hilts than I SD Population 68% = Become åmiliar with how to limits are less ) Frequency control Purpose Puruose chons when Hons ±Qhen when resul+s achons when resul+s 2140 710.0 Wider cont exceed ±Q6en when resal+s Frequency l) Frequency higher Frequency - detection but exceed QQQep+Q61e exceed Results Accuracy Accuracy. System to error 'have fewer false have fewer false I limits are less exceed QQcep+Q61e exceed QCCep%åle Results occep+oåle QQQep+Q61e Wider control accep+QåIe QCQep+061e b) The ad-tons +ahen when cesul+s q) OF re'. Ereor Mode: Median limits are less have fewer false have fewer false SDI (Standard Deviation Index, a.k.a. z-score) — the calculated value that indicates the number of SD units b) The achons +ahen when SDI (Standard Deviation Index, a.k.a. z-score) — the calculated value b) The achons +QKen ualit q) The rev, q) pectodvc rev al Error have false fewer false er false? vou when resul+s when Results ISD limits have have (ewer false Shave fewer false have fewer false. - have fewer false, falso false 5) The b) The achons ±ahen when resul+s b) The achons when QCQep+Q61e accep+oble ions when resul+s ResulIts seasiti« to that Quality detec b) Error (TE) Your laboratory enrolls in a inter-laboratory YourAåaboratory enrolls in a inter-laboratory false false. false b) The achons fahen when resul+s b) The achons +a6en when QQCep+aåIe Results *80/0 = Median a to compare their performance to that of a peer group. EQC is useful for estimating the bias or inaccuracy of a method. detection but detection bet have fewer Hh-ions +Qhen when resul+s Hons ±Qhen when 680/0 = Total Error Trueness Precision PEPFAR QC Workshop y) Frequency fewer What VOu What What vou Quality ua uali 'Results less 8hiftz than The *etuency The rev, 3/10 80/0 = < TEA To S when resul+s os when resulA-s ua It 68% Frequency fewer false achons +QHen YO = Yo Yo = 580/0 = Assigned On sensitive to error i8% = i80/o = i8% i80/o Population System Normat rule 'have fewer false _ sensitive to Goals ua 1 have have tewer Sing le Distribution sensitive to Wider limits are The achons +QKen when resul+s achons +ahen when resul+s achons ±ahen when achons fa16en when resul+s CA-ions +Q64en when resul+s chons ±ahen when ons +Qhen when resul+s have fewer false Change fewer ualit Single system to fewer have fewer false ns +a14en when resul+s achons when .4 Laboratory Labora less 3) Type TO •l rejections fewer ±QKen when resul+s Schemes rate ± I SD Population Results 68% = ,examining system in a manner as close as possible to patient samples. examining system in a manner as close as possible to patient samples. sensiti«e to e amples. 68% = detection but limits are less and Sys tel Single A ave fewer false rejections have tandom ± 1 SD Population have fewer Wider con Median . rejections smatter Mean m d acu:i detections rejections - Rejection Goals r ejections rejections . telections •ions of each result from the group's mean; describes the bias of a method in units of SD. SDI is expressed as — I_SE_ I_SE_ Res ± ISD exceed aecep+oåle iccep+061e zep+oåle Laboratory Goals detection but but -p+üle •re less We Are Mode: M Odea dept-able '-pi-ruble ali limits less Your Lab detection number sensitive to erroiu p+üle Wider less Control. Qüåfåt&SDl and -CVl—EVåluate detection but — exceed QCQep+oåIe exceed 56 rejections 214 2100 545 Omits are are åir , ten Trueness Seteetion but WHERE WE EXPECr BE WHERE WE ARE pairs, QQCep+061e ± 2SD limits Results . ISD calculate SDI and CVI. Evaluate calcu e SDI and CVI. Evalculate ep+üle llfnl+s exceed — SE limits limits are less pi-able ua lty and 950/0 Assigned on sensitive to but ua SE error ep+oåle p+oåle p+Q61e fewer false SD Population Accuracy, timßs less Mode: WE ARE The achons +ohen Laboratory telec achons b) The achons fahen when resa ISD Wider Results b) The achons SD SS Single Wider exceed QQCep+oåIe exceed qccep+oåle exceed accept-ruble exceed occep+oåle -d Wider control SE ualit 50/0 = 0/05 Population exceed QQCep+061e limi+s exceed acceptable exceed exceed QQCep+oåle limi+s ed qccep+oåle ed d QQQep+oåIe limi+s Capability Results SDI and CVI *Ste Inherent Wide•r Wider provide objective evidence for determining the acceptability detection qccepiüle limits are qccep+üble either a positive or negative value, indicating whether your result is either a positive or negative value, indicating Control to characteristics of a measurement either a positive or negative value, indicating whether your result is above or below the group's mean. For (ATE) Pop Stab'. WHERE WE EXPEC ualit'. ± ISD ± 2SD ualit) detection but Results b) The achons +aKen inconsistent Total Error Comparison Programs Comparison of EQA Programs exceed ua It rror (RE) Assigned on ed limi+s pi-ruble limi+s sensitive to Random to to err" sensitive Standard Deviation Index have limits are less Eimits less Single less •have fewer false, Con limits are jess are less to fess b) ach eed accept-ruble b) The achons ±aHen alit Population the Chart or ±3SD Wider less less program offered by your QC manufacturer. CVI Calculation CVI Calculation SDI and b) The achons faHen uall 95% = limits less limits 're less limits are less the chart = ua ua 1 limits arc the chart g exceed QQCep+061e 0/0 the chart Detections exceed QCQep+rÆle exceed OCCep+061e limits detection but have fewer false exceed qccep+061e exceed QCQep+QbIe exceed OQCQep+06(e exceed QQQep+Q61e licni+s exceed QCcep+06ie limi+s exceed QQQep+061e eed geed QQQep+rÆIe -eed QCCep+obIe ed QQQep+061e limi+s ed ed QQQep+obIe Results 68% Total Error Wider con ua detection but SD 12 QCQep+Q61e pi-able . ISD QQQep+061e OQQep+rÆIe uali Assigned on QCCep+061e acceptable the Chan Median : Median = Gaussian uali exceed uali ua It'. 680/0 680/0 = bet error Detection Capability have fewer false j ARE —7 inter-laboratory comparison reports, the SDI is a statistic that indicates the relationship between your yste 3.63 +3 3.638 limits are less ualit ewer tatse Assigned 68-95-99 but ualit Assigned on te error ua 1 R:qs I SD Populati« characteristics of a measurement Accuracy Wider ua systen Population 68% = to the chart : your method's performance.in your method's performance in your performange in performance 68% = Quality sensitive to the Chart the chart = Quality 68% to Quality S Quality have fewer Cap.abåUty WHERE WE ARE Proficiency Testing (PT) Accur acv of measurement of measurement limits are less System the chart z Wider control control Accuracy of measurement Inter-laboratory Comparison Control ISD 1:2s ru (ATE) limo •re less Error (TAE) Mean Median = q; ua Control WHERE WE EXPECT ro BE WHERE WE ARE WHERE WE WHERE WE ARE rejections uatlt Control Cont ± ISD limns Single your m t. ontrc Capability ( the chart • WHERE WE EXPECr ro WHERE WE ARE TE < TEA the chart Change Use the information supplied on '7dJse the information su plied on Use the information suppiied on . WE BE WHERE WEARE fewer Wider con Use the information suppliednon RE WXPECT ro WHERE WE ARE HERE WE ro WHERE WE ARE E EXPECT ro BE BE WHERE WE ARE Quality ua 1 Qualit: fewer Need for Ota/ detection but the chart Quality Control ua It) Con laboratonies. detection but CVI Calculation Laboratoryes. laboratoryes. Laboratories. Lab's Monthly CV detections —WHERE WE EXPECT ro BE Labo detections Laboratory. Single Assigned on the chart z WHERE mean value and the results reported by your group. Mea ontrol talk detectio' detection (ATE) Mean 7.9 rule Qualit ali (P cv Goals detection WHERE control so ualit EXPECr WHERE WE ua It WHERE WE EXPECT ro BE , 68% detection. Goals EXPECT ro BE —WHERE WE EXPECT ro BE WHERE WE EXPECT ro BE., Assignor. Quality (ATEA) PROFICIENCY EXPECr ro BE error ass€gnee ro BE TO BE ua 1 ua Laboratory WHERE WE method on an on-go;" 68% Capability Assigne Assign' . assi€ne Assigne•q Assi Assign', Description Distribution Laboratory Many laboratories analyzing PT samples from the same pool limits are less Many laboratories analyzing the same lot number of control WHERE WE EXPECT ro Assigned Goals WHERE WE EXPECT ro BE Observed he Assigned on YSte As TESTING ua i ua it ualit) Single Population foyer false fewer false of examination results. '00k for for v. need) Cont 350/0 ts oals ts llts (SDI) Index (CVI) Results Assign'. • Its its 'n ualit (Analytic Quality System WHERE we Observed uall ua it Enrollment Error Detection t ion Bias Absolute WHERE WE EXPECT ro WHERE WE ARE HERE WE ro BE WHERE WE ARE WE ARE yste Use ti 95% error Detection the chart ua ity Goals ua it 1:2s rule as Look for WHERE WE ro (Pa for detection but Err-or Observed fewer ua It ua It Look for . k for WE ARE Detections Assigned *00k for WHERE h • E WEARE sensitive to ua it Laboratory relationship to the presence or re.at•onship to the presence or Population WHERE WHERE WE ro BE , 80/0 Need for corr forr Need for WHEWE WHERE m k for WHERE We WHERE hE What vou need) What need) What vou need? Control systenr 50/0 950/0 error the chart z S rule . for for need? ua 1 test Error What OC the chart : "0k for EXPECr ro BE WHERE WE EXPECT ro BE conditi0's 50/0 = Single ua it o/ WHERE WE EXPECT ro WHERE WE ARE 9.7% = WHERE WE t WHERE ARE :crroBE sensitive to error ts ni or the ac u ac WHERE WE e, WHEkE WE EXPECr ro BE WHERE WE ARE Control Results Quality the Chan Control to • Monitor the accuracy of your Monitor the accuracy of your Accuracy WHERE WE EXPECT ro DE WHERE WE ARE of we me EO BE WHERE WE ARE easons the chart My Lab WHERE WE EXPECT ro BE WHERE WE ARE WE E WE ARE WHERE we Control the chart = the chart Your laboratory, along with other laboratories, WHERE WE We Are ARE detection but ua 1 the chy Its chart submit their results for each sample to the PT provider s mit resui•s sampie sensitive to er Jse the material submit monthly summary statistics (Mean, SD, number of WE EXPEcr ro BE WHERE WE ARE Truoness ua 1 68% = Capability ( (P the c Quality control materials shall be periodically examined with a 1. 23 24 its the chart t, Look for EAPECr ro BE WHERE WE ARE Ecr ro BE since your vaj. ua 95.70/0 Capability (P a) Your sensitive to WHERE WE ro WHERE WE ARE >yster ISD Capability (P„) (Pa) What Oc at hors cc&rar WHERE WE EXPECT ro BE , .1 Capability (P — 95% = capability z. Quality device Pro Cap-ability (P Random Ise the Look for Assigned on achons Worksheet 1 to calculate your Worksh et 1 to calculate yo r Worksheetl&o calculate your WHERE WE EXPECT ro WHERE WE ro DE WE EXPECT ro BE WHERE WE EXPECT ro ,E WHERE WE EXPECT ro BE. ard Capability Assigned on thet •bnorma —a hen to Ken a Ken Ideally, 10 since your ork Ideally, 1.0, your vaue Ideally, CVI s•nce your va Ideally, CVI sance your va Ideally, CVI sjsince your vai. Ideally, CVI s•nce your va•e Ideally, your LOOK for Ideally. your Ideally, CVI s•nce your vale Ideally, I O your van Ideally, C VIS 1.0, stnce your Ideally, CVI since your Ideally. Ideally, CVIS since your van Ideally, CVIS stnce your vane Ideally, CVI sance your van Ideally, CVI since your van: Ideally, CVI since your vie Ideally, CVI 10, sunce your Ideally, CVI since your vaue Ideally. C VI 1.0, s•nce your lly, CVIS 1.0, your va•e sensitive to error E WE ARE Q WE ARE Assi-ned sensitive to error ua - Y WE EXPECT ro WHERE WE ARE Population jbnorma mea 'E WE A RE ontrol SDI = (lab mean — group mean)/group SD your va•e WHERE WE A4RE WE PATIENT The WHERE WE EXPECr ro WHERE WE ARE ua i we SDI = (lab mean — group mean SDI = (lab me WHER Quality Vastriaution . 1:2s rule as "se the VHERE 68% the the WHERE WE EXPECr ro BE WHERE WE ARE WHERE HERE ro BE WHERE WE ARE 'HERE WE EXPEcr ro WHERE WE ARE ERE WE EXPECT ro BE WHERE WE ARE E WE Controlå 50/0 = Error ' 1:2s rule as Look for WHERE WE EXPECT TO BE WHERE WE ARE WE WE ARE 1:2s ru thate fohen Nbnormal Compare your results Laborator 1:2s ru detection but abnorma Normal E WE BE WHERE WE ARE have fewer false WHERE WE EXPECT ro EE ontrol chart abn åma •ma Group D ma ro BE WHERE WE ARE BE WHERE WE ARE WHERE WE ro BE , WHERE WE ARE WHERE we ARE PROFICIENCY Bias ass Compare vou WHERE WEARE System Control -e the abnorma abnorma •N. BE WHERE ua ua 1 Observed ua e the Labora 6. 'Ecr scr ro BE me uality Observed Use the use the the hither Nabnorma Compare ycur results PECT ro BE ro BE Control( ualit uali ISD Coefficient Of UJse the that jbnorma •ma control measurements) to a central data base ave fewer false TESTING (PT) WHERE WE ARE 95% •ewe-f Use the e the BE have fewer false PROFI 2SD Precision while the peer CV •s PATIENT 'e as preCIS10n WE ARE WHERE WE EXPECT ro ,E v WHERE WE EXPECT ro BE WHERE WEARE WHERE Con Yo = 68% Single lab, while WHERE WE EXPECT WHERE WE ro BE WHERE WE ro ua ity Contro absence bf SDI WHEARE, WHEW —WHERE WE EXPECT ro ,E WHERE WE EXPECT ro BE raise Rejection Single Control ua ualii Trueness Control YS the chart Total Error N WHA we TESTING TESTING i80/o = absence of SDI or CVI flags. 1:2s ua It Inconsistent absence fewer false Capability 680/0 = abnorma single lab, while the peer CV 2 WHEP t WHERE WHERE WE ARE BARE single lab, while the peer CV Single lab, while the peer CV single lab, while the peer CV •s -e lab, while the peer CV control TESTING Control:. Control Assigned on 68% = i Inherent alusterin Look for Need for WHERE WE ARE WHERE ua 1 99.7% Rat. charte Error Peer Monthly CV rrna WHERE we WHERE WE ARE ± ISD the h•ve fewer false Stable lusterin ts E WE ARE 0.03 the chart • fewer tatse wr Control ysterge•kk ntr Assigned on Assigned on — Assigned nd Concentrol 'lusterin •in •n ARE QUALITY rate error Detection Accuracy of measurement - Closeness of agreement ule as %Isterin A Re c 3.58 3.58 68% = Ise the TE < TEA 'ass -•:lusterin in . .1:2s rule as ass pule as ccuracv clusterin :lusterin rin WHERE WE ARE WE ARE core 680/0 • •masterin TESTI 950/0 = Use t R Ate as Äin •este Whenever possible, this mechanism shall utilize appropriate ontrol js rule as •lusterin abnorma WF Control? Data Submitted Results obtained from testing provided PT samples. -.usterin clusterin lusterin Monthly summary statistics calculated from routine QC testing as _ ) 1 WHERE WE ARE 95% = used Use the Observed Control. Control detections aim fewer false a scan Ac •gracy Single Sin way fr t:2s rule as laboratory's SDI and CVI. the same System PROFICIENCY TESTING (PT) a scan Capability fre is baG on the stability of the procedljfé'ifid the risk scanning Look for laboratory s SDI and CVI. iaoora W HERE WEARE SDI and CV;. Change analyze the same lot number of control materials O'2s rule as 2s rule as , ( WHERE WEARE • Peer-based measure of Control 1H:2S rule as rule as e?way fro Ale as '(e as f! 95% = 4 •.2s rule as •ule as Laboratory a oratory s PROFICIENCY frequency that is based on the stability of the procedure and the risk M As rule as •1:2s rule as '1:2s rule as '2s rule as Is rule as •ay fro rule as isk 2SD 95% = 1:29 as Systerr ysü :2as rule as Laboratoryes. Laboratoryes. RE PATIENT PROFIC 95% S 0.05 SS• Assigned on the chart My Lab or 'Ste r atones. SAMPLE average value average PROFI the cha gay fro LOOK for S O.OS System System "'fro Control yste QUALITY I •STING WHERE WE ARE IS in CVI (Coefficient Of Variation Index, a.k.a CVR) — the comparison or ratio of the laboratory's CV to the CVI (Coefficient of Variation Index, a.k.a CVR) — the comparison Or ratio Of the laboratorV's CV to the CVI (Coefficient Of Variation Index, a.k.a CVR) — the comparison Or ratio Of the laboratory's CV to the CVI (Coefficient of Variation Index, a.k.a CVR) — the comparison Or pATIENT i WHERE WE ARE B WHERE WE ARE Error Detection •s rule as WHERE WE ARE CONTROL PATIENT , the WHERE ystem PROFICIENCY SAI 950/0 1:2s rule as TESTING ( aim 0.05 SS V fro abnorma Zookfo 0.05 SS Population jaway fro per your laboratory's QC protocol schedule Rate Vastriaution Must be analyzed in the same manner as patient testing. Population@ vay fro IY fro fro co, use annin away fro *stem SAMPLE Capability (P„) 'he m fro TESTING (PT) SA Lim tot S or 680/0 TESTING syste 'ay fro 'way fro Syste 'i•.2s rule as Used Ascanning PROF Observed TESTING (PT) S O,OS or S•• vstem y fro yste -canning se t ISD ru xe the Nbnormal Laborator ar analvte analvte TESTING Popu tanning 'se the PROFI ± ISD ± •ISD • .neependert be SAMPLE tcanning •tnning abnorma PATIENT • should CONTROL Use t ± ISD 0.05 Use anning 'Ing Use the nning SS the chart z .A.2sruleas s anning scanning nning Activity 12: *lvice Assigned on rule as a scanning canning ng mg nea 99.7% = yste (Peer) SDI group's CV. QUALITY 99.7% false Rejection LOS SS away fro met' TESTI Error (SE) lhe ea TE TEA method for a particular analyte • .neependent S O.OS SS Syste YSte False have System System Change PROFICIENCY TESTING (PT) System 99.7% - "m S 0.05 R election S 0.05 S or S" stem Allowable canning abnorma Trueness . Trueness vour lab 50-100% less precise SAMPLE and an accept TESTING (F- S Or Systeyn • Receive and analyze 1-5 samples per discipline. Single orueness clusterin Fats. method on an on-going basis . scanning qusterin the mea T) 70/0 S SS ACTIVITY: TESTING (PT) PROF SAMPLE TESTING (F TEST' SAMP SAI asis 70/6 'evice TESTING ISD onea the mea PROFI S O.OS e mea - lusterin ,mea mea nean If SAMPLE Total Error Complete Worksheet 1 to evaluate a scan S OAS SS ice away mea 5 to 20, your lab 50-100% less precise If CVI : 1.5 to 20, your lab is 50-100% less precise If to 20, your lab is SO. 100% less precise Syste aluate Observed for the month device the mea if CVI to 20, lab is S If to 20, lab is SO mean - 1.5 to 20, your lab is S , 5 to 20, your lab is 5 to 20, your lab is 5 20, your lab is 5 If to 20, lab 50-100% less precise If to 20, lab is 50-100% less precise If CVI : 1.5 to 20, your lab is less precise If to 20, your lab is less precise If CVI 1.5 to 2 0, your lab is 50-100% less precise If CVI : to 20, your lab is 50-100% less precise If CVI :.5 to 2.0, your lab is 50-100% precise If CVI : 1.5 to 2.0, your lab less precise If CVI : 1.5 to 2.0, your lab is 50-100% less precise If CVI 1.5 to 20, your lab is 50-100% less precise If to 20, lab 50-100% precise System = 1.5 to 20, your lab is less precise •our lab is 50-100% less precise QUALITY Rhante 99.7% = ea Mean QUALITY yste have Ret e rate of harm to the patient from an erroneous result. materials. uste bias Total er PROFICIENCY Rate between a quantity value obtained by measurement and rule as sevice value Pmprecision imprecision cons' de reg cons. de reg TESTING CONTROL PATIENT S OOS SS The S 0.05 'ice essayes QUALITY PATIENT 1:2s ru ru SAMPLE Population Ise t! use t! Use t apply Oc device ass 95 % scanning Nevice vice ±3SD Samples may require special handling and reconstitution. lice an erroneou- TESTING \evice limits '1:2s rule as -pop SAMPLE aim S 0.05 99.7% Coefficient of • 11 P ROFICIENCY QUALITY usually reqwnng TESTING (PT) ± ISD Understanding Inter -laboratory SA ±3SD clusterin understanding Inter-laboratory TEST S 0.05 or SS Mean ISD I3SD away fro evi What QC S 0.08 fro S 0.05 than its PROFICIENCY CONTR PROFICIEI PROF 99.7% vs lway fro •eer group, usually requ•ring than its peer group, usually requiring peer group, usually requiring PATIENT TESTING (PT) peer group, usually re eer group, usually re er group, usually re than its peer group, usually requjring than its peer group, usually requ.ring than Its peer group, usually requtr•ng than its peer group, usually requtnng 's peer group, usually requiring CONTROL PROFICIENCY QUALITY PROFICIENCY i; CVI = lab CV/group CV CVI lab CV/group CV ±3SD PROFICIENCY reqwnng QUALITY PATIENT Messag S 0.05 aim S 0.0S dev PROFICIENCY TESTING QUALITY aim 0.03 or 0.03 aim SS aim 0.05 or SS S O.OS or SS What will you need? What will you needo? SAMPLE tot 0.05 S OAS S 0.0S or S 0.05 S•• S or SS 005 or SS need. 62 us 0.05 SS S 0.05 or S OOS S ODS S 0.05 SS OAS o' 0/05 0.05 SOS Messages PATIENT aim S 0.0S or Lim S 0.03 ot O.OS o r 3.53 LAB C nning QUAVI SAMP aim S 0.03 or SS aim S OSS or ss aim lot S aim S 0.0S ot S•• S OOS e S 0.0S 0.0B O.OS ODS S OSS SS S SS ODS SS 0.05 or or SS 2-4 events a year -refer to PT testing schedule "ning TESTING aim S aim S 0.03 or Lim S 0.0 99.7% = Frequency monthly use t! aim S o. S or 0.0S OOS but PATIENT ODS PATIENT CONTROL TESTING (PT) .ca to HOW 1 ill 11 Assigned on rule as SD QUALITY QUALITY PROFICIENCY Vastriaution Q' ru 1:2s ru I•.2s ru an Consistent mea your mnethod's performance.in your met o 's per ormance.in your method's performance in your method's performance. as ru 99.7% = 99.7% Control TESTING (PT) TESTI PROFICIENCY TESTING ru •ING (PT) TING (PT) PROFICIENCY TEST. C Dmpaaison Programs fro away fro invest! TESTING (PT) 95% = PATIENT PROFICIENCY Error Detection can QUA • a scanning peer Peer V fro (QC) investigation. (QC) S 0.03 S• False Mean Median atlon. tigat•on. oat•on. Peer Group PROFICIE 62 the mea ation. awa it•on. the mea 496 95% S SS PATIENT What OC CONTROL CONTROL TESTING (PT) Cd, bra em (PT) CONTROL (QC) PATIENT False Rejecti coy • ISD CONTROL PATIENT Values of Sample Not known at time of testing CON' the chart Assigned on easons Why appear Gaussian Disfr NT co SAMPLE CONTR easons bæs Protein. Total, Serum essayes I3SD the Chart the chart = Messages CONT' Known at time of testing - the new control measurements should Change Z-score 11 r 111 QUALITY eas:ns NOTE Examples of such materials may include: 68% = devi Ot SS QUALITY accuracy Messaye the true value of the measurand PATIENT Error ice 'less ages SAI Your laboratory submits your QC results through a 95% = 950/0 submits your QC results through a levice SD PROFICIENTY NT CON the S 0.09 1. 33 TESTING (PT) (u 99.7% = bes 0' easens bces nglO ±3SD • Describes how far our Capability QUA' Comparison of our Comparison Programs TESTING (PT) SAMPLE aim SO.C SAMPLE a scan PROFICIENCY What OC Appear Ga Ideally, CVI 1.0, since your values are from a .E S'. LE .E CONTROL 111 950/0 aim tor S 0.05 or SS TESTING (PT) TESTING a sca sca S 0.03 % age s SAMPLE 1:2s ru easons Why aim S O Cd, bra QUA'I eau ret. kinetc eqes •ROL TESTING (PT) easens Why He schedule O.OS "s ru S 0.05 or Capability capability QUA' SAMPLE OC Observed apply the mea PROFICIENCY rules to How many apply QC SD 2SD 19 N — the number of participating laboratories identified in the report 40 PAT' show the same distribution as the past control measurements if the rubes to How many easens Why He bass to HOW many PR device ey Appear Gaus, Why TESTING CPT) PATIENT CONTROL analytic System (QC) What OC rulesS to SAMPLE ± ISD TESTING PATIENT t ru • Worksheet 1: Calculating SDI (QC) (QC) ±3SD Use t! • Works et 1: alculat• SAW, . -o other— SA Obs SAMPLE Single ess s QUALITY appear Gaussian bi9f,bu+ed Appe.ot cat, ± ISD VARIATION Verify your responses with the pair Verify your respor»es with the pair rules to re •nses with the pair SAMPIE $ ODS Da+a PRE. 11 easons Why D BIOLOGICAL SAMPLE SAMPLE CONTROL devi device PROFICIENCY Why Da+a bæs CONTROL apply QC TESTING (PT) {les dev sAMPLE r Esr1NG ru easens Why Gaussian bis+ftbu+ Use t Tbe acueacy use t gppeor G roup F PROFICIENCY accueacy appt y QC aim C Why ± ISD ± 2SD Gaussian bütbu+ed Compare your results to other SAMPLE SAMPLE What OC What QC eascns Why Understanding Inter-laboratory CONT system is stable. NOTE 1 The laboratory should choose concentrations of control SAN 99.7% = easons W/hy Boes cow.'" CON' TESTING coy cor- cor sca Son an PATIENT PAT' Obse. • PATIENT to 95% PATIENT PAT' 10.7 11.3 CONTRO' 8.8 CONTROL Use t' a scan ±3SD OC eascns TESTING (PT) TESTING (P T) TESTING co The accuracy apply He Da+aboes QC data management program to a central facility. sdCLåatå management program to a central facility. cop Tie The He fm S 0.05 Gaussanbisfr,bu+ed _f:æ Tie (QC) Messayes 104U00 rulesS to TESTING — certified reference materials; Yd How SAMPLE PATIENT mean is from the peer or on / laboratory's CV to the SDI TEST CVI HOW Why -He 1:2s ru on on - eas:ns -He accuracy 3.48 BIOLOGICAL rubes to Reasons Why -He t. Tie and CVI What OC OC ac Appear Gaussian S 0.0 ANALYTICAL na TESTING (PT) SAMPLE rules to to aussian ppeor Gaussia QC CA! Ma PRE. PRE- Key messages messa3éS 1:2s ru essayes less ages Provider Accredited PT provider -ANALYTICAL y me. Mess/ Manufacturers of QC materials Manufacturers of QC messaqes messages Messages He SD ru fm CNAL located closest to you. located closest to v . esc aye S -He N VARIATION s/ ersayes ess to easons sages ssag€s SAMPLE ppeor Gaussian b. single lab, while the peer CV is from several "ICAL s09em 685.0 5620.0 osest to y QC samples Gaussian bis+fibu+ ANALYTICAL BIOLOGICAL CAL eotu 17 rules to How many ppeor Gaussianbis+fibu+ec How many Gaussianbis+r;bu+ed o. ts CK ppeor GaussianbisfrF• ere ANALYTIC ANALYTICAL/ ANALYTICAL ANALYTICAL, em 1:2s rul SD St) anal'/ ice PRE. PRE- K St) Mean (QC) appt' appt y Be T • 2 SD ± 2SD t: SAMPLE Your laboratory's mean value is essayes ayes Indicates for a specific control the rulesS to rules to apply to ts To BIOLOGICAL ANALYTICAL "ICAL TESTING (PT) •ayes ges es 1.0 ± 1.0 Why VT' CAL •m CAL VARIATION . eos e a chan Why Dc' SAMPLE Why Why -H Why Me VARIATION . System dew Sy*em dew ac Why He dewc do: OC Gaussianb,i ppent Gaussian to measure a sca Appenr Gaussianbis+r;b ppeor Gaussian -He gages Appear Gaussianbis+fib When to Appear Gaussia Gaussian b: He OC Me QC samples QC OC (QC) OC samples VARIATION ION 95% messages bes "len ppeot Gaussian • Captures a snapshot of your test performance for a given VARIATION - VARIATION VARIATION • • Describes your performance relative to your peers. Gaussianb)isfribu materials, wherever possible, especially at or near clinical decision matertats;wnerever possible, especiaily at BIOLYTICAL ANALYTICAL SAMPLE BIOLOGICAL Abs SF+em descrtk osn ppeor Gau Appeor Gaussian-I Gaussian Disfr,buf Gaussian bis+T,k SAMPLE. VARIATION sr apply appt y appty appt Y Appear ppeor Gaussianbis+fibu+ ppe.ot Gaussianbisfr,b HOW How ppenc/ nes SS ayes 100 Gausstan , ages 99.7% = Gaussian bic tan a scan the same as the group, no bias ey essayes e ss s imprecision is less than the group's L Abs Total PRE. dev ±3SD ppe.ot Gauss Gaussianbisfr,bu+ed Gaussigne)is+f Use t to S or ODS SS SAMPLE SAMPLE Appear Gaussiar TESTING Gaussianbisåf,ibu+ed t 0 to measure 99.7% = . (Pr) t a Cho 6u!uueos SAMPLE How VARIATION . False Rejection Use t OC samples Appear VARIATION schedule ppent Gaussianbisif,bu K SD Es+iwn+e IS QC OC • Calculation Device (e.g. phone, to Appeor Char + • phone, A' vec'ßon 2SD St) provided by the all lab methods' mean BIOLOGICAL peer or all-lab CV ppeot GaussiarDis+q,bu+ BIOLOGICAL ppeot ussian i •als Excellent source to determine Target Value for each control analyte at a specific point in time asc piec,310n K SD Es+im*e pp The central facility examines the data for outliers 4. The central facility examines the data for outliers devi anf in Curacy — samples previously examined; — samples previously exangined; false reiection False Rejection a scan ppeot auss 1. ethod CK Appear M eth o d average Impreclsvon average Impreclston average TICAL dev•ce ION 1:2s ru •K SD K St) Oppent Oppeor 99.7% When to ppent to measure SA Resolve any discrepancies between • Resolve any discrepancies between Resolve any discrepancies between U t a cha St) Jon peer Rate devi mute! Gaussiaöisif,bu+ QUALITY bra independent test results obtained under stipulated can 1. Calculate the CVI for each control. I. Calculate the CVI for each control. upeJsnp Appear GaussianÄ)is+f . 'A WHERE a Chanc in Appear Gauss—a 3. Gaussian t)iswtbt" How GaussianÄ)isWIbu+ed Gaussiant)is+flbu+ earl HOW b es Appear GaussianÄ)ié Gaussianl)isv laboratories performing the same dervice service em descftb es Abs t a change in ocuracy or can qppeot qppe.ot Gauss— +he L-J Gaussiqne Acceptable performance when 1.0 Indicates for a specific control the Acceptable performance when schedule Appen q ppeo Gaussian b Gaussian, Gaussio.. Gaussian yste Rate SDI: ± 1.0 o .53 o .36 Appear Gauss Appear n Appeot can .36 evaluations Gaussiane Gau The accuracy c Abs •ascns Why Change au easons aussian ist Grades your performance ussian i laboratoryes. laboratories. -an t a change in or mixed values, which ensure the validity of decisions made. vall es, which ensure the vaiidit t a change In octuracy or aussiar L-Tckat TICAL Key Messagef He Key Message When to ppenv aussi, to GaussianÄ)iV aussio pp 'asons descri b es calculator) Calculater) can Chan \ The accucacy and an a Change iIn calculator 'n ±3SD devi a sca Chat + ppeov au L-Jchat+ a scan ICAL •asens ass an How Median eun easons also compared to your group imprecision is the same as the compared to your group ATION eon The accuracy QqecÅSton The accu-racy and The accuracy and@qecÅslon an Vcucacy an CONTROL Gaussiany e and Xhe accuracy and an Oppen ppeot Chat assane -d ppe _ pp Assi The accucacy and@€ecÅSton an The accucacy and an and an fid TE in units of cells x 109/L a Change in ocur ppeLt O.OS or ec•son t mixet schedule ±3SD veclslon on aim S 0.05 or • Provides some idea about the amount Of bias present in the aess asso ass VARIATION Appear Gaussic Chante Idea •bout the asso Tbe accncy an The accuracy and Curacy and an Median cBton an Abs Comparison Programs pan tach *ou+fiers usr Chatge Are rcS re the pairs. The accucacy -ptecÅSton The accucacy j)fecÅs'ton IL and calculates the means and SDS for the peer and calculates the means and SDS for the peerpusm LAB C LAB D assi When to BIOLOGICAL *.ccucacy and an t a in ocuracy or lccucacy and ire Abs tack # Labs: Labs: accucacy an accuracy 'acy eclSton an aim for 0.05 or fid an evaluations or The and@€ec'ßlon an ac — material from cell or tissue repositories; — material from cell or tissue repositorie:v • Also known as Also known as t mixed poputa+ion t a change in m The accuracy The accucacy and The accumcy and an The accucacy an The accuracy and an •oho "on an schedule Accuracy K SD ISD ON accuracy and an accucacy and accuracy and@qecÅSton an *007 accucacy and an •i Qfld d@qeclslon d an •K St) •K t mixed populo' t a Change in acuracy analytical system. Key Messa (QC) acy Qqd an sac. r10N ANALYTICAL/ qd@vec'ßon an t mixed popula+io @€eclslon an vec'ßon an vecÅsion L-Tch (5 and ptecßon veclslon t a change in acuracy a Change in occuracy L-T cha conditions 1 284 Analytical rcu t a change In or evaluations Key can or TION t a change in ocuracy or a change in acuracy or eon Cm ± 1.0 to 1.5 Problem may exist, laboratory Son Problem may exist, Chary pn an sac. • Pencils— t mixed poputThion t a change 'In QCuracy a change ocuracy or ' t a change in • Pencils K SD ac aim S 0.05 or •K an 0/0 Total t mixed popula+ion IS *ou+hers used in calcula+ ANALYTICAL a Change in OCuracy Ooh • o/ a change occuracy a Change in t a change c fid@€ecÅSton a, Abs a change ocuracy accucacy and@€eclslon an I and Chan9e he analy+ica\ mixed m b es Analytical t a change 'In a change "n ocuracy or m descftbes descflb es descr•t b es Il t a change in or Cost Costr Additional costs to participate in a PT provider's testing scheme Usually no additional cost to participate; service provided with the IS Sys+em descrtb es *Ou+h ' a change CAL Sys+em es a change In ocuracy or ptecision a change or should investigate problem with imprecision is greater than the t a change or t a Change in (Curacy a change in ocuracy Syså-em b es sianbis+f,bu+ed on Ooh 4 Sys+em descri b es descfl b es fib es a Change in oc a change in acutacy or Change in a Change ocuq ate in Workshéefl's class a change in acur a change Cf LA SAMPLE OF SIAC. *ou+hers used in calc, Your Lab's CV -OF Abs m mol/l- Chavge Ooh, S, Sys+em descftb es a change in Change in ocuracy o Participate in Worksheet I's class Par9icipate in Worksheet 1 s class 'S ooh G ys+em b es VARIATION t mixed a Change •oho c sysfem descftbes descrté es sys+em descfl b es Chanf Sys+em descft b es sys+em b es Sysfem describes Sys+em b es Sys+em descflb es SysYem b es syså-em es Sys+em descf(bes Nm descrt b es descftb es a change oc Curacy a Change oocuracy a chor,8e in group and all-lab group, and SDI and CVI for your group and all-lab group, and SDI and and ysfem descnb es Yem descftb es m describes es m mol /L Abs a change ocuracy or sys+em describ es a Change QC used in calcuta±ion descri b es m mol IL b es m mol IL Syså-em es t rntxed *ou+hers used in calcula±ion *ou+hers used i t a change In ocuracy or a Change In or p€ecision tach t mixe method for a particular analyte Error t a change in occurac purchase of QC materials. bias group's average Imprecision group's avera e imprecision grou 's avera e imprecision grou 's average Imprecision sy*em es sp+em descftb es Sysfem descflb es fem b es Syså-em descftb es Shi-em describ es Syså-em descflb es m descflbes Chante in ocuqacy Coefficient of Variation Lem b es *ou+hers used in Fem descfl b es em descftbes m descflb es • describes descrt b es descft b es escft b b es noN y+ic Sys+em descftb es /l_ change In o/ descftbes — exchange of samples with other laboratories; Your laboratory, along with other laboratories, — exchange offampl — exchange of samples 'With other laboratories; used in ion used in Chante in æcuco o/ ages a chary 4. If the criterion for acceptable performance for platelet count is 25%, is the TE acceptable? Concentration thant a Chan, Chante Fem es o/ a Chant QC cm •hen t mixed poputa+ioo t mixed popula+ion as OC 'hen *ou+hets use Concentration o/ Your Lab's Peer Your Laboratory Your Lab's NOTE 2 Use of independent third party control materials should be 'end ent third pacty o - - O, anal) a change acutacy ± 2.0 or greater Troubleshoot bias problem and ± 2.0 or greater Troubleshoot bias problem and Troubleshoot imprecision problem Hers used ooh" - > 1.5 Change 3.639 o/ to fie t a change In accuracy Chant we IS e No + TAT for Report The time between receipt of PT samples for testing and returned Many manufacturers provide on-line services for up-to-date on-line services for ISD a Change accuracy a change ccuqacy Pruecision - The closeness of agreement between the Truenesson - The closeness of agreement between the asz a change in æcucac a Cho K SD fie ION a change acc a Change acuq ness -The cl a Change 'In a change In accur a Chanc In acuca (Ecu 4). assionbsfr,bufed on t a change 'In a chant cc a change acu a chanc Change 'In Chante occur Ooh v used in WHERE WE ARE a Change in Change in mixed WHERE * ach * a chant 'In a chan a Chante a chanc In a Change Chant 'l chavge discussion. t a cha a change a chi. a Cho chat Change Chart change to anal) Bys+em descfl b es Sys+em describ es soys+em descfl b es Bys+em b es a Chan b es Chanse he m mol/L t mixed t poputa•hof) Change anqly• WE ARE ARE and perform corrective action Abs Abs ERE ARE E WE ARE - ARE defines fie *poin+s HERE we ARE • fie anqly• laboratory. Laboratory. fie analy * mixed a change change in mixed TO ME ARE HERE WE ARE IRE *Ou+hers used in t mixed report for review may be a few months. Information information WHERE Ckar+ mixed Receipt fie t mix TEA in units ploæ c. t WHERE WE ARE TICAL mixe ass ardy€eal 23 24 25 26 27 28 29 30 23 (Curacy Error 24/5 28 24 30 31 3. or *poin+s plo correctly QC Material 12345 10/6 12 11 12 Used WHERE WE ARE [ERE WE ARE WE ARE we ARE r10N noN used in If CVI = 1.5 to 2.0, your lab is 50-100% less precise 1.5 to 2.0, your is 50-100% 'ess precise — WHERE ARE ARE WHERE ircorrec+ WHERE .0, your K SD Abs *poinfs p' Run Ratio (CVR) OF S(AC ICAL used in cat used in calculo used in calcul( +0 *ou+hets used in catcult±io and and t mild , E WE ARE used in cal a Chante has t a Change in Used ocuqocy concen aqd limi+s are fimi+s used in catcu! Part 1 m mol (L used in cal' Chavqe used Pitfall o/ If report TAT is extensive it may not allow laboratories to detect Manufacturer Of QC materials used for the method does not Abs t mixe used in calcula±io of cells x 109/L of cells x 109/L- of cells x 109/1- corfdefd ei er instead of, or in addition ctntrol f' used in calcul considered, either instead of, or in addition to, any control materials — control materials that are tested daily in interlaboratory used catcuta±ion WHERE Used in catcutm *Ou+hers used in catcuto+ Change used in calc used assiqne Used in calcut Used calcula+ Used catcuto and % Total used in calcula In addition to the SDI and CVI, inter-laboratory reports often include: +0 +0 bein9 *poinfs plo used calc assiqned. inq used in catcui Ooh O mixed Used i used in colcuto Concen+ratien Bas ry used . Group SS used in ca usec WHERE WE ARE bein9 — is 360.0 620.0 610.0 635 620.0 655.0 used in 655.0 620.0 used calcula 100 635 _o 655 _o 655 _o 710.0 6605.0 Value 6100 615.0 610.0 6250 t mixed Used in catc WHERE WE WHERE mnixed -ub irrorr moixed Used in calcula±i, used catcuto used in catc assiane mixed mtted average value obtained from a large series of test results Used calcuta±io ge value obtain and correct in a timely manner problems that affect quality provide this service Oon - and graph se+-up ircorrec+ly and Error ooh 4 Ooh o se+-up ircorrec+lY o/ 'ufed on — and fid used dcu ) ANALYTICAL pncen used calcut dc graph se+-up irrorrec+ly se+-up incorrec+ly used calcu E WE ARE ARE amly+icql chniQlty me Change is WHERE WE * Ou±hets US racy or preasion graph se+-up ircorrecklY we Abs ERE we o/ Used "n ERE we ARE sRlbufed on Change in Used IS Concen LS LS LS IS LS In itials AM AM AM AM graph ircorrec+lY and chon on o/ Your laboratory receives a report indicating your plo Error a change acc 1. 3. 2. m mol/L This month's group mean, SD, CV, and number Of measurements involved. Your month's mean, SD, CV. and number amly%cnl US plof4d correctly cc *hen QC cm IS plof4d plof4d corr Compliance used colcuta€to Required by many regulatory and accrediting bodies mixed ssol • Optional, but highly recommended if service is offered has occurted. SD ISD plo+4d 3.639 3.539 Mean Median mmcl/L o/ used calc Used calc Used CO! plof4€d corr Used call Used calcu I acy pccision cesqt+ cfiniQ/ty defines fie *Ou+he Used dc 3/6'. Used In cal used d -2 13 -3 WE Expecr ro 'IS the Used a 230/5 29/5 231/5 230/5 230/5 ' 'IS *offers Used 'l Used In supplied by the reagent or instrument manufacturer. than its peer group, usually requiring Date 1/6 2/6 Date 11/6 cliniQi/ly o/ *ou+ess used Inco used 'In co suVpLied by ne reag-nt or instrument Cho used cn used 'l used in a ac used In VARIATION ro æh g+-up used Concen+a+i Used Concen+ra+ien ircorrec+l>/ corr ERE IS 'IS * potnås 1.0 25 oof corr correc±ly comparison programmes. comparison programmes, o/ on Used •tn a or ) +hen many se\-up ccorrec+ly' WHERE WE EXPECr ro +hen Peer M eth o d M ethod M ethc•d ERE we EXPECT ro ERE WE ro ERE WE RE WE ro is QC cm QC WE EXPECT ro • Participation may fulfill regulatory and accreditation we qe+-up and * ou\hess 'RE WE m IS *poinås plof4d —4WHERE EXPECR WE EXPECR ro —'WHERE WE EXPECT ro g IS corr mes. —WHERE EXPECt ro 8 —'WHERE ro o/ ISD — IS nples ANALYTICAL —WHERE we EXPECr ro EXPECT ro g sac-.— IS S+ima+e * potnås C —WHERE'VE EXPEcr ro EXPECT E.rpecr ro sac: - handled incor incorræ+ly —WHERE IS TE < TEA incorr WE EXPECr ro beir used in dc PRE- Your month's group mean, SD, CV, and number of measurements involved. . Your month's group mean, SD, CV, and number of measurements involved. •e we EXPECr ro WE ro WE ERE WE EXPECr ro con 3. WE expecr ro g we we EXPEcr ro we EXPECT ro g This month's group mean, SD, CV, and number Of measurements ir is and an accepted reference value EXPEC* ro expecr ro Be— EXPEcr ro EXPEcr xpecr ro Concen+&ien rro+enials hordled incor so 1: —o -27 —0.27 —o .31 —o -31 - EXPECT ro ro ro con Con rro+enals incorræ+ *poin QC samples used IS the Use the IS anqlyiica\ iical raph se+-up ircorr requirements for the analyte in some countries. analytical performance. * mixed bash - ppeov SAC. sac. ir hange the anqly+ical t rnited Cumulative or lot-to-date (LTD) laboratory mean, SD, CV, and the number of measurements involved. c. Cumulative or lot-to-date (LTD) laboratory mean, SD, CV, and the numb Cumulative or lot-to-date (LTD) laborator•i 2. 3. 3. fie analyiica I analyiical 0.8 Anamly+icql alyiicall S(AC. SQCI PRE- OF SAC. OF "SAC. shS—oF SAC. 'SQL. SQCZ S(ACJ hse+-up - sac: ce+-up inc sac. se+-up ircorre sac: sacs t mig BIOLOGICAL -up ircorrec+l co ANAL "ICAL t rniy ircorr 80/0 qraph se+-up Abs OF asure t mix: mix c.e+-uD irco. 00k 004 Abs -ub .ircorrec+l> c. acc < ± ISD +0 Used h..ce+-up ircor incorrec+l ircorrec+l 33 h...se+-up ircorr graph Your Laboratory Laboratory 100 # Labs: Labs: 25 33 horded MAY (cihi'éÅtmon'th) MAY (month -1) aralyfe beins amlyfe beins +es ardyfe beins specific specifi investigation. WHEREØÆARE 1. MAY (current month) _ircorrectly/ ircor VA JUNE (current mont!ü+ (current month) -up trcor Con CP+-UD as OGALYTICAL ce+-up ircor 950/ c.e+-uD iftor JUNE occuffed. Abs ce+-up ep+-uD —G IS IS -up -ub iftor ire trcor lany QP+-uD se+-up ircorre Concen+ra+ion itcorr has occuvted WHERE mE Abs P Oints: 1 295 1 2gs 1 61 6 1 61 4 c.p+-uD has occuvced Abs se\-up if-corr no-k aly+ioal clini@lly and applied. Concentra±ien Concen+ra+ion * incor WE EXPECr ro ed. 15 ± Concen+atien urocy applied. specific + * 00k 00k 10 11 1. 14 12 13 <+lsb+o VARIATION se+ WE 15 inks eek-up se±-up SIS Is ne+ Peer Group C peer Group B peer Group D «lsb+o I Abs deficed and applied. Con WHERE 'IS - is no+ defieed and applied. -s do —'WHERE used I WE gorre&ly e no/ iftcorrec+l>/ D +0 no-k 20 20 minutes art WHERE we arc WHERE WE 1.4 COncen#a+ion Concen+rat{, cr€of€ ERE WE SDawqred we nples U. iftorre WHERE WE WE ARE IS +hen m mol h mmol/l 14 Run -be Activity 12: Understanding Inter-laboratory Comparison Programs Activity 12: WE ARE expecr ro mmol/ mmol/L — IS do Concen+ra+inn Activity 12: Understanding Inter-laboratory Comparison programs Activity 12: Underst Underst' ro BE' Co 17. and . grams grams concentrn IS no WE A Re IS