Return to CAP Home

  Point of Care FAQs - Quantitative Testing

title
 

Is calibration verification required for ACT instruments?

Calibration verification is not applicable to coagulation testing that reports patient results in units of time (e.g., ACT, PT/INR, aPTT).

Are establishment of the analytic measurement range (AMR) and the clinically reportable range (CRR) required for coagulation test instruments?

Neither calibration nor AMR verification are applicable to coagulation test instruments that report patient results in units of time. The CRR for such instruments can be established using previously analyzed patient specimens, control materials, and/or other standards.

Does regular calibration verification need to be performed on all i-STAT instruments, or can it done on one instrument for each new lot number? For example, can we perform calibration verification on one unit at each campus and then compare i-STAT units?

Single-use devices such as i-STAT instruments or glucose meters are exceptions to the rule that calibration verification must be performed on all instruments. Since large numbers of such devices may be in use at a given time within an institution, it may be impractical to perform calibration verification using a special set of specimens for each device. In these situations, alternative approaches for calibration verification of instruments are acceptable. If calibration verification materials include low, midpoint, high values that are near the stated analytic measurement range (AMR), and calibration verification data are within a laboratory’s acceptance criteria, the AMR has also been verified.

One possible approach is to perform calibration verification on each instrument when it is put into use and after maintenance or servicing. Calibration verification can be performed on representative devices every six months. Acceptable performance of the other devices can be demonstrated by satisfactory QC results. If an institution has more than one device type, lot of reagent strips, or lot of cartridges in service, the process should be repeated for each combination.

Another method of ensuring ongoing acceptable instrument performance is to monitor the extent of agreement between POCT results and parallel blood specimens that have been sent to the main laboratory for analysis. Note that there may be some discrepancies in results in depending upon the analyte concentration and a patient’s condition. This type of comparison is facilitated when the POCT results are downloaded to a central data management computer. Comparisons should be conducted regularly such that over a six month period each POCT device is included in a comparison at several concentrations. This latter approach can also concurrently document correlation between POCT devices and the main laboratory method.

Other approaches for calibration verification of multiple single-use devices may be acceptable as long as the process documents calibration verification for the POCT instruments. Manufacturers’ instructions for calibration verification and AMR verification must be followed.

How often must we re-calibrate the glucose meters that we use for point-of-care testing?

Your glucose meters should be calibrated in accordance with a frequency that is established by the instruments’ manufacturers. Calibration of POC glucose meters is typically set at the factories prior to shipment of the instruments. With such instruments calibration verification, as opposed to calibration, must be performed on a representative sample of devices every six months. Calibration verification must include three points (low, mid, and maximum) that span the analytic measurement range (AMR). The materials used must be matrix appropriate (blood-based, not aqueous). Please contact your manufacturer to obtain recommendations for calibration frequency if you have a calibratable meter. You must, at a minimum, follow manufacturer requirements for calibration frequency.

We are having difficulty correlating the results of blood gas analytes between our several i-STAT instruments because the samples are not stable. Do you have any recommendations on how to correlate these instruments?

Generally, all instruments must be included in instrument-to-instrument and method-to-method comparisons. Such comparisons are ordinarily performed with fresh human samples. Institutions that utilize large numbers of identical instruments, particularly when such instruments, measure unstable analytes, typically perform comparisons on subsets of instruments, using representative instruments to perform batch-to-batch and method-to-method correlations.

Single-use devices such as i-STAT and glucose meters are an exception to the general rule that all instruments must be included in instrument-to-instrument comparisons. For institutions that utilize large numbers of these devices, alternative procedures may be used to document correlation of patient results among the devices and with the central laboratory. One approach is to document the agreement between patient results produced by a laboratory method and representative POCT devices for a lot of reagent strips or cartridges, while concurrently collecting results for QC materials that are run on the POCT devices. In addition, users can correlate POCT single use devices by comparing quality control results from meters that have used identical lots of reagent strips or cartridges and identical QC materials to perform their analyses. Another method of instrument correlation is to monitor the agreement between analyte measurements performed at the POC on whole blood samples, and those from simultaneously collected blood that has been sent to the central laboratory for analysis. Blood that is collected within thirty minutes of a POCT measurement generally provides acceptable agreement if the specimen is immediately processed by the central laboratory. However, there may be some discrepancies in results depending upon the particular analyte concentration and a patient’s underlying medical condition. Other correlation schemes may be acceptable. An appropriate correlation method will document the agreement between POCT devices and the central laboratory’s analyzers.

What are the CAP requirements for validation of new reagent lots for our glucose meters and other POC instruments?

The procedures and materials that are appropriate to validate new reagent lots must meet minimum manufacturer requirements, but are otherwise within the discretion of your laboratory director. Most often, two or three levels of control materials are run for multiple days on new lots of reagents. The means from each level are compared with those of the preceding lot numbers to validate the new reagents.

Please explain the difference between the analytic measurement range (AMR) and the clinically reportable range (CRR).

The analytic measurement range (AMR) is the range of analyte values that a method is capable of directly measuring for a given specimen type. In contrast, the clinically reportable range (CRR) is the range of analyte values that the method can report as quantitative results. Thus, the AMR describes the limits of the method’s ability to measure an analyte, while the CRR can represent an extension of the AMR that is generated by dilution, concentration, or other pretreatment of specimens.

Can control materials be used for calibration verification?

Routine control materials generally are not suitable for calibration verification unless their manufacturers have specifically designated the materials as valid for calibration verification. The use of control materials ordinarily does not allow for verification over the entire analytic measurement range.

In order for control materials to be used for calibration verification they must be matrix specific, and have analyte values that reflect the minimum, middle, and maximum points of the analytic measurement range (AMR). It is unusual for quality control materials to meet these requirements.

According to the manufacturer of our POC glucose meters, the lower limit of these instruments’ AMR is 10 mg/dL. However, 32 mg/dL is the minimum concentration of the materials that are included in our linearity verification kits. Can we extend the AMR beyond that achieved with the linearity kit using materials from sources other than the linearity kit?

The analytic measurement range (AMR) of an instrument-reagent system is the range of values that can be achieved without the use of dilution or concentration protocols. The most common way to establish the AMR is through linearity testing, although the AMR can be verified by a number of different procedures. If the linearity materials you are using to establish the AMR do not verify the range of values that you desire, you may include additional data points to extend the AMR beyond that which you are able to verify with your linearity kit. This data may be obtained with previously tested patient samples (either neat or diluted), compatible materials from other vendors, previously tested proficiency samples, or other reference materials. The AMR must be validated every six months. If you were to use only your linearity kit for AMR verification, any result below 32 mg/dL should be reported as < 32 mg/dL.

For instruments that utilize electronic quality control to satisfy the CAP’s daily QC requirement, does the electronic QC need to be run by the personnel who actually perform patient testing?

The CAP requires the personnel who perform patient testing to run quality control testing. This requirement applies to systems that utilize electronic quality control in combination with external, i.e., “wet,” control materials. Quality control testing establishes that an instrument/reagent system is functioning properly, and that it is producing accurate results. It is the instrument operator’s (testing employee’s) responsibility to ensure that an instrument/reagent system produces correct patient test results before he or she tests actual patient samples.

What is meant by the term “matrix appropriate” materials as used in the CAP POCT Checklist?

As used in the CAP POCT Checklist, matrix appropriate materials refer to solutions or suspensions in which analytes are suspended or dissolved, that are expected to exhibit identical or nearly identical measurement properties as the patient specimens on which clinical laboratory testing is performed. Samples that are spiked to yield identical concentrations of an analyte, but that contain solvents with different matrix characteristics, can display different analyte concentrations upon measurement. These discrepancies in measurement are known as matrix effects. Measurement deviations of this type that occur in prepared samples when they are compared with actual patient samples, are attributable to such matrix effects. Optimally, comparisons between instruments and other quality control measures are performed with matrix identical materials in order to avoid such errors. For example, comparisons between instruments that test whole blood patient samples should be made with actual whole blood samples. At a minimum materials that have been demonstrated to be matrix appropriate should be used in instrument correlation studies and other quality control activities.

Please define “when applicable” in the checklist question “When applicable, are all patient results reported with accompanying reference (normal) or interpretive ranges?”

The CAP requires age and/or sex-specific reference ranges (normal values or interpretive ranges as applicable) to be reported with all patient test results. This is important to encourage proper interpretation of patient data. The “when applicable” is intended to refer to the appropriate use of reference versus interpretive ranges. This latitude is purposeful and necessary to allow medical directors the flexibility to set meaningful, appropriate standards that reflect the unique needs of the physicians and patients served by individual laboratories. All tests should have defined reference or interpretive ranges reported with patient results. The sole exception to this rule is for test results that are reported as part of a treatment protocol, e.g. an insulin sliding scale, in which clinical actions are taken in response to test values.

We have developed a whole blood platelet function test that can be used at the point-of-care. The test produces quantitative values, but uses a numeric “cut-off” to define “positive” or “negative” results. The results are reported qualitatively, as either positive or negative. The FDA considers the reported results to be qualitative. Does initial validation need to include linearity testing, as with a quantitative test?

Given the information that you have provided, i.e. that the FDA considers the test to be qualitative and that you are reporting a result based upon a numeric cut-off, the test can be validated as a qualitative test. Therefore, linearity testing is not required for the assay. Laboratories will need to run controls that verify the accuracy of the cut-off point. As a qualitative test instrument, positive and negative controls should be run each day of patient testing, prior to testing patient samples. Laboratories must also devise proficiency testing methods for the test, which must be performed every six months. If multiple instruments are used within an institution, the instruments’ function must be compared using patient samples every six months.

We use the i-STAT instrument to perform POC arterial blood gas measurements. The instrument design includes an electronic control regimen that is recommended for use by the manufacturer. What daily quality control requirements does the CAP have for this instrument?

Traditionally, laboratories have been required to run two levels of liquid controls on quantitative instruments, at a frequency within which the accuracy and precision of the measuring systems have been expected to remain stable. This frequency has been based upon manufacturers’ recommendations, but has not been less often than once each day of patient testing.

However, the daily use of two levels of liquid controls may not be required for a test system for which the use of instrument or electronic controls demonstrates that the calibration status of the system is maintained within acceptable limits. Thus, the CAP accepts the use of daily electronic or internal controls in lieu of liquid controls after a laboratory has verified the adequacy of electronic controls for monitoring instrument performance. The daily use of two levels of instrument or electronic controls as the sole quality control system is allowed only for unmodified FDA approved “waived” or “moderate complexity” systems.

The laboratory is expected to provide documentation of the validation that was performed on the test system to ensure the adequacy of the quality control regimen. This documentation must include the federal complexity classification of the test system. It must also include data that demonstrates that the calibration status of the test instrument is monitored. The laboratory must also have a quality control policy that specifies the type of quality control materials used and the frequency of their use. Documentation demonstrating that the policy has been implemented and is followed by the testing personnel must be maintained.

What performance verification does the CAP require on new lot numbers of reagents?

The CAP requires laboratories to verify the performance of new lot numbers of reagents, prior to using the new lots for patient testing. Your laboratory may determine the appropriate procedure for verifying new lot performance, e.g., performing QC, calibration verification, patient comparisons, etc.

How frequently must linearity studies be performed on an instrument?

Linearity studies are only required at the time of installation of a new instrument. Linearity studies may be used to establish the reportable range for the assay. By contrast, the analytic measurement range (AMR) must be established before the instrument is used clinically, and must be verified every six months. Calibration verification may be used to verify the AMR.

For testing of activated clotting times in our cardiac catheterization laboratory, we currently test two levels of liquid controls weekly, and two levels of electronic controls twice a day on Monday through Friday, our scheduled days of operation. For weekend emergencies, is it sufficient to use the electronic controls alone although more than twenty-four hours may have passed since controls were last tested?

The quality control requirements for coagulation testing are that at least two levels of control material must be tested every eight hours on each day of patient testing. The control material can be either electronic or liquid. In the situation that you have described, it would not be necessary to run liquid controls on the weekend unless this was your regularly scheduled time for doing so, or unless you were troubleshooting an instrument.

What are the meter-to-meter and method-to-method requirements for POC glucose meters?

Generally, all instruments must be included in instrument-to-instrument and method-to-method comparisons. Such comparisons are ordinarily performed with fresh human samples. Institutions that utilize large numbers of identical instruments, particularly when such instruments, measure unstable analytes, typically perform comparisons on subsets of instruments, using representative instruments to perform batch-to-batch and method-to-method correlations.

Glucose meters are an exception to the general rules that all instruments must be included in instrument-to-instrument and method-to-method comparisons. For institutions that utilize large numbers of glucose meters, alternative procedures may be used to document correlation of patient results among the devices and with the central laboratory. One approach is to document the agreement between patient results produced by a laboratory method and representative POCT devices for a lot of reagent strips, while concurrently collecting results for QC materials that are run on the POCT devices. In addition, users can correlate POCT glucose meters by comparing quality control results from meters that have used identical reagent strips and QC materials to perform their analyses. Another method of instrument correlation is to monitor the agreement between glucose concentration measurements performed at POC on whole blood samples, and those from simultaneously collected blood that has been sent to the central laboratory for analysis. Blood that is collected within thirty minutes of a POCT measurement generally provides acceptable agreement if the specimen is immediately processed by the central laboratory. However, there may be some discrepancies in results depending upon a patient’s blood glucose concentration and his or her underlying medical condition. Other correlation schemes may be acceptable. An appropriate correlation method will document the agreement between POCT glucose meters and the central laboratory’s glucose analyzer.

Is it necessary to perform precision checks on ACT instruments every six months?

Ordinarily, precision checks are only required as a component of the initial instrument and method validation that is performed at the time of an instrument’s installation. Your laboratory may need to perform precision studies for the purpose of troubleshooting, or to investigate possible changes in the precision of an instrument. As part of your ongoing quality control program, monthly QC statistics should be reviewed for accuracy and precision. Accuracy and precision are usually determined by analyzing the percent coefficient of variation and the standard deviation of the data.

The CAP requires that laboratories use repetitive analysis to establish a statistically valid QC range of acceptability for commercial controls that have assigned mean values. The analysis must be performed for each lot of quality control material. How many analyses must be performed in order to meet this CAP requirement?

The number of repetitions that laboratories must perform to establish QC ranges is within the discretion of their laboratory directors. The determination of a statistically valid range requires the inclusion of a sufficient number of values to minimize the impact of any single value on the results. It is common for laboratories to use the statistics that they gather during the first month of use to establish the QC ranges of acceptability for each lot of QC material.

Does the CAP really require the same individual who performs patient testing to perform the quality control testing?

Yes, individuals who actually perform patient testing must perform the quality control testing on the instruments that they use. This requirement ensures that the personnel who operate the instruments have verified that the instruments and reagents are performing correctly. The CAP believes that those individuals who are responsible for operating the instruments are best suited to assess their function, and from this evaluation to decide whether or not their use for patient testing is appropriate.