click below

click below

Normal Size Small Size show me how

# Labops8

### Week 8: Quality Assurance and Quality Control: Part I

Question | Answer |
---|---|

Quality assurance | 1.is an overall system of checks and balances. 2.All actions ...ensure quality patient results. 3. Begins before the sample is collected and ends with the result being seen, interpreted and acted upon by the physician compasses three phases: |

Quality control | one component of the quality assurance process and involves the analytical phase. It is designed to recognize and minimize analytical errors in specific procedures to prevent reporting erroneous patient results. |

Standard | substance with a known concentration |

Medical decision level (during QC) | levels at which a physician may implement or adjust medications or take other actions |

What is Standard Deviation (aka precision)? | statistics that quantifies how close numerical values (i.e , QC values) are in relation to each other. |

What is imprecision? | used to express how far apart numerical values are from each other |

Acceptability of a reported CV | An acceptable limit for control data is < 5% For example, if the CV is 2.8% for a set of control data points, it indicates the size of the SD in relationship to the mean is acceptable and therefore the distribution is tight around the mean. |

Describe what steps must be performed before a new lot number of a control is used in the laboratory | |

Explain the use of Levey-Jennings charts | a graph that quality control data is plotted on to give a visual indication whether a laboratory test is in a statistically appropriate fashion... visual representation of day-to-day variations over a one month period. |

Westgard multirule system for monitoring quality control. | Also known as multirule Dr. James Westgard developed this system in 1981 to evaluate analytical run quality for medical lab testing. There are six basic rules that are used individually or in combination to evaluate the quality of analytical runs |

Which Westgard rule is violated, if One QC value is 3 standard deviations (SD) away from the mean. | 13S |

Differentiate between internal and external quality control. | required by lab QA, QC ... Proficiency testing(required by accrediting agencies and by CLIA'88. An outside agency (e.g., CAP) provides unknown samples to be analyzed for an annual fee) |

the ability to obtain a true value to verify the correctness of a result | accuracy |

the ability to obtain the same value for repeat measurements of a sample over time | precision |

shift pattern? | Values abruptly change and continue in this pattern. May be due to: introduction of something new into the procedure slight malfunction in instrument that results in immediate and somewhat permanent change in performance other possible causes |

trend pattern? | Values gradually change and continue in this pattern. May be due to: deterioration of standards and/or reagents aging of light source (bulb) in an instrument other causes |

Random error vs Systematic error | Random:shows no real pattern statistically expected characterized by a single "out" value Systematic:continuous pattern affecting all results shift or trend appearance |

Preanalytical components of quality assurance | includes tasks involved before the sample is actually tested in the laboratory. -Physician orders and their implementation -Specimen identification -Lab accessioning... |

Analytical components of quality assurance | includes processes within the lab itself that involve the actual testing of the specimen -Preparation of the specimen for testing -Testing of the specimen -Performance and interpretation of quality -Control on reagents and instrumentation |

Postanalytical components of quality assurance | postanalytical - includes processes between completion of the analytic process and receipt and follow-up by the physician. -Reporting of results -Receipt and review of results by physician... |

Possible causes for changes ( shift, trend and/or imprecision ) on Levey-Jennings chart | When an analytical process is within control, aprox. 68% of all QC values fall within ±1SD( LJ chart), 95%...±2SD About 4.5% of all data will be outside the limits(±2SD limit) when the analytical process in in control. |

Calculate the reference range for a control when provided with pertinent data. | |

Non Analytical factors of QA (7) | 1.Qualified personnel 2.Established laboratory policies 3.Laboratory procedure manual 4.Test requisitioning 5.Patient identification, specimen procurement, and labeling 6.Specimen transportation and processing 7.Preventive maintenance of equipment |

Error Analysis (Two types of errors) | Latent errors [ Equipment malfunctions (e.g., old error-prone analyzers),Teamwork factors...] and Active errors[Errors with collection tubes,Errors with transportation system (e.g., pneumatic tube),pt ID errors] |

Quality Control (QC) | one component of the quality assurance process and involves the analytical phase. It is designed to recognize and minimize analytical errors in specific procedures to prevent reporting erroneous patient results. |

Quality Control (QC) | involves running and evaluating controls to monitor test performance |

Control specimen | is material or solution with a known concentration of the analyte being measured |

Controls should: | - preconstituted (commercially manufactured controls are usually used) - similar to patient samples.Run alongside patient samples -include at least 2 medical decision levels -have the same lot# for running over extended periods... |

Evaluating Controls | A statistical analysis of variability( results of different techs) is necessary to evaluate controls. Controls must be assayed for a period of time to ensure enough data points are gathered. |

Evaluating Control Validity (run once a day or per shift for a month.) | A minimum of 20 data points are needed to ensure validity - controls are usually run once a day or per shift for a month. |

Quality Control Statistics | Measures of Central Tendency: mean, median, and mode Measures of Dispersion: range, standard deviation, and coefficient of variation Sources of variance or error: Sampling factors / Procedural factors |

Central Tendency | refers to the concept of the clustering of data points around one value |

Frequency | the number of times a value appears |

What does a peak in a frequency plot represent? | the value most often obtained |

If points to the right of the peak are about equal in number to points to the left, data is said to | have normal (Gaussian) distribution |

the mean, median, and mode are approximately the same value. | normal distribution |

Measures of Dispersion | once central tendency has been proven, the dispersion or spread of data around the point of central tendency must be considered. There are several ways to evaluate dispersion to include range and standard deviation |

The mean | is the average of all data points (values). It is the laboratory's best estimate of the analytes true value for specific level of control |

In the following set of data points {1, 2, 3, 4, 4, 5, 5, 5, 6, 6, 6, 6, 7, 7, 7, 8, 8, 9, 10, 11, 12}, what is the median value? | 6 |

The mode | is the value occurring with the greatest frequency. |

In the following set of data points {99, 99, 100, 100, 100, 101, 101, 101, 101, 102}, what is the mode? | 101 |

RANGE ( a measure of dispersion) | is the difference between the highest and lowest values. Example: If the low value in a set of data points is 96 and the high value is 106, the range would be 10. |

Standard Deviation (a measure of dispersion) formula | SD= Sqrt( ∑(x1 – x) 2/(n – 1 )) Where (x1 – x) 2 = each data point minus the mean squared; n – 1 = number of data points minus one degree of freedom. set of #(data points) x1, x2...xn |

Acceptable intervals for lab controls | set at ± 2SD that is 95.5% within the mean. 4.5 out of 100 control runs may be outside this interval for a normal value – So, it is normal to have one control occasionally outside the acceptable interval. |

COEFFICIENT OF VARIATION (CV) | describes the degree of variation (dispersion from the mean) for a series of values. |

5 of six basic Westgard rules that are used individually or in combination to evaluate the quality of analytical runs | 1. 12s when one control value > ±2SD (warning) 2. 13s when one control value > ± 3SD (reject) 3. 22s when 2 consecutive control values are on the same side of the mean and > +2SD...(reject) 4.41s // >+1 or mean - 1 SD limits (reject) 5 10x // reject |

CV formula | (SD/X) * 100 where x is the mean |

Repeatedly of a test is consistent | low standard deviation/ low imprecision |

Repeatedly of a test is inconsistent 9 due to chemicals involved or malfunction) | high standard deviation/ high imprecision |

A method may be precise in that repeat measurements are nearly the same, but if the repeat measurements are not the true value, they will be inaccurate. T or F | True |

Questions to consider when examining control data: | Are the same number of values above and below the mean? Are the controls rising or falling? Is there a sudden change? How many values are outside ± 2SD? Any outside of ± 3SD? Is there precision? |

Precision versus Imprecision * understand sd charts | indicated by a tight group of values above and below the mean indicated by a wider distribution of values above and below the mean |

Charting Method On the x-axis, the date/times (or the number of the control run) are plotted. On the y-axis, the values are plotted. Lines run across the graph at the mean and at 1 SD, 2 SD, and 3 SD (standard deviations) on either side of the mean. | Charting method for Levey-Jennings chart |

The numerical system proposed by Westgard has gained the most favor and involves the use of Control Rules to determine acceptability of control values. T or F? | True |

What are decision limits? | ±1SD, ±2SD, ±3SD from the mean on y-axis |

Calculate the decision limits ±1SD and ±2SD ranges given the following data: mean= 4.1mmol/L and SD= 0.1mmol/L | 4.1 - 1.0(1) = 4.0 4.1 + 1.0(1) = 4.2 therefore ±1SD range is 4.0 to 4.2mmol/L |

Some labs consider values out of the limit ±2SD to be out of control. If only one control value falls out of this limit but within ±2SD and ±3SD, the laboratory should discard the results. Yes or No ? | No. Because 4.5 % of all valid values are between ±2SD and ±3SD limits. |

a lab that uses a ±2 standard deviation limit, typically reject good runs. T or F? | True. Patient results are unessarily repeated, leading to cost amd waste of time |

The reference interval (or, reference range) is a set of test values that | The reference interval (or reference range) is a set of test values that reflects a healthy population of patients. |

Two pipettes are used to pipette patient specimens and QC material at the same volume. Pipette A has a Coefficient of Variation (CV) of 2.3 %. Pipette B has a CV of 1.7%. | pipette B has better precision because its CV is lower than the CV of Pipette A. |

Given QC data on a Levey-Jennings chart,When six or more successive QC results fall on one side of the mean, What has happened? | A shift |

The component of Quality Assurance that compares results from peer labs run on the same sample is called? | Proficiency Testing ( note it is an external quality assurance assessment) |

True or False. Patient results within the reference range can be reported when the quality control result is outside the control range. | False. The results of patient specimens can be accepted and reported only if the control results are within the control range, or "in control". Confidence in the accuracy of the patient result is essential. |

Which of the following studies is NOT part of a Method Validation for automated analyzers? 1.Carryover 2.Method comparison 3.Reproducibility study 4.Result verification | 4.Result verification |