They are just one factor that comes into play.
By John C. Yundt-Pacheco, MSCS, and Curtis A. Parvin, PhD
Laboratories know that appropriate quality control (QC) rule selection depends on the in-control performance of a test method (the Sigma value). Higher performing tests may allow “easier” rules, while lower performing tests require more powerful rules. In recent years, QC frequency has undergone a shift in thinking and approach related to patient risk-based QC design. That is, what we are asking now is, what is an appropriate QC frequency to assure that the risk of patient harm from an erroneous reported patient result is acceptable?
In that context, is the Sigma value for a test method all that is needed to determine an appropriate frequency for QC testing to adequately mitigate patient risk? The simple answer is no: having two different analytes with the same sigma value does not necessarily mean the same QC frequency is warranted. The Sigma value is an important and necessary piece of information, but it is not sufficient to appropriately mitigate patient risk. There are at least four other factors that should influence decisions about the appropriate QC frequency for a test method:
- The reliability of the test method;
- The expected length of time between reporting a patient result and its being acted on;
- The likelihood that an erroneous reported patient result will lead to an inappropriate medical decision or action; and
- The expected severity of patient harm if an inappropriate decision or action occurs.
Test method reliability
The Sigma metric measures the ability of a test method to limit the number of erroneous patient results produced when the test method is operating in its stable in-control state. However, patient risk not only depends on the likelihood of producing erroneous results when the test method is performing in its stable in-control state, but also on the increased likelihood of producing erroneous results when test methods are out-of-control. Therefore, knowing how often a test method goes out-of-control (the reliability of the test method) is an important consideration when assessing an appropriate frequency for QC testing in order to adequately manage overall patient risk.
An out-of-control condition that produces a single erroneous albumin result that occurs only once every two years would be acceptable. What if the same out-of-control condition produced a single erroneous result every month? That might be of concern. What if the same out-of-control condition happened every two days, so the lab would release about 15 erroneous albumin results each month? Depending on test volume, this may be acceptable—or it may be completely unacceptable. What is clear is that out-of-control conditions occurring every two days have a very different impact than out-of-control conditions occurring every two years.
Time until clinician acts
Even if the QC rules are powerful enough to detect a significant out-of-control condition at the first QC event after the condition occurs, if patient results are being reported as soon as they are produced and verified, then there is a possibility that a number of erroneous patient results will have been reported between the occurrence of the out-of-control condition and its detection. If those erroneous results are not identified and corrected before they are acted on, then the risk to those patients has not been mitigated. How likely it is that the lab will be able to identify and correct erroneous patient results already reported will depend on the expected length of time between the lab’s reporting of the result and its being acted on. No matter how capable the test method, if the time between QC events is too long relative to the time needed to correct results before they are acted on, then patients will be at increased risk of harm any time an out-of-control condition occurs.
Inappropriate medical decisions
The amount of actionable information contained in a patient result depends on the analyte. Some analytes are critical to the medical decisions and actions taken; others are peripheral. Therefore, for some analytes it is much more likely that an erroneous reported patient result will lead to inappropriate decisions or actions. For analytes where the likelihood is low that erroneous reported results will lead to inappropriate actions, the lab can tolerate reporting more erroneous results without significantly increasing patient risk. In this case, less frequent QC testing may be acceptable. Conversely, for analytes where the likelihood is high that erroneous results will lead to inappropriate actions, the lab needs to be much more stringent in not allowing erroneous results to be reported, and more frequent QC testing is advised.
For instance, compare troponin to albumin. Troponin results form a significant part of the clinical diagnostic decision. While other parameters are considered, most of the decision about what to do with a patient presenting with chest pains relies on the troponin result. If there is a credible, positive troponin result, chances are very high that clinical action will be taken because of it. So it is unlikely that a lab will have the same tolerance for erroneous troponin results that it has for erroneous albumin results. Laboratories are more likely to establish tighter QC programs for troponin than they might for albumin due to the potential for harm resulting from an incorrect result.
Severity of harm
Finally, depending on the analyte and how it is used in patient decisions and actions, the severity of harm to the patient from an inappropriate decision or action can differ widely. If the expected severity of harm to the patient from wrong decisions or actions is negligible, then the lab can allow more erroneous reported results and still maintain acceptable patient risk. If the severity of harm from inappropriate decisions or actions is expected to be critical or catastrophic, then the lab must assure that the number of erroneous results reported is kept to a minimum. (Again, the troponin vs. albumin example applies.) This will indicate that more frequent QC testing is needed.
Account for other factors
In summary, the Sigma value of a test method is a metric that will influence how frequently QC testing should be performed, but to depend on it exclusively does not sufficiently take into account patient risk. Other factors, such as those described above, can result in two different test methods with the same Sigma value having very different requirements for QC frequency if the goal is to mitigate patient risk to acceptable levels.