Chemistry QC, Calibration&Reportable Range REFINED.pptx

laboratory chemistry calibration and linearity

1. Explain the role of quality control testing in ensuring accurate patient results. Discuss calibration and calibration verification, identifying when each procedure is required. Define analytical measurement range and describe which actions need to be undertaken when a result exceeds this range. Identify the requirements for comparing results if the laboratory uses more than one instrument/method to test for an analyte.

2. Q uality control (QC) refers to samples with known values that are tested as if they were patient samples. QC testing helps assure us that a test method or instrument is working as designed. However, traditional quality control testing is generally not able to assess the preanalytic phase of the testing process. QC results that are within expected ranges tell the operator that the assay is performing within established parameters. Since QC is a critical measure of a system's performance, when QC is out of range, patient testing should not be performed until the problem is resolved. At least two levels of control materials should be tested daily for every laboratory assay. Quality Control (QC )

3. Built-in electronic controls or onboard controls are sometimes used within instruments, these can be used as an alternative QC method and must be tested  at least  as often as required by the manufacturer. Test methods with built-in controls require an extra step for laboratory compliance; they must be part of a written Individual Quality Control Plan (IQCP) if QC is not performed daily and with each new lot and shipment. In addition to testing daily QC samples, laboratories must also run at least two levels of external liquid controls each time a new lot of reagent is used and each time there is a new shipment of reagents. This allows the technologist to determine if new lots are performing the same as the old lot. A lways check the manufacturer's requirements when setting up QC for a new assay. Quality Control (QC )

4. If blood gas analysis, at least one level of QC must be tested every eight hours that patient testing occurs. Over each 24-hour period of testing, control materials for pH,  p CO 2 , and  p O 2  must be run and must cover both high and low ranges. For example, if a low control was used in the morning, when QC testing is again performed eight hours later, a high control should be used so that both levels are covered during a 24-hour period. QC results must be documented. If this is not an automatic function of the instrument's data management system or laboratory information system (LIS), this may require the entry of results onto a worksheet, retaining the printout from an instrument, or manual entry of results into the LIS or QC software. Quality Control (QC )

5. Quality Control for Qualitative Chemistry Tests Some chemistry test results, such as results from a drugs-of-abuse screen, are reported qualitatively rather than quantitatively. However, even though these results are a simple "positive" or "negative," the test result itself is still based on a quantitative measurement within the instrument. In the case of a drug screen, there are concentration cutoffs assigned to samples such that when a drug is present at, or above the cutoff, the result is released as "positive." In these types of qualitative tests, the cutoff concentration should be assessed and calibrated every six months. This is usually done using materials provided by the manufacturer for this purpose. The material is usually referred to as a calibrator but may be a control that is specifically designated for this purpose .

6. Daily QC for qualitative assays like this should assess or 'challenge' the cutoff value. If a QC value is very high or very low, it provides little challenge for the test. Instead, choose controls that are within +/- 20% of your cutoff value. Using controls that are only 10% from cutoff concentrations provides an even more rigorous control. Such controls will be much more sensitive to fluctuations in the assay's performance and will quickly reveal a problem when there is one present. There is always a balance to strike between having rigorous controls and yet, avoiding specifications that are so tight that QC fails too often and hurts the workflow of the laboratory, slowing the release of patient results. In addition to daily QC, the cutoff value should be re-verified: When there is a change of reagent lot After replacement of major instrument components After major service to the instrument If QC fails to meet established criteria Quality Control for Qualitative Chemistry Tests

7. Quality Control (QC) Failures If QC values are not acceptable, corrective action (troubleshooting), must be undertaken. These actions should be defined by the laboratory's quality control procedure. Some typical steps are: Repeat testing on a fresh vial of control. Check instrument (verify daily and weekly maintenance was performed). Recalibrate the assay or verify the calibration. Recalibrate with a different lot of calibrator. Contact the instrument vendor for service. For quantitative tests that fail QC testing, calibration or calibration verification may be required. Once an assay is recalibrated you should verify the assay once again. Some ways to do this include: Using test materials that are supplied by the manufacturer of the test system for the purpose of calibration verification. Measure the current calibration materials as unknowns (run them as samples, not as calibrators). Retest previously reported (and appropriately stored) patient samples that were tested during a run when QC was acceptable. Documenting corrective actions

8. When QC issues have been corrected, the person responsible for taking the necessary actions needs to document what was done to correct the problem. confirming that the problem has been corrected. Verifying patient results Patient samples will need to be re-tested following correction of the problem that caused the QC failure. If patient results were released during a bad run, it is important to repeat the patient samples once the assay is back in control. If the new patient results are significantly different than the results that were originally reported, then the patient results must be corrected in the patient's medical record. Clinicians may also need to be directly notified depending on the severity of the differences. The laboratory's medical director should be involved in the decision to revise patient results and inform clinicians of the error. Quality Control (QC) Failures

9. A delta check is a quality control (QC) tool that compares laboratory test results with results obtained on previous samples from the same patient. A delta check is programmed into the laboratory’s computer/information system (LIS) and can help detect some patient testing errors. Delta checks are particularly useful for detecting errors in specimen identification, specimen integrity, errors in manual data entry, or possible analytical errors. These are quality issues that usually cannot be detected by testing QC materials . For most tests, it is unlikely that consecutive results obtained on one patient will vary significantly unless a substantial change has occurred in the patient’s medical status. However, if multiple delta checks fail on several tests performed on a single patient, there is a strong possibility that the patient or specimen was misidentified . Delta Check

10. Proficiency testing (PT) is an external QC measure. It is used to evaluate a laboratory's testing performance by comparing them to peer laboratories. Unlike QC, PT samples have unknown values. The laboratory must analyze the sample(s) and submit their result(s) to the PT provider. The PT provider will then evaluate the laboratory's results and compare them with the results from other institutions that run the test (these are known as 'peer labs'). Laboratories must perform within acceptable limits on PT testing. Failure to obtain acceptable PT results can have severe consequences. Under CLIA regulations, repeated failures a laboratory will be unable to continue reporting the analyte in question. In PT testing the sample must treated exactly as they would a patient sample. Under no circumstances can the PT sample be sent to another laboratory for testing or verification, even if that would be done for a patient sample. Proficiency Testing (PT)

11. Calibration Calibration refers to the process of testing and adjusting an instrument or test system to establish a correlation between the measurement response and the concentration or amount of the substance that is being measured by the procedure. In short, it determines the relationship between the reagent system/instrument response and the corresponding concentration of an analyte. The instrument or method manufacturer establishes the minimum frequency for calibration. However, the laboratory may require more frequent calibration. Automated systems have become increasingly stable and calibration is usually not required as frequently as it once was. A notable exception to this are blood gas analyzers that typically auto-calibrate at least once every 30 minutes of use . Calibration and Calibration verification

12. Calibration verification Calibration verification refers to the testing of materials of known concentration in the same manner as patient samples in order to substantiate the instrument or test system's calibration across the reportable range of patient test results. In short, it is the process of checking that the calibration that is in use is acceptable for the range of patient results that are being released. The College of American Pathologists (CAP) uses the term calibration verification to mean a process that confirms calibration settings are still valid, but then uses a different term, "analytical measurement range validation," to refer to the process that verifies the reportable range. Calibration and Calibration verification

13. Some suggested methods for calibration verification include: Purchase of assay calibration verification materials from the test manufacturer. Use your current test calibrators but run them as unknowns and verify that the correct target values are recovered. However, if this procedure is done immediately following a calibration, one should not use the same calibrators that were just used for calibrating the instrument. Use a different set or a different lot of calibrators. Retest previously reported, properly stored patient samples that were tested during a period when QC was acceptable. Calibration and Calibration verification

14. Calibration verification is the process of verifying agreement between calibrators (or other materials of known analyte concentrations), and their measured values. To verify calibration is to establish that an instrument or test system's calibration is correctly set throughout the reportable range for patient test results. Linearity, on the other hand, does not require knowledge of the actual analyte concentration. Interestingly, the term “linearity” does not appear in CLIA documents . Calibration Verification: Confirms the accuracy of your measurement of patient samples by proving that the values you receive are what you expect to receive . Linearity: Along with proving measurement accuracy, linearity verifies that the assay is linear and therefore not a curved relationship. While most clinical laboratory assays are linear with respect to analyte concentration, assays are not required to be linear (they can use nonlinear regressions in their calculations). However, since the vast majority of laboratory assays are linear, verification of the linear relationship between instrument signal and specimen concentration is often assessed during the initial validation of a method. So, while it may be appropriate to sometimes check the linearity of an assay, this is not the same as performing a calibration verification. Calibration Versus Linearity

15. Calibration is to be performed at whatever frequency is required by the manufacturer. Typically, this is every 30 days or with each new lot of reagent. Calibration verification is usually performed every six months. However, calibration and calibration verification may be needed at other times. These activities may also be required when: There is a reagent lot change (especially when QC values shift). QC fails to meet established criteria (as a troubleshooting step). After major instrument maintenance or a significant change to the instrument (such as relocating the instrument).

16. The analytical measurement range (AMR) is the range of concentrations that an instrument can measure without any pretreatment of the sample (eg, dilution or concentration). The AMR is established at the time of method validation and is determined using matrix-appropriate materials with known values that verify the reportable limits of the test . The AMR must be checked or verified at least every six months, using a minimum of three samples that span the measurement range. Five samples spanning the range is better practice; the more samples or points you have, the more assured you can be that the assay is consistent (accurate) across its measuring range . It is essential that the AMR results be graphed, and a regression analysis performed. Regression calculation allows the user to see how well (usually this means how linearly), the assay is performing. The intent of the AMR verification is to assess the linearity of the assay; to determine how "true" the assay performs over its entire range.

17. Although the terms 'AMR' and 'calibration verification' are sometimes used interchangeably, they are different processes. The College of American Pathologists' (CAP) differentiate the two; calibration verification aims to detect how "true" a result is. It is assessing whether the value arrived at is what the calibration would expect it to be. AMR verification , on the other hand, assesses if that trueness extends over the entire range of the assay. AMR is similar to linearity; however, linearity is a narrower term that just refers to how (or whether) linear the assay is ( ie , how good the linear regression and correlation numbers are). The term reportable range is yet another term that is sometimes used synonymously with AMR. It's best to think of the term 'reportable range' as referring only to the range of values that a laboratory will report for an assay. For example, a laboratory may not report urine morphine levels above 5,000 ng/mL, even though their morphine confirmation method is able to give a result up to 7,500 ng/ mL. In this example, the AMR would go up to 7,500 ng/mL but the reportable range would only go up to 5,000 ng/ mL. More commonly, the reportable range exceeds the AMR because laboratories can perform dilutions. An assay may only have an AMR that goes to 2000 ng/mL, but with dilutions, a technologist may be able to report a much higher number .

18. How does one perform an AMR verification? Calibration or calibration verification materials can be used if there are at least three levels and if the values of the materials extend across the entire AMR. This means having at least three samples; one that is near the lowest point of the AMR, a midpoint value, and a value that is near the highest level of the AMR. According to the CAP, the guidelines (and thus concentrations,) chosen for the AMR study should be determined by the laboratory medical director who considers the medical usefulness of low and high result values. The decision tree on the following page summarizes what steps need to be taken to ensure the analytical measurement range is being verified adequately and at the required frequency.

19. Calibrate at least every 6 months? Perform calibration verification at least every 6 months Use at least three calibrators? Are the materials used for calibration or cal. verification near low, midpoint and high values of the AMR Perform AMR verification using at least 3 materials of known value near low, midpoint, and high values of the AMR AMR has been verified Yes No No Yes Yes No

20. As a general rule, one should find samples that are within 10% of the concentration of the highest range point of the assay and within 1% of the lowest detectable concentration of the assay. Using a sample with zero concentration of the analyte is also useful. Often times it is clinically useful to extend the reportable range of an assay's linearity. This can be done by using a dilution (to extend the upper range) or by concentrating samples (to extend the lower range ). However, since the AMR, by definition, refers to the range of the assay in the absence of pretreatments, the AMR will remain the same, even if dilutions or concentration steps are used to extend the reportable range. An assay that has had its analytical range or reportable range extended becomes a laboratory-developed test.

21. Never report a patient result that is lower or higher than the analytical measurement range (AMR). 'But can't dilutions be used to extend the reportable range and thus the AMR?' No, remember that the AMR does not change with dilution; a dilution, if performed, must still bring the concentration of the sample into the AMR. If a pretreatment can be performed which brings the measurement back into the AMR, the result can be released (of course, the result must be mathematically adjusted for the dilution or concentration). In some instances, an analyte simply can't be diluted or concentrated enough to bring the result into the validated AMR range. It is also possible that the manufacturer or the laboratory has defined a maximum dilution that may be used. In these cases, it is necessary to report the results as "greater than" or "less than" the limits of the AMR. Most general chemistry analyzers have automated dilution options that will perform a dilution and calculate the concentration of a diluted specimen.

22. Be certain that you know what diluent to use, how to make a proper dilution, and how to calculate the final reportable result. It may also be good practice to have another technologist verify your result. Examples of dilutions are shown below: 1:10 dilution = 20 µL of sample + 180 µL of diluent 1:15 dilution = 20 µL of sample + 280 µL of diluent 1:20 dilution = 10 µL of sample + 190 µL of diluent 1:100 dilution = 10 µL of sample + 990 µL of diluent When making a dilution, it is important to know the approximate range of the AMR into which the diluted result should fall. Dilution of a sample into the extreme low end of the AMR is usually less precise than diluting the sample into the central region of the test's range. The CAP requires that a laboratory's dilution protocols be clearly described for each assay that allows for a dilution. The actual steps (volumes and diluents) need to be part of the operating procedure for the assay.

23. If the laboratory uses more than one instrument or method to report the same analyte & located at the same CLIA address, these methods must be checked against each other twice per year. C omparisons between such instruments helps to ensure that similar results are reported no matter which instrument is used to measure the analyte. In the United States, the Clinical Laboratory Improvement Amendments (CLIA) requires checks for comparability of results at least twice a year. This applies to tests performed on the same or different instrument makes/models or by different methods, but only when the instruments/reagents are producing the same reportable result ( ie , have the same reference intervals). If your laboratory is accredited by the College of American Pathologists (CAP), the twice per year comparison requirement applies to nonwaived instruments/methods accredited under a single CAP number.

24. The brand or model of the instrument is not important; if the same analyte is reported, then the two methods must be compared. For example, if pCO2 is performed using a laboratory blood gas analyzer made by vendor A and the same laboratory or hospital also measures pCO2 with a point of care device made by vendor B, and the surgical area has their own blood gas instrument made by vendor C, then all three of these instruments must be compared to each other if the testing is being done at the same address (or under the same CLIA license). It needs to be established that all three results are within established comparability limits . The criteria for comparability are determined by the laboratory director as are the number of samples to test. Most instruments should give a result within 10% or less compared to another instrument. However, the criteria will vary depending on the analyte and method . Quality control material can be used in some cases as a substitute for pooled specimens. If the same QC sample is used across the instruments, this can suffice as the comparison sample .

25. Bishop ML, Fody EP, Schoeff LE. Clinical Chemistry Principles, Procedures, Correlations. 8th ed. Philadelphia, PA: Lippincott Williams & Wilkins: 2018. College of American Pathologists. All Common Checklist. Northfield, IL: College of American Pathologists: 2021. College of American Pathologists. Chemistry and Toxicology Checklist. Northfield, IL: College of American Pathologists: 2021. College of American Pathologists. Laboratory General Checklist. Northfield, IL: College of American Pathologists: 2021. Miller WG, Sandberg S. Quality management. In: Burtis CA, Bruns DE. Tietz Fundamentals of Clinical Chemistry and Molecular Diagnostics. 8th ed. United States of America: Elsevier Saunders; 2019: chap 7. Westgard J. Westgard Rules. Westgard QC. https://www.westgard.com/westgard-rules.htm. 2009. Last accessed August 8, 2022. Westgard JO. Basic QC Practices. 3rd ed. Madison, WI: Westgard QC Inc ; 2008. References

26. Quiz A Medical Laboratory Scientist has installed a new urine drug screen on her immunoassay analyzer. What should be her starting assumption about the frequency of running control samples ? Please select the single best answer QC samples must be run at least daily. QC samples should be run monthly. QC samples that span the range of the assay must be performed during the initial validation but then afterward, only as needed. QC samples should be run with every batch of patient testing. The minimum requirement for QC is daily testing. Manufacturers will outline the specific number of controls and may require more controls to be run than is typical. Monthly QC is rarely permitted by manufacturers of non-point-of-care assays and should not be a starting assumption. QC that challenges the range of an assay (or the clinical decision point range of assay) is not a bad idea, but it is not usually part of an initial validation. Instead, patient samples, not QC, that span the range are used to compare the new assay. Running QC with every batch of patient testing (such as with 'every rack') is sometimes required but is quite rare due to the inefficiency and the cost that such frequent QC would require. Batched QC should not be a starting assumption, daily QC is the most typical frequency, is the CLIA minimum, and is the starting assumption.

27. Quiz A Medical Laboratory Scientist working in a toxicology laboratory is looking for controls for a urine oxycodone assay. The cutoff for the assay is set at 100 ng/ mL. Which set of controls would be best for the daily QC of this assay? Please select the single best answer Pos = 300 ng/mL, Neg = 0 ng/mL Pos = 1000 ng mL, Neg = 0 ng/mL Pos = 115 ng/mL, Neg = 85 ng/mL Pos = 150 ng/mL, Neg = 50 ng/mL All assays need controls, even qualitative assays. Of the options given, a positive control of 115 ng/mL and a negative control of 85 ng/mL are the best options. These controls are 15% above and below the assay cutoff and so will adequately assess the 'medical decision point' of the assay and more sensitively reveal any performance issues with the assay or instrument. Using controls that are within 20% of the assay cutoff is the common "rule of thumb" for such assays. Using QC that is within 15%, is even better, and more robust than 20% since it will reveal analytical issues with the assay sooner and with more sensitivity.

28. Quiz A Medical Laboratory Scientist (MLS) receives a package of proficiency testing (PT) samples for her laboratory's serum protein electrophoresis test. Which of the following should be avoided? Please select the single best answer The samples should be run with patients, on the same gel. The samples should be 'randomized' to whichever MLS is working the bench at the time the samples are tested. The PT samples should be interpreted by the same people who read and report patient specimens. The PT samples should be run in duplicate, ideally in another facility. It is imperative that PT specimens not be sent to another address or facility. PT samples should only be run in duplicate if that is how patient samples are normally analyzed. All attempts should be made to run PT samples exactly as patient samples are run. They should not be processed, tested, or interpreted differently or by alternate personnel.

29. A hospital had a plumbing issue that resulted in water flooding the floor of the laboratory. As a result, a large chemistry analyzer was unplugged and moved to another room. After building repairs are made and the instrument is returned to its original location, what should be performed? Please select the single best answer Assays on the instrument should be recalibrated. Ten days of QC should be run to re-establish QC means. The instrument vendor should come and check the instrument before powering it back up. Patient samples should be performed in duplicate for the first eight hours or so. Calibration is required when there is a reagent lot change, QC fails to meet established criteria, or after major instrument maintenance or movement. Moving the instrument would likely qualify as a 'major' change to the instrument and so re-calibration would be appropriate. It is not necessary to reestablish QC means or run patients in duplicate. And while an instrument vendor may offer to assist with a move, they are not necessarily required to be part of re-deploying an instrument that was shut down or moved.

30. A Medical Laboratory Scientist is asked to perform a linearity check on an AST assay that is being validated in the laboratory. Which of the following is true? Please select the single best answer Linearity is the same as a calibration verification (if the assay is known to be linear). If five to seven days of QC results are plotted, linearity can be verified. Running samples with concentrations spanning the assay's range can determine the linearity. Linearity verification is no longer required for any FDA-approved assay. A check of linearity simply verifies that the assay is linear and not curved and is often done as part of an initial validation. Linearity is not the same as calibration verification. One cannot assess linearity using QC. It is not true that verification of linearity is never required for an assay, a manufacturer may require such a check, and most validations will perform a linearity study and report a correlation coefficient (r2) that describes how linear the assay response is.

31. Which of the following is the testing of materials that have known concentrations in order to substantiate an instrument's calibration? Please select the single best answer A linearity check An analytical measurement range (AMR) study Calibration verification Between-day precision Calibration verification refers to the testing of materials of known concentration in the same manner as patient samples in order to substantiate the instrument or test system's calibration across the reportable range of patient test results. A linearity check assesses the correlation or 'straightness' of an assay. An AMR study validates, or re-validates, the linearity and accuracy of the assay over the range of the assay. Between-day precision is an assessment of QC values across different days.

32. Which of the terms below is closest in meaning and practical use for the term AMR? Please select the single best answer QC mean Linearity Calibration Recovery Although not the same thing, the AMR will reflect the linearity of the assay. An AMR analysis requires a linear regression and assessment of the linearity of the assay, with the expectation that the assay is indeed linear over the measuring range of the assay. The other terms are not as directly related to an AMR study. The QC mean is just the average of the latest QC values, it does not address the range of the assay. Calibration obviously impacts the performance of the assay and sets the characteristics of the assay's range, but calibrating or recalibrating 'resets' the relationship between instrument signal and expected concentration, it does not specifically assess the range of analyte concentrations for a given assay. Recovery studies involve the addition of a known amount of analyte to a sample and then determining what percent of the amount added is detected.

33. A TPO antibody test has an AMR of 0-150 IU/ mL. The laboratory has validated a dilution protocol. The reportable range for the assay is 0-300 IU/ mL. If a sample had an anti-TPO concentration of 1000 IU/mL and a 1:10 dilution was made, what should be the outcome? Please select the single best answer A 1:10 dilution would bring the sample to around 100 IU/mL and be within the AMR and reportable range. A 1:20 dilution would be needed to bring the assay into the AMR. Dilutions cannot be used when the AMR and reportable ranges are different. The result should be reported as >100 IU/ mL. Dilutions are useful when a result is outside of the reportable range or AMR. In this scenario, a 1:10 dilution should lower the concentration 10-fold, bringing it within the AMR and reportable ranges, so a calculated result could be released. A 1:20 dilution is more than is needed to bring the analyte within range. Since the dilution will bring the result into the reportable range, a ">" should not be reported.

34. At the request of physicians, a laboratorian validates a dilution protocol for her laboratory's ACTH assay. Which of the following would be true? Please select the single best answer If the analytical measurement range (AMR) or reportable range are extended, the assay becomes a laboratory-modified test (LMT). The dilution would extend the AMR but not the reportable range. It is non-compliant to alter the reportable range or AMR for a commercial assay. A dilution should only be used in samples that exceed the reportable range. Dilution of samples can be used to extend the reporting range for an assay. It is compliant to extend the range if one validates the utility of the dilution protocol. But to do so renders the assay 'laboratory-modified.' When an assay is modified like this, it is no longer 'FDA-approved' since it is being used outside of the original specifications. Dilutions don't extend the AMR, they extend the reportable range.

35. A body fluid assay for CEA has a validated range of 0-100 ng/ mL. When the instrument reads ">100 ng/mL," a technologist performs a 1:10 dilution and reruns the sample, reporting the dilution-adjusted result in the patient's chart (though they will not dilute the sample further if the result is over 1000 ng/mL). What are the analytical measurement range (AMR) and reportable range for this assay? Please select the single best answer AMR and reportable range are both 0-100 ng/ mL. The AMR is 0-100 ng/mL, and the reportable range is 0-100,000 ng/ mL. The AMR is 0-100 ng/mL, and the reportable range is 0-1000 ng/ mL. The AMR and reportable range are both 0-1000 ng/ mL. The AMR is the range that can be assessed without dilution. For this assay, the validated range is 0-100 ng/ mL. However, the reportable range is the concentration that can be reported in the patient's chart with dilutions; this range goes from 0 ng/mL up to 1000 ng/ mL.

36. Which of the following is true concerning the extension of the analytical measurement range (AMR)? Please select the single best answer Dilutions can be used, but the type of diluent must be validated. The AMR can be extended to 10 times the reportable range, but no further. The AMR can be extended only if the FDA-approval for the test is maintained. Dilutions can extend an AMR, but a concentration procedure cannot. Extension of the AMR can be achieved with a dilution or concentration protocol. If a dilution is used the protocol for the dilution must be clearly defined and validated. This validation would obviously include validating the type of diluent to use. There are no theoretical limits to extending the AMR, it depends on the assay or instrument's signal measuring range.

37. When a Medical Laboratory Scientist calibrates or recalibrates an assay, what is actually being performed or being assessed? Please select the single best answer The QC is being evaluated against a peer group mean. Within-day QC precision is being compared to within-day accuracy. Proficiency testing samples are being reconciled to peer group means. The relationship between the reagent/instrument response and the concentration of an analyte is being established. Calibration is a determination or re-determination of the relationship between the reagent system/instrument response and the corresponding concentration of an analyte. This is done by running calibrators of known concentration to re-align or establish the performance of the assay. Without calibration, a quantitative assay is not able to predict unknown concentrations. None of the other choices describe calibration but instead refer to other processes or actions around quality control. Calibration must be done before QC or in response to failed QC, but calibration cannot be done simply by evaluating QC.

38. A hospital laboratory has two instruments that perform troponin testing. One is a high-sensitivity troponin assay that has an analytical measurement range (AMR) of 5-500 pg / mL. The other instrument runs a traditional troponin assay with an AMR of 0.01-100 ng/ mL. Which of the following would be true? Please select the single best answer The two instruments must have comparison studies performed twice per year. The AMR study samples for one instrument can be used for the other instrument. These constitute different assays and as such, they do not need to be compared. The AMR check from one instrument can be used to complete the annual comparison study for the two instruments. These two tests are clearly different tests as evidenced by their very different AMRs (high sensitivity troponin is not the same as a traditional TnI or TnT assay). Since these assays do not share the same AMR or even the same reporting units, it would not make sense to compare them to each other. While a patient's clinical outcomes with regard to each assay's result may be interesting to compare, the instrument values don't lend themselves to direct comparison nor harmonization. Since these are two distinctly different assays, twice a year comparison study is not needed (if such a study were needed it would be twice a year, not annually).

39. A laboratory supervisor receives a call from an irate physician who is complaining that when her patient had his creatinine measured at the hospital the result was 5.2 mg/ dL but when re-ordered at a reference laboratory across town, the result was 2.7 mg/ dL . The physician is wondering how the hospital laboratory resulted such a falsely-high result. Which of the following is a reasonable response? Please select the single best answer Hospital laboratories tend to use faster, but less accurate, methods, and so the outside laboratory is likely the true result. The hospital laboratory is not required to compare or harmonize results with another laboratory at a different address, repeat testing of the patient should be performed to troubleshoot this discrepancy. The two results are reasonably close given that they were from different tubes. Hospitals are required by CLIA to be harmonized to reference laboratories located in the same geographic area, and so a root-cause analysis should be undertaken to investigate the discrepancy. It is true that the hospital laboratory is not required to compare or harmonize results with another laboratory located at a different address. If the hospital and laboratory were part of the same system ( ie , owned and operated by the same agency), then periodic comparisons across the two sites would be reasonable but are not specifically required. CLIA does not require all laboratories in the same geographic to be harmonized. It is unknown which method is more accurate (or which result should be used). The difference between the two results is significant and should be investigated. Comparing the reference ranges from the two laboratories for creatinine maybe a good first step. Repeat testing of the patient should be performed to troubleshoot if there is a true discrepancy.

40. A regional health system uses portable point-of-care (POC) chemistry analyzers at multiple locations throughout the city. They also have a regional laboratory that performs centralized automated chemistry testing. Which of the following is required? Please select the single best answer The laboratory is required to compare the POC instruments with the main laboratory instruments once per year. The laboratory is required to compare the POC instruments with the main laboratory instruments twice a year. The laboratory is not required to compare these instruments since they are located at different locations. The laboratory should do the same method region-wide to ensure harmonization of results . Laboratories are required to compare instruments that are housed within the same CLIA address. Comparisons across a regional system that has multiple sites (such as in this example), are not required (and may not be feasible for extremely large systems). However, verifying that laboratory results are consistent across a health system is prudent since patients may access care at multiple locations and providers expect laboratory results to be commutable, but a strict twice a year comparison is not specifically required. Having a single method for a large or multi-site system is not required and may not be possible.

41. A large volume laboratory has a backup immunoassay instrument that is sometimes used when hepatitis B testing orders increase. What should be the practice for this laboratory? Please select the single best answer Laboratories are not required to do comparisons on backup instruments. Only primary instruments require comparison checks. The laboratory should do annual comparisons for the hepatitis test(s) in question. The backup instrument and the primary instrument need to have results within 25%, and this must be checked twice a year. The two instruments should have results compared twice a year if the backup instrument is used at all during the year . If the backup instrument is used at all during the year, then a comparison check must be conducted twice a year to verify that results are consistent between the two platforms. While a 25% acceptability target may be reasonable, the 25% number is not mandated. The acceptance criteria are up to the laboratory's medical director (and for qualitative tests a percentage may not be possible). The requirement is twice a year, not annually.

42. The analytical measurement range (AMR) for your laboratory's serum glucose assay is 10-800 mg/ dL . A sample is tested and the result is >800 mg/ dL . The clinician would like an actual concentration in this case. Which of the following is true? Please select the single best answer An upfront dilution could be made to bring the sample into the AMR. The result for the diluted sample should be mathematically adjusted to take the dilution into account. If a manual dilution were made, the instrument reading for that sample would be the value you would report. A result greater than 800 mg/ dL can't be released without a new validation (to extend the AMR). A result greater than 800 mg/ dL is not clinically meaningful . In this case, a simple up-front dilution could be done to bring the sample into the measurement range (AMR). However, when tested, the instrument result for that diluted sample cannot be reported without FIRST accounting for the dilution factor and adjusting the result (multiply the result by the dilution). Dilutions do not affect the AMR. A new validation is only needed if the AMR is changed. A result >800 mg/ dL may be clinically meaningful if the provider thinks this will affect patient management.

43. A laboratory supervisor is considering establishing a delta check for serum iron and ferritin. Which of the following would be true? Please select the single best answer This delta check could save patients from being falsely diagnosed with anemia. A delta check is reasonable for ferritin but not for iron. These two laboratory tests are not good candidates for a delta check. Delta checks are never used in hematology since values fluctuate wildly. Ferritin and iron can change significantly over a short period of time. Such changes are often normal or expected, e.g., in response to diet, inflammation, and stress. Therefore, these two analytes are not good candidates for a delta check. Many healthy people would fail a delta check for ferritin and iron, making it a poor-quality control indicator. A delta check for ferritin or iron would not help with anemia diagnosis. However, many laboratory parameters are relatively stable (more often those in the hematology section of the laboratory as opposed to chemistry, e.g., analytes such as RBC count and MCHC). A delta check could potentially be useful in some other setting or patient population but not for ferritin and iron.

44. You are preparing to work for the day and while running QC for prolactin you notice that level 1 failed. Level 2 and level 3 are within acceptable limits. What would be a good first step? Please select the single best answer You can proceed with testing patients but monitor patients for trends and biases. The assay should be recalibrated with a new lot of calibrator. A new lot of reagent should be loaded and QC repeated. Repeat the QC with a new vial of level 1 material . If QC is out, at any level, patients should not be run or reported. The assay may need recalibration or a new lot of reagent, but those steps should only be done if simpler solutions are not successful. It would be most reasonable to first try a new vial of the QC material to make sure that the QC sample itself hasn't deteriorated, been contaminated, or in some way expired.

45. Your laboratory subscribes to a proficiency testing (PT) service that provided a sample to be screened on the immunoassay instrument for urine amphetamine. The test is positive. Like all positives, you send the sample to the toxicology laboratory for confirmation. Which of the following is true? Please select the single best answer You can only send the sample for confirmation if that is the way you would treat a patient sample. The amphetamine result for the PT sample should always be positive; therefore, it does not require confirmation. The PT result should be repeated to verify accuracy. The PT sample must not be sent for confirmation. The PT sample should not be repeated and never sent for confirmation, even if this is how normal patients are treated. The PT survey must be performed blinded and repeat or confirmatory testing is akin to 'cheating on the test' by double-checking one's answer. PT samples can be positive or negative, the result is unknown and variable and only revealed when the survey results are returned back to the laboratory.

46. You have been given a set of proficiency testing (PT) samples to run on the chemistry instrument that you are operating for the day. Which of these procedures should you follow in performing the PT? Please select the single best answer Finish running all the patient samples and then calibrate the instrument prior to testing the PT samples. Have someone else repeat the testing and compare your results. Perform all the tests in duplicate to verify that the results are correct. Incorporate the PT samples into the patient run. PT test samples should be integrated into the routine workload following the same procedure that is used for patient testing. Calibration was not performed prior to patient testing in this scenario. Therefore, it should not be done prior to PT testing. PT testing cannot be repeated by someone else. Repeat testing is only acceptable if patient samples are routinely repeated. If the laboratory policy is to repeat patient tests that are critical results, this can be done for proficiency testing samples as well, but samples that are not in the critical range cannot be repeated, as patient samples with similar results are not repeated.

47. Your technical supervisor has asked you to verify the calibration for all your electrolyte assays. Which of the following is not an acceptable way to test method calibration? Please select the single best answer Use verification materials from the test system's manufacturer. Use calibrators but run them as unknowns, to verify that the correct target values are obtained. Use reconstituted QC to see if you obtain target values. Retest previously reported patient samples . Feedback When checking calibration, all these options are acceptable except for using QC samples. QC samples serve the purpose of being controls and should not be used to validate a calibration. Instead, materials with identical matrices or actual patient samples should be used. If calibrators are run as unknowns, you should not use the same calibrators that were just used for calibrating the instrument. Use a different lot of calibrators.

48. A Medical Laboratory Scientist (MLS) is asked by a supervisor to run three "assay verification samples" provided by the manufacturer. The instrument and this assay were last calibrated two weeks ago. The last time the MLS ran the assay verification samples was six months ago. Which process is the MLS undertaking? Please select the single best answer Calibration verification Reference range verification Recalibration Instrument comparisons Feedback The scenario above describes calibration verification; the testing of the current calibration to recover known results. The assay verification samples are target-value-assigned samples. Assay verification samples would not be used for the verification of reference ranges. Samples from healthy individuals would be used for this purpose. Calibration materials rather than assay verification materials are used for calibration or recalibration of the instrument. Testing assay verification samples ensures that the current calibration is still acceptable. The materials are not being tested on multiple instruments. Therefore, this testing is not being done to compare results from instruments that are used to test for the same analyte.

49. Major maintenance was performed on the chemistry analyzer. Calibration verification is performed by testing the calibrators as unknowns to validate the calibration settings. However, calibration verification fails. What should be done in this case? Please select the single best answer Recalibrate the instrument. Use a different method for calibration verification. Run the daily QC materials to see if those values are acceptable. It is acceptable for the calibration verification to fail once in a while. If calibration verification fails, the instrument should be recalibrated. Testing the calibration materials as unknowns is a straightforward way to determine that the correct calibration target values are recovered. Another method of calibration verification would also fail as it has already been shown that the current calibration settings are not correct. It would also not be acceptable to test the QC materials because these materials will not validate the calibration settings. It is never acceptable for calibration verification to fail.

50. A chemistry laboratory supervisor discovers that serum creatinine testing has been performed for the last 8 hours despite the fact that the low QC failed at the beginning of the shift. What should be done? You answered the question correctly. The correct answer is highlighted below Since only the low QC was out, any creatine value that is normal or high is unconcerning . Low creatinine results are clinically less significant, so no further action is needed. The assay should be checked and once QC is back in, all reported patients since the shift started should be repeated. The laboratory's medical director can then decide if results need to be corrected in patient charts. QC should be repeated. If all QC levels pass, then no further action is needed. All low creatinine values should be run in duplicate and the average value of the duplicates should be reported. Feedback In this scenario we have an assay that is 'out of control' but was still used for patient testing. In such a case, once discovered, reporting should stop and the assay's performance must be corrected. Once troubleshooting is complete and QC is again within acceptable limits, patients that were run during the 'failure period' should be repeated. The laboratory's medical director should then be consulted. Technical personnel can then use the laboratory director's criteria for what constitutes a clinically meaningful change and look to see if any results were significantly different upon retesting. Results that are deemed significantly different should then be corrected in the patient's medical record.

51. A Medical Laboratory Scientist (MLS) is performing his biannual analytical measurement range (AMR) study for beta-hCG. Since his laboratory is used to seeing pregnant women with hCG values >10,000 mIU /mL, he designs the AMR study to range from 5 mIU /mL to 20,000 mIU / mL. The assay's validated range is 5 mIU /mL to 5,000 mIU / mL. Which of the following is true? You answered the question incorrectly. The correct answer is highlighted below Samples above 5,000 mIU /mL will need to be diluted for the AMR study. The AMR should not exceed the validated range of the assay. The AMR should extend to the highest concentration that can be seen in a laboratory's patient population. The AMR should test at least five samples ranging from 0 mIU /mL to 10,000 mIU / mL. Feedback Since the AMR is the range of concentrations that an instrument can measure without any pretreatment of the sample, dilutions should not be performed on samples used in the AMR. The AMR range should match the validated range for the assay and not factor in dilutions. One can change the reportable range for an assay (with dilutions) but in this example, the MLS is trying to change or validate a new AMR. The AMR should not extend below the bottom of the validated range, so a sample of 0 mIU /mL should not be used.

52. Chemistry QC, Calibration & Reportable Range