Home

   Documentary Procedure for validation of examination of result (Internal Quality control procedure)

Purpose of Examination: To lay down standard operating procedure for Internal Quality control programme in clinical Biochemistry Section.

Principle: Delineate control of quality in laboratory processes to ensure generation of laboratory data of highest quality for the benefit of studies and participating volunteers and/or patients.

Responsible Personnel: Laboratory Technicians, Section Incharge, Quality Manager, Laboratory Director.

Precautions: Standard precautions should be followed when conducting running IQC material Check if the temperature of the storage (refrigerator/freezer) has remained within allowed range. Make sure that the quality control samples are analyzed after room temperature is attained. (Avoid rapid thawing of QC material).

Procedural Steps:

i. Selection of Quality Control material:

Commercially available material: ∙ Assayed controls (with assigned values) are preferred. ∙ Third party controls (not manufactured by manufacturer of reagents) as far as possible. ∙ Manufacturer provides a range of values for parameters as guidelines only and cannot be used as “true value”. ∙ In the case of non-availability of Assayed controls, laboratory will use known value samples as Quality control with established Mean, Standard Deviation for acceptance of results. ∙ Strict adherence to reconstitution and storage to be followed as per the manufacturer’s instructions. The QCs and Calibrators, wherever applicable, are aliquoted immediately after reconstitution and stored. Document the details of reconstitution, and the storage stability date & time. ∙ Ensure that QC material is reconstituted by calibrated & well-maintained pipette & type I distilled / deionized water. ∙ Generally QC material of the same Lot No. is procured, to last for 6-12 months, as long as possible for ease of operation. ∙ Establish mean, SD for QC Lot No in advance/parallel to current Lot No.

Finding Mean-SD for New Parameter

• Choose right method of parameter from kit literature
• Check that, is new laboratory mean falling under manufacture range or not?
• If Yes, Laboratory mean will be utilized for further evaluation
• If No,

1. Check for method – principle of laboratory with defined in IQC literature.
2. Laboratory will do maintenance of the instrument
3. Check for instrument calibration
4. Check for Reagent quality – lot verification. Expiry of reagent, storage stability, contamination
5. If needed, re-calibrate the parameter
6. Than, re- calculate the data or re-run the IQC for 20 times.

• Take Cumulative mean as Running Mean

To find laboratory New Own SD.

• Laboratory will used directly SD which derived from 20 data point, if total error % of this parameter does not exceed than TAE% defined by guideline (CLIA, Ref Annexure I), with this new SD & last month EQAS bias%
• Laboratory will use peer group CV% for finding new own SD.
• Laboratory will use CV% from previous IQC data of same parameter.
• And if TE%>TAE% , laboratory will calculate SD as per following formula.
• TE%= 2(Cumulative CV%) +Bias(take from EQAS)
• Formula for calculation of new SD
• New SD= TAE% X Cumulative Mean X 100
• Example :
• Glucose New IQC Finding Mean & SD
• Manufacturer’s Mean (Running Mean) = 279
• Manufacturer’s SD = 21.50
• Manufacturer’s Range = 236 to 322
• Method: GOD-POD Method

1. Laboratory’s 20 Data points for Glucose Level 3:
2. 256,251, 307, 310, 250, 261, 281, 271, 245, 265, 268, 278, 239, 248, 263, 271, 268, 259, 320, 259
3. Cumulative Mean = 264.27
4. Cumulative SD = 19.31
5. CV%= SD/Mean X 100 = 7.31 %
6. Last Month EQAS Bias% = -1.81%
7. Total Error % (TE%)= (2CV% + Bias) = 16.41
8. TAE% for Glucose as per CLIA guideline = 10 %
9. TE% < TAE% ,
10. So Cumulative SD is not acceptable
11. Calculate SD will be derived
12. New SD= TAE% X Cumulative Mean X 100

= 10 x 264.27 X 100 = 8.8

• New Mean = 264.27
• New SD = 8.8

Finding Interim Mean-SD for Expensive parameter

1. Laboratory will run for 5 IQC data points for expensive parameter to find laboratory own mean – SD before implementing IQC for evaluation
2. Remaining all process is same as Finding Mean-SD for New Parameter

Procedure for Updation of Mean

• Calculate SDI by following formula
• SDI= Cumulaive Mean-Running Mean/

Running SD

• SDI is within defined SDI (⇐0.5)
• If SDI < =0.5 , there is no need to update Running mean and same mean will be utilized till further evaluation
• If SDI > 0.5, Mean will be updated with cumulative mean and it will be taken as running mean till further evaluation

Procedure for Updation of SD

• Calculate Difference % from cumulative SD & Running SD
• Difference %= Cumulative SD-Running SD X 100

Running SD

• If Dif.% is ⇐ 20%, No need to update SD and Running SD will be taken till further evaluation
• If Dif.% is > 20%, cumulative SD taken as new SD till further evaluation

Finding Mean-SD for New Lot of IQC

1. Run for 20 new IQC run lot in parallel with the old IQC lot, prior to used for evuation.
2. Old IQC lot will be used to evaluate of laboratory performance.
3. From 20 data point of IQC, laboratory own Mean – SD will be derived before implementing IQC for evaluation.
4. New derived mean should be with manufacturer range , provided with quality control literature.
5. New SD will be find form

Laboratory will use peer group CV% for finding new own SD. Laboratory will use CV% from previous IQC data of same parameter. Laboratory will used directly SD which derived from 20 data point, if total error % of this parameter does not exceed than TAE% defined by guideline (CLIA, Ref Annexure I), with this new SD & last month EQAS bias%

1. From first day of implementation of new lot of IQC, CAB will used own mean & sd.
2. After consecutive , 20 , 60 & every 60 data point , CAB will be update mean-sd as policy of updation.

After Establishing range, QC results are documented, monitored, and evaluated for acceptability by L-J chart as follow

ii. Levy-Jennings Chart Graphical representation of variability of Quality Control material for an analyte helps to compare and understand performance of a method at a glance. For each analyte level/concentration a chart is prepared as shown below:

• A graphical method for displaying control results and evaluating whether a procedure is in-control or out-of- control.
• Control values are plotted versus run time.
• Lines are drawn from point to point to accent any trends, shifts or random excursions.
• A Figure of Levy Jennings Chart:

iii. Westgard Rules

• “Multirule Quality Control”

Uses a combination of decision criteria or control rules.

• Allows determination of whether an analytical run is “in-control” or “out-of control”

1-3S, 2-2S, R-4S are the rejection rules of Westgard that will be followed. 1-2S, 4 - 1S, 10 X will be considered as warning rules.

iv. General guidelines for Out-of Control Quality Control Trouble shooting: ∙ Repeating the control or trying a new aliquot of control without actually inspecting the value obtained and the reason behind it is not recommended. Repeating the control will often give us a value that may be “within the limits”, careful inspection of the actual repeat result will often show that “we may have just squeaked by” and what we really have done is to delay the trouble shooting and problem solving until future run.

* Is the QC run in appropriate QC file * Check the lot number of QC, lot/generation of reagent/cartridge * Is integrity of QC material doubtful? ∙ Make sure that the appropriate quality control was run. This will not happen if the QC is identified beforehand and analyzed using barcodes when possible. ∙ Make sure that the quality control samples are analyzed after room temperature is attained. (Avoid rapid thawing of QC material). ∙ Check the dates on which the control was opened or reconstituted and when will it expire? ∙ Check if the temperature of the storage (refrigerator/freezer) has remained within allowed range. ∙ Check if the QC material was reconstituted by calibrated & well-maintained pipette & type I fresh distilled / deionized water and well trained technologist.

• Are reagents at fault?

∙ Check the dates on which the reagent was opened and will expire (on board expiry where applicable)? ∙ Check if the temperature of the storage (refrigerator/freezer) has remained within allowed range.

• Is calibration ok?

∙ Check the calibration status (“calibration due/overdue/approaching”) of the particular parameter in question or the analyzer as a whole? ∙ Was the calibration verification satisfactory?

• Is analyzer ok?

∙ Many kinds of malfunction can disturb the analysis .Check sampling, reagent delivery, mixing, lamp integrity, and reaction temperature. ∙ Check if there are any maintenance related issues (due/overdue/approaching) of the analyzer. ∙ Check if any critical part of the analyzer has been changed recently, is a recalibration required as a consequence of the part change which is affecting the parameters in question. Note: If all of the above are okay, check by analyzing known value sample (preserved samples, PT samples or a calibrator with different lot number). At the same time reanalyze at least 03 samples that were analyzed prior to QC failure to ensure that all the reported results were acceptable.

• Remarks:-

i. In case the QC analyzed after 1st run or during the first run of the day, for a particular parameter fails (unless it is an apparent random error or suspected QC deterioration which does not affect the system), at least 02 samples analyzed prior to the QC analysis should be rerun to ensure that the results that were released are acceptable. Check if the difference between the results is within the Quality Requirement for that particular parameter. Report the values in consultation with the Section Incharge/Laboratory Director / Quality Manager.

statistical calculation:

Calculation of Mean: Where Σ = operation of summation or addition of all xi values. Xi = Individual measurement n = The number of observations.Calculation of SD Σ = Sum of X = Average value Xi = Any single observed value n = Total no. of observed values n-1 = Degree of freedom The degree of freedom represents the number of independent data points that are contained in a data set. The mean is calculated first, so the variance calculation has lost one degree freedom.

Calculation of CV: SD CV% = Mean X 100

Acceptability limits: According to CLIA/5 BV/ 4 CAP guidelines

Record: IQC Register, IQC programme chart, L-J file, Excel file for Upadation of Mean and SD, lot verification file, EQAS Report file for Bias

Reference: College of American Pathologists (CAP) 2012 Commission on Laboratory Accreditation, Laboratory Accreditation Program; Laboratory Checklists

1. 2012 (01-04-2012) Chemistry and Toxicology checklist - CHM.12900 “New Reagent Lot Verification”, p.15
2. 2012 (01-04-2012) Hematology and Coagulation checklist – HEM.24575 “New Reagent Lot Verification”, p.27
3. 2012 (01-04-2012) Immunology checklist IMM.41400 “New Reagent Lot Verification, p.39-41
4. 2012 (01-04-2012) POC Checklist-POC.0500 “New Reagent Lot Verification” p. 14
5. DAIDS Guidelines for GCLP Standards: (Version 2.0, 25 July 2011) p17.

• https://resources.psmile.org/resources/process-control/parallel-testing dated 18/03/2020
• NABL 112: Specific criteria for accreditation of medical laboratories,
• www.westgard.com
• CAP Checklist