Lab QC_ Monitoring

Slide 1: 

Laboratory Quality Control Program Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Objectives: 

Objectives Define Lab quality program Setting lab quality program & Design Definition of quality materials Uses of QC material Characteristics of quality control materials Laboratory SPC (Statistical process control) Patients’ data for controlling analytical quality Identifying the sources of lab errors Corrective actions to lab errors Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Laboratory Quality Control Program: 

Laboratory Quality Control Program It is the lab quality monitoring design, as an essential aspect to ensure appropriateness of laboratory quality objectives for data release (Accuracy & Precision) Free Of Errors Result Report Timed Right result for Relevant Test in Right Time by Right Method Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 4: 

Setting up a quality control program Define & select control procedures Set appropriate concentration ranges Design the frequency of control analyses Design the position of control samples in an analytical run Follow the proper guidelines for interpretation of quality control Daily evaluation of quality control Long term evaluation of quality control data Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 5: 

Uses of quality control program Detecting accuracy and precision of the lab analytical process Measurement of uncertainty (detecting analytical errors) Method validation Methods/instruments comparison Personnel comparaisons or compétences Evaluation of proficiency testing Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Criteria of suitable QC materials: 

Criteria of suitable QC materials Stable & Long shelf life (expiry) Suitable size of vials Less lot to lot variability Suitable matrix (protein matrix is the best when serum is the test material) Safe on use (Bovine source) Have normal and abnormal levels of testing Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Quantitative QC Materials: 

Calibrator : a solution which has a known amount of analyte weighed in or has a value determined by repetitive testing using a reference or definitive test method Control : material or preparation used to monitor the stability of the test system within predetermined limits Quantitative QC Materials Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Calibration: 

The process of testing and adjustment of an instrument, kits, or test system, to provide a known relationship between the measurement response and the value of the substance being measured. Calibration Concentration Abs Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 9: 

Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 10: 

Indications for Calibration Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 11: 

Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 12: 

Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Internal quality control design: 

Internal quality control design Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 14: 

Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 15: 

Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Establishing Control Ranges: 

Select appropriate control levels correlated with linearity of the method and the medical decision levels Make sure any procedural variation is represented: different operators different times of day Determine the degree of variability (SD) in the data to establish acceptable range Establishing Control Ranges Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

NCCLS methods for QC data collection: 

The National Committee for Clinical Laboratory Standards (NCCLS) describe several methods for estimating the mean and precision for a control level as following; Method 1: N > 20 (20 or more runs) (QCX1daily RunX1replicateX20days) NCCLS recommends that, at a minimum, 20 data points from 20 or more separate runs be obtained to determine an estimate of mean and precision. Method 2: Provisional Ranges N > 20 (Fewer than 20 runs) (QCX1daily RunX3replicateX7days) If 20 runs cannot be completed, a minimum of seven runs (three replicates per run) may be used to set provisional ranges. A mean and standard deviation can be calculated and used to set provisional ranges. The mean and limits derived from the abbreviated data collection should be replaced by a new mean and limits calculated when data from 20 separate runs becomes available. NCCLS methods for QC data collection Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

NCCLS methods for QC data collection (Continue): 

Method 3: N=80 (40 Runs) (QCX2RunsX2replicateX20days) The most detailed NCCLS-recommended protocol involves running an assay for 20 days, collecting 80 data points. Each level of material is run twice daily in replicates of two. The collected data can then be entered into NCCLS-provided software to determine estimates of within run, between run, between day, and total precision as well as an estimate of the mean. Method 4: N=40 (20 Runs) (QCX1 RunX2replicateX20days) This abbreviated version of the N=80 data collection is also discussed by NCCLS. It makes use of only one run per day of two replicates for a total of 40 data points. NCCLS methods for QC data collection (Continue) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 19: 

20 points 40 points 80 data points > 20 data points Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 20: 

Evaluation of Quality Control Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

SPC (Statistical process control): 

SPC (Statistical process control) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Measurement of Variability: 

A certain amount of variability will naturally occur when a control is tested repeatedly. Variability is affected by operator technique, environmental conditions, and the performance characteristics of the assay method. The goal is to differentiate between variability due to chance from that due to error. Central tendency & Dispersion indices are used to detect normal variability (X, S 2 , SD, CV%) Measurement of Variability Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

: 

Data are frequently distributed about a central value or a central location There are several terms to describe that central location, or the ‘central tendency’ of a set of data The distance between the target and the observation is termed “Bias” Measures of Central Tendency Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 24: 

Indices of Central Tendency Median = the value at the center (midpoint) of the observations Mode = the value which occurs with the greatest frequency Mean = the calculated average of the values Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 25: 

Calculation of Mean X = Mean X 1 = First result X 2 = Second result X n = Last result in series n is Total number of results Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 26: 

Normal (Gaussian) Distribution All values are symmetrically distributed around the mean Characteristic “bell-shaped” curve Assumed for all quality control statistics Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Normal distribution: 

Normal distribution Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 28: 

Normal Distribution Mean Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 29: 

Accuracy and Precision The degree of fluctuation in the measurements is indicative of the “precision” of the assay. The closeness of measurements to the true value is indicative of the “accuracy” of the assay. Quality Control is used to monitor both the precision and the accuracy of the assay in order to provide reliable results. Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 30: 

Precise and inaccurate Imprecise and inaccurate Precision and Accuracy Systematic Error Random Error Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 31: 

Precise and Accurate Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 32: 

D B Target level B D B = Bias from the target D = Dispersion between observations Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 33: 

Measures of Dispersion or Variability There are several terms that describe the dispersion or variability of the data around the mean: Range Variance Standard Deviation Coefficient of Variation Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 34: 

Range Range refers to the difference or spread between the highest and lowest observations. It is the simplest measure of dispersion. It makes no assumption about the shape of the distribution or the central tendency of the data. Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 35: 

Calculation of Variance (S 2 ) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 36: 

Calculation of Variance Variance is a measure of variability about the mean. It is calculated as the average squared deviation from the mean. the sum of the deviations from the mean, squared, divided by the number of observations (corrected for degrees of freedom) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 37: 

Degrees 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 to loose one degree of freedom (n-1) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 38: 

Calculation of Standard Deviation Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 39: 

Calculation of Standard Deviation The standard deviation (SD) is the square root of the variance SD is commonly used (rather than the variance) since it has the same units as the mean and the original observations SD is the principle calculation used in the laboratory to measure dispersion of a group of values around a mean Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 40: 

Standard Deviation and Probability For a set of data with a normal distribution, a value will fall within a range of: +/- 1 SD 68.2% of the time +/- 2 SD 95.5% of the time +/- 3 SD 99.7% of the time 68.2% 95.5% 99.7% Frequency -3s - 2s -1s Mean +1s +2s +3s Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 41: 

Calculation of Coefficient of Variation The coefficient of variation (CV) is the standard deviation (SD) expressed as a percentage of the mean Ideally should be less than 5% Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 42: 

Standard Deviation and Probability In general, laboratories use the +/- 2 SD criteria for the limits of the acceptable range for a test When the QC measurement falls within that range, there is 95.5% confidence that the measurement is correct Only 4.5% of the time will a value fall outside of that range due to chance; more likely it will be due to error Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Flow of centralization and dispersion statistical indices: 

Flow of centralization and dispersion statistical indices Mean Variance Standard Deviation Coefficient of Variation Sigma metric σ = (TE a – Bias obs ) / SD obs Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Evaluation of the control of lab analytical process: 

Evaluation of the control of lab analytical process Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

QC Graphic monitoring: 

QC Graphic monitoring Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 46: 

Statistics use many C ontrol C harts : V ariable D ata Measured & Plotted on a Continuous Scale (Run Chart) n = 1 2 < n < 9 median n is ‘small’ 3 < n < 5 n is ‘large’ n > 10 X & R m X & R X & R X & S Schewart Levey- Jenning Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 47: 

PAT 1 : One point plots beyond zone A on either side of the mean PAT 2 : Nine points in a row plot on the same side of the mean PAT 3 : Six consecutive points are strictly increasing or strictly decreasing PAT 4 : Fourteen consecutive points which alternate up and down Control Chart Interpretation: Pattern Analysis Tests (PATs) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 48: 

PAT 5 : Two out of three consecutive points plot in zone A or beyond, and all three points plot on the same side of the mean PAT 6 : Four out of five consecutive points plot in zone B or beyond, and all five points plot on the same side of the mean Control Chart Interpretation: Pattern Analysis Tests (PATs) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 49: 

PAT 7 : Fifteen consecutive points plot in zones C, spanning both sides of the mean PAT 8 :Eight consecutive points plot at more than one standard deviation away from the mean with some smaller than the mean and some larger than the mean Control Chart Interpretation: Pattern Analysis Tests (PATs) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 50: 

CL U1SL U2SL UCL L1SL L2SL LCL A B C C B A Control Charts: Colors Used * * * * * * * * * * * * * * * Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 51: 

CL U1SL U2SL UCL L1SL L2SL LCL A B C C B A Control Charts: Colors Used * * * * * * * * * * * * * * * 1 Row 9 Shift slope 6 Trend 2 +1 4+1 1+3 15 14 3+1 Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 52: 

Laboratory Monitoring QC Data (Chemometrics) Monitor over time to evaluate the precision and accuracy of repeated measurements Review data at defined intervals, take necessary action, and document Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Schewart X & R control charts: 

Schewart X & R control charts Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait Days CL in X chart= X± 1.05 S CL in R chart= 4.76 X S

Levey-Jennings Chart: 

A graphical method for displaying control results and evaluating whether a procedure is in-control or out-of-control Control values are plotted versus time Lines are drawn from point to point to accent any trends, shifts, or random errors Levey -Jennings Chart Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Levey-Jennings Chart: 

Levey-Jennings Chart Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Levey-Jennings Chart - Record Time on X-Axis and the Control Values on Y-Axis: 

Levey-Jennings Chart - Record Time on X-Axis and the Control Values on Y-Axis Time (e.g. day, date, run number) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Levey-Jennings Chart - Plot Control Values for Each Run: 

Levey-Jennings Chart - Plot Control Values for Each Run Time (e.g. day, date, run number) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Levey-Jennings Chart Calculate the Mean and Standard Deviation; Record the Mean and +/- 1,2 and 3 SD Control Limits: 

Levey-Jennings Chart Calculate the Mean and Standard Deviation; Record the Mean and +/- 1,2 and 3 SD Control Limits Mean Day + 1SD +2SD +3SD -1SD -2SD -3SD Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Levey-Jennings Chart - Record and Evaluate the Control Values: 

Levey-Jennings Chart - Record and Evaluate the Control Values Mean Day +1SD +2SD +3SD -1SD -2SD -3SD Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Findings Over Time: 

Ideally should have control values clustered about the mean (+/-2 SD) with little variation in the upward or downward direction Imprecision = large amount of scatter about the mean. Usually caused by errors in technique Inaccuracy = may see as a trend or a shift, usually caused by change in the testing process Random error = no pattern. Usually poor technique, malfunctioning equipment Findings Over Time Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

When does the Control Value Indicate a Problem?: 

Consider using Westgard Control Rules Uses premise that 95.5% of control values should fall within ±2SD Commonly applied when two levels of control are used but modern rules applied for more than 2 levels of QC Use in a sequential fashion When does the Control Value Indicate a Problem? Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

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” Westgard Rules Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Westgard Rules (Generally used where 2 levels of control material are analyzed per run) : 

1 2S rule 1 3S rule 2 2S rule R 4S rule 4 1S rule 10 X rule Westgard Rules ( Generally used where 2 levels of control material are analyzed per run) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Westgard – 12S Rule: 

“warning rule” One of two control results falls outside ±2SD Alerts tech to possible problems Not cause for rejecting a run Must then evaluate the 1 3S rule Westgard – 1 2S Rule Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

12S Rule = A warning to trigger careful inspection of the control data: 

1 2S Rule = A warning to trigger careful inspection of the control data Mean Day +1SD +2SD +3SD -1SD -2SD -3SD 1 2S rule violation Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Westgard – 13S Rule: 

If either of the two control results falls outside of ±3SD, rule is violated Run must be rejected If 1 3S not violated, check 2 2S Westgard – 1 3S Rule Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

13S Rule = Reject the run when a single control measurement exceeds the +3SD or -3SD control limit: 

1 3S Rule = Reject the run when a single control measurement exceeds the +3SD or -3SD control limit Mean Day +1SD +2SD +3SD -1SD -2SD -3SD 1 3S rule violation Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Westgard – 22S Rule: 

2 consecutive control values for the same level fall outside of ±2SD in the same direction, or Both controls in the same run exceed ±2SD Patient results cannot be reported Requires corrective action Westgard – 2 2S Rule Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

22S Rule = Reject the run when 2 consecutive control measurements exceed the same +2SD or -2SD control limit: 

2 2S Rule = Reject the run when 2 consecutive control measurements exceed the same +2SD or -2SD control limit Mean Day +1SD +2SD +3SD -1SD -2SD -3SD 2 2S rule violation Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Westgard – R4S Rule: 

One control exceeds the mean by –2SD, and the other control exceeds the mean by +2SD The range between the two results will therefore exceed 4 SD Random error has occurred, test run must be rejected Westgard – R 4S Rule Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

R4S Rule = Reject the run when 1 control measurement exceed the +2SD and the other exceeds the -2SD control limit: 

R 4S Rule = Reject the run when 1 control measurement exceed the +2SD and the other exceeds the -2SD control limit Mean +1SD +2SD +3SD -1SD -2SD -3SD Day R 4S rule violation Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Westgard – 41S Rule: 

Requires control data from previous runs Four consecutive QC results for one level of control are outside ±1SD, or Both levels of control have consecutive results that are outside ±1SD Westgard – 4 1S Rule Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Westgard – 10X Rule: 

Requires control data from previous runs Ten consecutive QC results for one level of control are on one side of the mean, or Both levels of control have five consecutive results that are on the same side of the mean Westgard – 10 X Rule Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

10x Rule = Reject the run when 10 consecutive control measurements fall on one side of the mean: 

10 x Rule = Reject the run when 10 consecutive control measurements fall on one side of the mean Mean Day +1SD +2SD +3SD -1SD -2SD -3SD 10 x rule violation Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 75: 

2of3 2s - reject when 2 out of 3 control measurements exceed the same mean plus 2s or mean minus 2s control limit; In situations where 3 different control materials are being analyzed Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 76: 

7 T - reject when seven control measurements trend in the same direction, i.e., get progressively higher or progressively lower. A related control rule that is sometimes used, particularly in Europe, looks for a "trend" where several control measurements in a row are increasing or decreasing: Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 77: 

Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

When a rule is violated: 

Warning rule = use other rules to inspect the control points Rejection rule = “out of control” Stop testing Identify and correct problem Repeat testing on patient samples and controls Do not report patient results until problem is solved and controls indicate proper performance When a rule is violated Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Solving “out-of-control” problems: 

Policies and procedures for remedial action Troubleshooting Alternatives to run rejection Solving “out-of-control” problems Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Corrective Action: 

When quality control results are out of range identify the type of error causing the quality control failure determine the best possible approach to resolve the deficiency. Corrective Action Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

What kind of errors could occur in your lab? : 

What kind of errors could occur in your lab? Random error? Systematic error?

Random Error: 

Random error is defined as imprecision of the test system causing a scatter or spread of control values around the mean. The exact magnitude of random error cannot be predicted. It is estimated by repetitive testing or precision studies. Random Error Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Common causes of random error: 

air bubbles in the reagent pack, reagent lines, sample, or reagent syringes improperly mixed/dissolved reagent pipette tips not fitting properly a clogged pipetter (clot) imprecise pipetter the power supply fluctuations Common causes of random error Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Systematic Error: 

Systematic error is defined as a systematic change in the test system that displaces the mean of the distribution from its original value. Systematic error of an analytic system is predictable and causes shifts or trends on control charts. Systematic Error Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Common causes of systematic error : 

a change in the reagent or calibrator lot number wrong calibrator values improperly prepared reagents deterioration of reagents or calibrators inappropriate storage of reagents and calibrators change in sample or reagent volumes due to pipetter maladjustments or misalignments change in temperature of incubators and reaction blocks deterioration of a photometric light source change in procedure from 1 operator to another. Common causes of systematic error Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Periodic Review: 

Periodic review of control charts to check assay performance is essential and detects problems such as shifts, trends, random errors, imprecision, and outliers. Periodic review can include such elements as: daily review of control values and control charts by the technologist before accepting assays weekly or monthly review of quality control data and charts by the supervisor or director of the laboratory periodic audits Periodic Review Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

CLIA recommended Corrective actions : 

CLIA requires procedures to manage an out-of-control situation; Some possible control procedures include: Review procedures used Search for recent events that could cause change such as: New reagent kit or lot New control bottle Instrumentation component replacement Instrument maintenance Instrument move Examine the environmental conditions (temperature, humidity, etc.) Prepare new control materials Follow manufacturers troubleshooting guide Contact manufacturers of: Instrumentation Reagent materials Control materials CLIA recommended Corrective actions Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Other QC data analysis methods: 

Other QC data analysis methods Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 89: 

“Q” TEST FOR DATA REJECTION Q (90%) n 0.76 4 0.64 5 0.56 6 0.51 7 0.47 8 0.44 9 0.41 10 Results 1 Results 2 42.45 42.45 42.67 42.67 42.21 42.21 41.98 41.98 43.55 44.21 gap 0.88 1.54 range 1.57 2.23 Q calc 0.56 0.69 Q table (n = 5 ) 0.64 0.64 it is possible to reject the 44.21 value in Result 2 since Q calc > Q table at the 90% confidence level Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Q test for data rejection It is used in selecting data of QC that is used in re-calculation of the mean Recalculation of the mean in your lab is a recommendation by NCCLS: 

Q test for data rejection It is used in selecting data of QC that is used in re-calculation of the mean Recalculation of the mean in your lab is a recommendation by NCCLS Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 91: 

Cumulative SD (Lot to Date) deviation Where; - n t (   ∑ x i 2 ) t is the total of the sums of all the squared individual values, - ( ∑  x i ) 2 is the square of the total of the sums of all the individual values, - n t is the total number of measurements in the time period of interest. Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 92: 

This is a long term estimate of the central tendency observed for a control material based on a large a number of control measurements collected over a long period of time. A long period here is at least two months and could be several months, even a year. Changes in the accuracy of a method could lead to shifts in the mean observed for a control material. Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Example: 

Example Month Monthly total (cumulative total) Calculated statistics Control Limits n ∑ x ∑ x 2 Mean s Mean +/- 3s 1 20 3983 793465 199.15 3.63 188.3 - 210.0 2 20 3993 797537 199.65 4.20 187.1 - 212.2 (40) (7976) (1591002) (199.40) (3.86) (187.8 - 211.0) 3 20 4002 801138 200.10 4.22 187.5 - 212.7 (60) (11978) (2392140) (199.63) (3.97) (187.7 - 211.6) 4 20 4020 808182 201.00 2.92 192.2 - 209.8 (80) (15998) (3200322) (199.96) (3.77) (188.7 - 211.3) 5 20 3995 798259 199.75 3.68 188.7 - 210.8 (100) (19993) (3998581) (199.93) (3.73) (188.7 - 211.1) Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Z-score & DI: 

Z-score & DI It is calculated by taking the difference between the control result and the expected mean (Bias), then dividing by the standard deviation observed for that control material. For example, if a control result of 111 is observed on a control material having a mean of 100 and a standard deviation of 5, the z-score is 2.2 [(111- 100)/5]. A z-score of 2.2 means that the observed control value is 2.2 standard deviations from its expected mean, therefore this result exceeds a 2s control limit but not a 3s control limit. Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 95: 

Is z-score useful? It is very helpful to have z-scores when you are looking at control results from two or more control materials at the same time, or when looking at control results on different tests and different materials on a multitest analyzer. You can quickly see if any result exceeds a single control limit, for example, a z-score of 3.2 indicates that a 3s control limit has been exceeded. You can also look for systematic changes or trends occurring across different control materials, for example, consecutive z-scores of 2 or greater on two different control materials. Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 96: 

There is similarity between the calculation of the SDI and the z-score. They're basically the same thing, but the z-score tends to be used in internal QC programs to compare an individual QC result with the expected values for that material, whereas the SDI tends to be used in external QC programs to compare the performance of the lab with the overall mean for a defined comparative group or with an established target value. Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

SDI (or z-score) interpretation: 

SDI (or z-score) interpretation Excellent < 0.5 Satisfactory < 0.5 – 1.0 Acceptable 1.0 – 2.0 Alarming (Calibration check) 2.0 – 3.0 Rejected > 3.0 Apply Westgard’s rules for alarming zone generated by calculating Z-score or DI In PT, sum of last 6 DI results are multiplied by 6 A score of PT-DI > 100 indicates unsatisfactory performance Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 98: 

Cumulative sum control chart . The cumulative sum control chart is a more sensitive control chart that can use information from an entire set of points to draw conclusions about the process.  Basically the cumulative sum (or cusum) chart plots the cumulative sum of measurement deviations from an average.  Therefore, if an abnormal amount of measurements fall on only one side of the average this sum will grow and indicate a systematic out-of-control condition. Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Cusum chart: 

Cusum chart Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait Tabulate the differences between the observations with preserved mathematic signs – or + 2. Plot the differences with time run A steep slope suggests systematic error Undulating values around the mean denotes variability Zeroing of cusum denotes stable method

Slide 100: 

Exponentially weighted moving average (EWMA) control chart . This chart is similar to the cumulative sum chart but instead of weighting each measurement the same, recent measurements are more influential because measurements are weighted exponentially based on when they were analyzed. This is a sophisticated but more indicative for quality control Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

QC software development: 

QC software development Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

OPSpecs chart: 

OPSpecs chart Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Real time QC plot (Youden Plot): 

Real time QC plot (Youden Plot) Y X Y X Y X 10 x 2SD 3SD >2 2s R 4s - 2s +2s Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

OUT WITH/WITHIN CONSENSUS METHOD: 

Widely used in coagulation studies For PT/INR, PT, APPT; the DI % from the MEDIAN is calculated Within consensus performance if DI % is less than 15 % Out with consensus if DI % is more than 15 % OUT WITH/WITHIN CONSENSUS METHOD Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Ranking Zones: 

Group Group A 25 % of results adjacent to & above the median + 25 % of results below the median Group B Next 10 % on each side of A Group C Next 5 % on each side of B Group D Next 5 % on each side of C Group E Next 5 % on each side of D Unsatisfactory when assessed two consecutive runs D - D E – E D – E E – C Used in assessment of factor assays in coagulation studies Ranking Zones Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Is QC analysis enough for your judgment about your lab performance?: 

Is QC analysis enough for your judgment about your lab performance? Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Patients’ laboratory data guide the lab performance: 

Patients’ laboratory data guide the lab performance Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Control the analytical quality using patients’ data: 

Control the analytical quality using patients’ data Single patient data Clinical correlation Lab results correlation Intra-lab duplicate Ladenson third day delta checks Multiple patients’ data Test distribution (Scattering & Spanning the measuring levels) AON (Average Of Normals); the daily Mean or Median Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

1. Clinical correlation It allows identification of impossible test results that are do not correlate with relevant clinical data or diagnosis Examples - Normal bilirubin in severely jaundiced patient - Low ferritin in BTM with transfusion - Normal creatinine in ESRD on HD 2. Correlation with other laboratory tests: Pattern Recognition It depends on combinations of multiple related results of the same patient which raise the possibility of analytical error Examples - Marked increase of creatinine with normal BUN - Marked ALT elevation with normal AST : 

1. Clinical correlation It allows identification of impossible test results that are do not correlate with relevant clinical data or diagnosis Examples - Normal bilirubin in severely jaundiced patient - Low ferritin in BTM with transfusion - Normal creatinine in ESRD on HD 2. Correlation with other laboratory tests: Pattern Recognition It depends on combinations of multiple related results of the same patient which raise the possibility of analytical error Examples - Marked increase of creatinine with normal BUN - Marked ALT elevation with normal AST Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

3. Ladenson Delta checks: 

3. Ladenson Delta checks Analysis of a patient sample and correlate the results with previous results. It could be done on volunteers The expected variability depends upon; Time course of the analyte behaviour Time interval between two samples (3 – 5 days) Intra- individual variation Test Delta limit Calcium, total 15 % Creatinine or BUN 50 % CK 99 % Hb 10 % WBC count 20 % Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

4. Arithmetic checks: 

4. Arithmetic checks On analysis of two or more related tests from the same specimen Examples Anion gap exceeds the range of 10 – 20 Osmolal gap 5. Limit checks Results those exceed the limits that incompatible with life Examples - Dropped digits e.g. Sodium 42.0 instead of 142.0 mmol/L - Misplaced decimal point e.g. Potassium 40.0 instead of 4.0 mmol/L Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 112: 

6. Bull’s group means algorithm Computerized algorithm calculating and plotting or comparing the means of inter-group data each is formed of 200 – 500 samples 7. Split sample Bias Testing 10 samples in duplicate and the difference between each other to be less than 2 SD Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

General scheme for QC evaluation: 

General scheme for QC evaluation Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Levels of QC verification: 

Levels of QC verification Operator level ; Check Levey Jennings chart QC supervisor ; Analyze daily QC data, give decision for acceptance/rejection QC officer (coordinator); Documentation of QC management HOD; Method, calibration, QC, SPC and directing the QC plan and improvement Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait

Slide 115: 

Dr. Mahmoud A Abdelwahab YIACO- RNMLC Adan- Kuwait Thanks