Page 1
The business case for optimized quality control practice Zoe Brooks, AWEsome Numbers Inc. Philip Morehouse, Cape Breton District Health Authority Reprints available at www.awesome-numbers.org Contact zoe@awesome-numbers.org
Page 2
Purpose of the Study Background Clinical Laboratories commonly monitor their testing procedures by repeatedly analyzing quality control materials and assessing the results to determine acceptable performance. In most jurisdictions this requirement is regulated and results in a significant volume of testing and cost to the health care system.
It is recognized however that most laboratories do not follow generally accepted standards of practice concerning Quality Control practices. The negative impact of this, inside and outside the laboratory, is not well understood or easily quantified.
In the current environment of limited resources laboratories need to ensure that their quality control system design maximizes both efficiency (lowest possible costs in the laboratory) and effectiveness (best possible protection of patients from clinical harm of misleading laboratory results.)
The Study To understand this better, we analyzed our current Quality Control practices for30 tests across6 laboratories and compared our practices with Q.c processes recommended by a prototype version of Quality OptimiZerTM Software (Awesome Numbers Inc.) to determine:
- the gaps in practice, and
- the impact of the identified improvements to overall healthcare system reliability and costs.
Page 3
Process Flow Define Objectives
Record Nov. 2011 Q.C. Data & Costs for 65 Q.C samples from 8 analytes in 30instrument-analyte combinations
Choose True Value for samples in order of: PeerMean-allmethods-cumulative PeerMean-thismethods-cumulative Peer Mean - local lab group Midpointofpackageinsertrange
Examine Total Error in OptimiZerTM
SelectattainabledefensibleTEaLimits foreachsamplebasedonStockholmHierarchy
Performancemeets Clinicalstandards?
No Yes
Reviewvalidationof Stop and Improve Method analytical & Q.c. processes, clinical risks, & costs
Page 4
Q.C. in Theory Routine Quality Control practices detect change fromstableacceptableperformance that produces clinically-meaningful results
- Events that change methodperformance Event can&dohappen frequently in labs.
- Scientifically-referenced Q.C. processes vary Process Q.C. rules & actions to match method performance (Margin for Error.). Evidence
- Evidence e(Q.C. Flags) alert staff to clinically- significant change. Interpretation 4. Staffinterpret the evidence consistently. Q.C.flags are evidence of change, not failure. Consequences Change canbefor the betterorworse!
- Staff understand the Action consequences of misleadingpatient results;
- Take the right action. Impactof 7. Report results only Event when performance meetsclinicalgoals
Page 5
OptimiZed Quality Q.C. rules are selected to match the size of shift tolerated before results become clinically misleading:
- the ‘Margin for Error’ ExampleA’canshift 3.3 SD. B Example‘B’canonly shift 1.3 SD.
OptimiZed Q.C. Q.C. Practice Example Asg mean & SD = Meas Asg SD = 2x Meas SD. Rules match ME Rules’1-2s’ &‘1-3s’
B
OptimiZed Q.C. Q.C.Practice:Stop&Act Asg SD= Meas SD for ‘1-3s’ reject. Rules vary with Margin for Error Asg SD = 0.8 to 3.7x Meas SD 20 200** 米 3 Nosampleswouldtake Over50%ofsamples 15>10runs toflagclinically >Change 5 would take > 10 significant change. runs to flag clinically significant change 10 Runsto1stFl
- MAIACA 2.0 2.0 6.0 10.0 14.0 -2.0 2.0 6.0 10.0 14.0 Marginfor Error Margin for Error APTT PT NaCa Chol BHCG TSHDig -·1:1Line
Page 6
Example Bottom Line MYLAB Sunshinometer Serum Sunshine 06/01/2012To06/30/2012 AWE some Quality OptimiZerTM ValidationofAnalytical Process: ValidationofEffectiveness: ValidationofEfficiency: NuMBERs Efficient,Effective, Pass-A.AIlOK Fail-Stop&Fix Pass-A.AlIOK org ReliableQuality Q.C.Sample PerformanceBarZ? Analytical QCRating: Effectiveness Efficiency Reliability Q-Score Level1 OptimizedQuality ★★★★★ ★★★★★ ★★★★★ 100% ★★★ Lot:QC1111 ME:3.0 Lab Practice ★★★★★ NoFlags 40% Level2 OptimizedQuality ★★★★★ ★★★★★ ★★★★★ 100% ★★★★ Lot:QC2222 ME:4.0 Lab Practice ☆ ★★★★★ 43%
QualityMetric HowItIsMeasured Level1 Level2 Level3 ActionstoimproveQCDesign, Whatitmeans Impacttopatients&lab Quality Lab Quality Lab Quality Lab Effectiveness,Efficiency Optimizer Practice Optimizer Practice Optimizer Practice andReliability Q.C.Design ActualQCruleused 2.5 6.0 3.5 5.3 AnalyticalProcess:Lev1 23 Assigning theright QCvalues; RatioofActual QCruleusedto 1.0 2.0 1.0 1.8 LowQualityRating Using therightrules theoretical QCrule. Q.C.Effectiveness Probabilityoferrordetection 78% 0% 78% 16% Improve precision Abilitytoflagclinicallysignificant Runs to firstFlag 1 20 1 6 Improve accuracy change thatcancause>5%of Patientsatrisk/year 10 200 10 60 Improve BOTH patientresults tobeclinically Patientsharmed 1 10 1 3 QCProcess: misleading. Clinical Cost S20 S300 S20 S90 ComparetheConsequencesofthe Q.C.Efficiency Probabilityof falserejection 0% 0% 0% 0% OptimiZed and Laboratory processes. 1.Whatis the clinicalimpactof Abilty of theQCprocess to NOT Falseposruns/year 0 0 0 0 OptimiZed Q.C.(patientsrisk&cost)? flagresultsthatareclinically Reagent&SupplyCosts S0 S0 S0 $0 2.HowwouldOptimiZedQCimacttime acceptable(few falsepositiveQC &costs in thelaboratory? Staff Costs SO S0 SO SO 3.Would thechangeinreliability flags.) TotalLabCosts S0 $0 S0 $0 improve staffconfidence inQ.C Q.C.Reliability Stopreporting/improve ProbabilitythataQCFlagsignifies TruePositiveFlagsperYear/ 100% NoFlags 100% 100% OptimiZeQ.C.Process significant change. True+FalseQCFlags OtherProcess
Theclinicalcostofpatientsatrisk: The clinical costofpatients atrisk: $1.00 The OptimizedQualitycouldpreventrisktoupto190patients;andupto9patientscouldbe Financial Cost of clinically misleadingresults: $10.00 savedfrompotentialharm.ClinicalsavingscouldtotalS280. Assumptions Reagent andSupplies costtest: $1.00 Salarywithbenefitsmodeledat: $50.00 Bottom OptimiZedQualitywillcostS0moreforreagentsandsuppliesandS0moreforstaff. Total=s0 peryear. Average=100runspermonthwith10patientsperrun Line Combinedfinancialimpacttoclinical andlaboratory:SavingsofS280 Simulatedchangewouldcause5%ofresultstoexceedTEalimits,with thissignificantchange 1xinayear.
Page 7
ChartComparison MYLAB Sunshinometer Sunshine 06/01/2012T006/30/2012 AWE some Quality OptimiZerT Average=100runspermonthwith10patientsperrun Labpractice=CheckQCafterreportingpatients.Repeat NuMBERs Efficient,Effective, forRejector2consecutivewarnings. org ReliableQuality Simulatedchangewouldcause5%ofresultstoexceedTEalimits,withthissignificantchange1xinayear. WhereYouAre WhereYou AssessmentofAnalyticalQuality Where yousay you The Asg TheAsg SD “Stop& “Stop& Want ExpectToBe Mean is is Act Rule” ActRule” (andexpecttobe) ToBe Abilty toproduce clinicallymeaningfulpatientresults. AssignedChartValues SD Optimized LabReject from the xMeasSD Quality Mean SD True TEa Bias TE ME OK? Asg Mean Asg SD Meas Mean 1-3s Level1 100.0 2.00 100.0 10.00 0.0 4.0 3.0 OK 100.0 4.00 0.0 2.00 1-2.5s 1-6.0s Level2 101.0 1.50 100.0 10.00 1.0 4.0 4.0 OK 100.0 3.00 -0.6 2.00 1-3.5s 1-5.3s
All Leveb Facts:Measured FromOpinion(Policy) Mesn-True (Biss+2SD) (TEa-TE)VSD TE<TEa? AsssignedChartValues (Asg-Meas)SD Asg/Meas AsgRule Modeled PerformanceChartsplotQCvaluesagainstTrue OptimizedQualityassignsthemeasuredmean& LabQCpracticemayassignothermean&SD ValuesandPerformanceStandards(TEaLimits) SD;selectsQCrulesbasedonMarginforError values,modifyingQCrulesfromtheory. Performance Chart OptimiZed Q.C. +3SD 112.0 Lab Q.c. +3SD 110.0 +TEa 106.0 Meas. Asg.
100.0 True 100.0 Mean 100.0 Mean
90.0 -TEa 94.0 -3SD 88.0 -3SD Meas. Asg.
Performance Chart OptimiZed Q.C. 109.0 Lab Q.C. +3SD 110.0 +TEa 105.5 +3SD Meas. Asg.
100.0 True 101.0 Mean 100.0 Mean
90.0 -TEa 96.5 -3SD 91.0 -3SD Meas. Asg.
Example
LabQCWarning(withoutaRejectFlaginthesamerun.) LabReject OptimizedQualityActionFlag Chartsshow50simulatedQCresulitsbasedonyourmeasuredmeanandSD.Aftertheredine,20pointsshowa shiftwith5%ofresuits>TEa(Clinicallymisleading.)
Page 8
Real Bottom Line LAB2 INSTRUMENTB PT 11/11/2011To12/21/2011 AWE Quality OptimiZerTM ValidationofAnalyticalProcess: ValidationofEffectiveness: ValidationofEfficiency: some NuMBERs Efficient,Effective, Pass-B.Review Fail-Stop&Fix PasS-A.AlIOK org ReliableQuality Opportunitiesforlmprovement:OptimiZeQ.C.processtoimprovepatientcareandreducelaboratorycosts. Q.C.Sample PerformanceBarZ Analytical QCRating: Effectiveness Efficiency Reliability Q-Score Level1 Optimized Quality ★★★★★ ★★★★★ ★★★★★ 100% Lot:1111111 ME:3.0 Lab Practice ★★★★★ No Flags 0% Level2 OptimizedQuality ★★★★ ☆ 30% Lot:2222222 ME:1.6 LabPractice ☆ ★☆★★☆ ★★★★★ 43%
QualityMetric HowItIsMeasured Level1 Level2 Level3 ActionstoimproveQcDesign, Whatitmeans Impacttopatients&lab Quality Lab Quality Lab Quality Lab Effectiveness,Efficiency Optimizer Practice Optimizer Practice Optimizer Practice andReliability Q.C.Design ActualQCruleused 3.0 5.7 2.0 3.0 AnalyticalProcess:Lev1 2 3 Assigning therightQCvalues; RatioofActual QCrule used to 1.0 1.9 1.0 1.0 Low Quality Rating X Using the right rules theoreticalQCrule. Q.C.Effectiveness Probabilityoferrordetection 66% 1% 49% 16% Improveprecision X Ability to flag clinically significant Runs to first Flag 1 20 2 6 Improveaccuracy change that can cause>5%of Patients atrislyear 33 661 66 198 ImproveBOTH patientresultstobeclinically Patientsharmed 2 33 3 6 QCProcess: misleading. ClinicalCost $53 S991 S96 $288 ComparetheConsequencesofthe Q.C.Efficiency Probabilityoffalserejection %0 0% 5% 0% OptimiZedand Laboratoryprocesses. 1.Whatistheclinicalimpactof Abilityof theQCprocess toNOT Falseposruns/year 0 0 36 0 OptimiZedQ.C.(patientsrisk&cost)? flagresultsthat areclinically Reagent&SupplyCosts SO S0 S648 S0 2.HowwouldOptimiZedQCimacttime acceptable(few falsepositive QC &costs in the laboratory? Staff Costs SO S0 $675 S0 3.Would thechangeinreliability flags.) Total Lab Costs SO S0 $1,323 S0 improve staffconfidence inQ.C Q.C.Reliability Stopreporting/improve ProbabilitythataQCFlag signifies TruePositiveFlagsperYear/ 100% No Flags 3% 100% OptimiZe Q.C.Process significant change. True+FalseQCFlags OtherProcess
Theclinicalcostofpatientsatrisk: The clinical cost ofpatients atrisk: $1.00 The OptimizedQualitycouldpreventrisktoupto628patients;andupto31patientscouidbe Financial Costofclinicallymisleadingresults: $10.00 saved frompotential harm.Clinicalsavingscould totalS938. Assumptions Reagent and Supplies cost/test: S1.00 Salarywithbenefitsmodeledat: $37.50 Bottom Untilanalyticalqualityisimproved,OptimiZedQualitywillcostS648moreforreagents and suppliesandS675moreforstaff.Total=S1323peryear. Average=60runspermonthwith16patientsperrun Line Combinedfinancialimpacttoclinicalandlaboratory:CostofS385 Simulatedchangewouldcause5%ofresultstoexceedTEalimits,with thissignificantchange2xinayear.
Page 9
ChartComparison LAB2 INSTRUMENTB PT 11/11/2011To12/21/2011 AwE some Quality OptimiZerTM Average=60runspermonthwith16patientsperrun Labpractice=CheckQCafterreportingpatients.Repeat NuMBERs Efficient,Effective, forRejector2consecutivewarnings. .org ReliableQuality WhereYouAre WhereYou AssessmentofAnalyticalQuality Whereyousay you TheAsg The Asg SD “Stop& “Stop& Want ExpectToBe Mean is is ActRule” ActRule” (and expect to be) ToBe Abilitytoproduce clinicalymeaningful patientresults. AssignedChartValues SD Optimized Lab Reject from the xMeas SD Quality Mean SD True TEa Bias TE ME OK? Asg Mean Asg SD MeasMean 1-3s Level1 12.2 0.26 12.8 1.90 -0.6 1.1 3.0 OK 12.2 0.50 0.0 1.92 1-3.0s 1-5.7s Level2 37.9 1.48 38.2 5.70 -0.3 3.3 1.6 REVIEW 37.9 1.50 0.0 1.01 1-2.0s 1-3.0s
AllLevels Facts:Messured FromOpinion(Polioy) Mean-True (Bias+2SD) )(TEs-TEVSD TE<TEs? Asssigned Chart Values (Asg-Meas)/SD Asg/Mess Asg Rule Modeled PerformanceChartsplotQCvaluesagainstTrue OptimizedQualityassignsthemeasuredmean& LabQCpracticemay assignothermean&SD ValuesandPerformanceStandards(TEaLimits) SD:selectsQCrulesbasedonMargin forError values,modifyingQCrulesfromtheory. Performance Chart OptimiZed Q.C. +3SD 14.3 Lab Q.C. +3SD 14.7 +TEa 12.9 Meas. Asg.
12.8 True 12.2 Mean 12.8 Mean
10.9 TEa 11.4 -3SD 11.3 3SD Meas. Asg.
Performance Chart OptimiZed Q.C. 42.7 Lab Q.C. +3SD 43.9 +TEa 42.3 +3SD Asg. Meas.
38.2 True 37.9 Mean 38.2 Mean
32.5 TEa 33.4 3SD 33.7 3SD Meas. Asg.
Real Data
LabQCWarning(withouta RejectFlagin thesamerun.) LabReject OptimizedQualityAction Flag
Page 10
Root Cause: The Q.C. Gap Q.C. practice diminished Q.C. effectiveness and efficiency. Assigning 4.0 Assigned vs.Measured SD 3.6 chart mean &SDatvalues 2.8 NOT measured 2.4 SDusedforCharts&Rules 2.0 by eachlab& 1.6 method cripples 1.2 Q.C. rules. 0.8 No SDs were 0.4 0.0 assigned at 0.0 0.4 0.8 1.2 1.6 2.0 actual values. SDCalculatedonQ.C.ResultsNovember2011
13.0 Meas SD = 2.0 12.0 Rules Used vs.Optimized 11.0 Asg SD = 2.0 10.0 1-3s rule flags -RejectRuleApplied 9.0 8.0 云公 at +/- 6.0 units (adjusted forSD) 7.0 6.0 If Asg SD = 4.0 Q.C.Practice 3.0 `1-3s’ rule flags 2.0 at +/-12.0 units 1.0 0.0 Rule Used =1-6s 1.5 2.0 2.5 3.0 3.5 4.0 OptimizedRulestomatchQ.C.Theory
20 Labsuseda 1-3.5s rule (or higher) 15 #Runs from for 89% of controls. significant RunsatRisk-Q.C.Practice 10 change to Q.C.Reject Only40%had Marginsfor Error high enough to flag significant change witha 1-3.5srule. RunsatRisk-Q.C.Theory
APTT PT Na Ca Chol BHCG TSH Dig-1:1Line
Page 11
Impact on Reject Rule Inflated Q.C. rules delay flagging significantchange&putmorepatients at risk of harm from misleading results.
NumberofPatientsatRisk
1,500 Inflated rules create
1,250 few true positives.
,000 Lab practice to stop
and act for a1-3s' Q.C. Practice 750 reject or two1-2s’
500 warnings placed
250 more patients at risk
than optimized
100 200 300 400 500 Quality Control.
Q.C.Theory
Inflated reject rules generate fewer false positives than Optimized Quality. 160 Numberof Lab practice created 120 False Pos fewerfalsepositive Q.C. Runs. Q.C. flags than Q.C.Rejects(Practice) 80 optimized Q.C.. OpitimiZer”uses 40 tighter Q.C.rules to detect change soonerinmethods 25 50 75 withlowMarginfor OptimiZedQuality Control (Theory) Error. The number of false positive runs is determined by the Q.C. rule.
to flag results at 2.9 to 12.1 SD from the mean.
APTT PTNaCaChol BHCG TSHDig-·1:1Line
Page 12
Impact on Warning Rule Wrong J mean and SD values caused the lab ‘1-2s’ rule to actually flag results at +/- 2.0 to 8.7 SD from the mean. Numberof PatientsatRisk 1,500 If staff had acted on 1,250 every single ‘1-2s’ flag 1,000 there would still be Q.C.Practice 750 thousands more 500 patients at risk 250 than with optimized Q.C. reject rules. 100 200 300 400 500 Q.C.Theory OptimiZerTM simulated a shift in the mean thatwould cause5%ofresultstoexceedTEalimits.Thenumber ofpatientsatriskisbased on thenumberofrunsto the first Q.C.flag after this change, and patients/run. 160 Numberof Lab practice created 120 False Pos twice as many false Q.C.Runs. positive Q.C. flags RunsWithFalsePositives-Q.C.Practice 80 as optimized Q.C. 40 You cannot fix this problemby just assigning the real mean 25 50 75 RunsWithFalsePositives-Q.C.Theory andSD on charts. Effective, efficient quality control requires using the right mean & SD AND the right Q.C. rule. APTT PTNaCaChol BHCG TSH Dig -·1:1Line
Page 13
Costs with Reject Rule Clinical Costs were 6.5 x times higher with the current lab`1-3s’ rule than with optimized quality. Three methods failed to meet TEa at the time of the study. $6,000 These 3 methods were $5,000 already producing $4,000 misleading results Q.C.Practice $3,000 results that failed to $2,000 meet TEa limits.
$1,000 Clinical Cost If that significant $O error remained $0 $1,000 $2,000 $3,000 $4,000 undetected for the Q.C.Theory full upcoming year, clinical costs would be 20x as high, with 20x as many patients at risk and potentially harmed, compared to OptimiZed Q.C.. We assigned a costof $1.o0 for each patient at risk and $10.00 for each patient receiving a misleading $6,000 result. Clinical costs $5,000 of missed or delayed diagnosis, or incorrect $4,000 treatment changes Lab Rejects(Q.C.Practice) $3,000 can be much higher. $2,000 Actual costs depend $1,000 on the analyte and its $OX impact on clinical $O $500 $1,000 $1,500 $2,000 OptimiZedQuality(Theory) decisions.
Laboratory costs were lower with inflated Q.C. rules. APTT PTNaCa Chol BHCG TSH Dig -·1:1Line
Page 14
Costs with Warning Rule Clinical Costs would be 3.6 x times higher, even if staff had stopped and acted for every`1-2s’ flag. Lab costs would double compared to OptimiZed quality. $6,000 Costs were almost 17x $5,000 higher, if we included $4,000 the failed tests that
Q.C.Practice $3,000 were currently producing results $2,000 with clinical impact $1,000 Clinical Cost to patients & cost $0 to the system now. $0 $1,000 $2,000 $3,000 $4,000 Q.C.Theory The question remains:if unacceptable performance was not detected to date, when would it be?
$6,000 Lab costs were based Laboratory $5,000 Cost on actual cost/test for repeat tests. $4,000 Labour was calculated Lab Warnings(Q.C.Practice) $3,000 based on the time $2,000 spent for Q.c. flags. $1,000 $O If the labs had chosen $0 $500 $1,000 $1,500 $2,000 OptimizedQuality (Q.c.Theory) to use their`1-2s’rule, that would cost $5300 more than OptimiZed quality. It would also cost more for clinical care by placing more patients at risk of misleading results.
APTT PTNaCaChol BHCG TSH Dig -·1:1Line
Page 15
Results Quality OptimizerTM Softwarevalidatedcurrent analytical quality, quality control effectiveness and qualitycontrol efficiencyfor eachmethod. Currentpracticewastostop and actonlyforrejectflags. Analytical Process Criteria APTT PT CHOL Ca Na BHCG Dig TSH Pass-A.All OK AlllevelsME>2.5 2 3 5 2 Pass-B.Review 1ormorelevelsME>1.0<2.5 2 2 2 2 Pass-C.Check NOW 2ormorelevelsME>0.0≤1.0 Fail-Stop&Fix 1ormorelevelsME<0 3 Q.C.Process-Effectiveness PaSS-A.AlI OK Change detectedin 1run 2 0 Pass-B.Review Change detectedin2runs 2 Pass-C.CheckNOW Changedetectedin>2runs 2 2 1 Fail-Stop&Fix ChangeNOTdetectedfor>10runs 3 2 5 5 Q.C.Process-Efficiency PaSS-A.AlI OK 0%falsepositives 5 3 5 5 5 1 0 Pass-B.Review 0-2%falsepositives 2 1 Pass-C.CheckNOW 2-5%%falsepositives Fail-Stop&Fix >5%falsepositives
Effectiveness OptimiZed Current Current and Efficiency Quality Process Process of Current Q.C. Practices Ifstaffreact If staff react only to Reject to every compared to ‘1-3s’Flags single OptimiZed Warning Quality ‘1-2s’Flag Actual Q.C. (1-2s to (1-2.9s (1-2.0s Rules Applied 1-3.5s) to 1-12.1s) to 8.7s) Maximum numberof runs 68 418 166 withpatientsat risk Maximum number of 4,046 63,966 53,080 patients at risk Maximum $5,648 $111,454 * $94,638 * clinical cost Maximum numberof false 304 42 655 pos runs Maximumcostfor $4,524 $1,049 $9,809 false pos runs
*includes $72,0o0 for tests that currently failed TEa &errorswere undetected by curent Q.C.practice
Page 16
Discussion Analytical If the analytical process produces clinically-misleadingresults Process
…and the Q.C.process sdoesnot Q.C. flag the change (or staff do not Process react appropriately) …
Patient ..incorrect patient diagnosis or treatment decisions can be made. Results Patients &healthcare suffer.
By simulating clinically-significant change, we discovered that over 50% of current Q.C. failed to detect this change in less than 10 runs. The Gap between Q.C. Theory and Practice
- Staff designing Q.C. processes were not familiar with expert publications & recommendations.
- There was no process to ensure that analytical processes meet performance standards.
- Staff used consistent Q.C. processes for analytical processes with very inconsistent performance.
- The mean and SD values that were assigned on Q.C. charts, and used to generate Q.C. rule violations, were set at values that did not reflect the analytical processes they were meantto control. Laboratories collectively spend billions of dollars on quality control’ to (theoretically) ensure that patients receive clinically meaningful results. Q.C processes should be validated for effectiveness & efficiency.
Page 17
Conclusion Based on our analysis, we estimate that with these current Quality Control practices as many as 63,966 patients are at risk annually of clinically- unacceptable results, which is a significant patient safety issue. Based on our assumptions, this could result in unnecessary healthcare system costs for the 30 instrument-analyte combinations of approximately $111,454.
When we extrapolate the impact of this to all 327 analytical tests across 7 Laboratories in our District Health Authority the potential avoidable clinical cost could be as high as approximately $1.2 million per year. With optimized Q.C practices, we estimate that this cost will reduce to approx. $110,874 annually. This represents a significant opportunity for improvement.
Closing the gap between knowledge and practice would require a commitment by Laboratory Administration to recognize this opportunity to improve the efficiency and effectiveness of their laboratories. This could be achieved by adopting generally acceptable standards of practice and consideringrelatively / minimal investments such as developing dedicated Quality Control deep- knowledge experts, wide spread technical staff education and training appropriate to their role, and software capable of easily analyzing performance and automating a customized approach to Quality Control for each test in their Laboratories.