Although point-of-care testing (POCT) provides rapid test results and the opportunity for faster medical decisions, the unique risk of errors with POCT raises concern over the quality and reliability of test results. In contrast to the central laboratory, where errors predominately occur in the pre- and postanalytic phases, POCT errors occur primarily in the analytic phase of testing (1, 2). This might be related to the non-laboratory staff involved in POCT, but might also be due to test limitations and misuse of POCT in extreme environmental conditions (3, 4).
Clinical personnel with minimal laboratory skills and experience, such as nurses and patient care technicians, perform the majority of POCT. These operators are focused on patient care and do not necessarily understand why they must handle POCT—a task viewed as a laboratory role and not a job for clinical staff.
Yet regulatory standards hold the laboratory director responsible for managing and supervising POCT quality. In a clinic setting, the laboratory director may be a physician, but in a hospital or health system, the chief of pathology and head of the central laboratory often become responsible. POCT is thus at odds with both the clinical staff performing the test as well as the laboratory staff responsible for supervising the test. This conflict creates a situation ripe for errors.
The Evolution of Quality Control
The analysis of quality control (QC), a liquid sample of known analyte concentration, has historically been used to prove the stability of a test system and ensure quality results. The concept of QC arose from the factory models of the early 1940s in which products on a factory line were periodically inspected to ensure that they met standards of production. If not, the line was shut down until the problem could be corrected. Just as in the factory, the periodic performance of QC in a laboratory—normally two levels each day of patient testing—ensures that the test system is performing as expected and the risk of errors is minimized to clinically acceptable levels.
The standard for two levels of QC each day became the default QC frequency in an era when batch analyzers sampled QC from the same bottle of reagent as patient samples. Patient results were held until QC results from before and after a batch of patients were acceptable, indicating that the test system was stable over the group of specimens. Now, modern instruments produce results continuously and automatically verify test results before the next QC is analyzed.
In the laboratory, QC does a good job at detecting systematic errors—those that occur from one point in time forward, and affect the QC samples in the same manner as patient samples. Reagent degradation, incorrect calibration setpoints, and pipetting errors can all be detected by QC because the same reagent and instrument settings are used for both patient and QC samples.
However, QC does a poor job at detecting random errors, which are unique to single samples, such as clots, hemolysis, or drug interferences, and affect a patient sample differently than QC. As a result, the requirement for two levels of QC each day of testing does not ensure that a test system will be free from errors and have zero risk.
Many POCT devices are single unit-use cartridges and test strips. With unit-use formats, analysis of liquid QC can verify the performance of an individual test, but the analysis of QC consumes the test cartridge and cannot guarantee the quality of tests from other cartridges. Thus, unit-use tests often contain internal control processes built into each test to ensure result quality on each cartridge.
For example, pregnancy tests contain an internal positive and negative control built into each test to ensure the viability of each cartridge. However, some POCT devices, like bilirubinometers, are non-invasive and have no means of analyzing a liquid QC sample. Others, like newer molecular arrays and diagnostic chip technologies, perform hundreds of test reactions on a single cartridge.
How does the laboratory control such tests? Does the operator have to control each reaction occurring on the chip each day of testing? This could be cost prohibitive and duplicative of internal control processes built into the test system. With so many different devices and control processes available, laboratories need a systematic approach to ensure quality and strike the right balance of liquid QC in concert with internal control processes. That approach is risk management.
Understanding the New Guidelines
The Clinical and Laboratory Standards Institute (CLSI) guideline EP-23 introduces risk management principles to the clinical laboratory (6). EP23 describes good laboratory practice for developing a quality control plan based on the manufacturer's risk information, applicable regulatory and accreditation requirements, and the individual healthcare and laboratory setting. This guideline helps laboratories identify weaknesses in the testing process that could lead to error and explains how to develop a plan to detect and prevent those errors from happening.
The Centers for Medicare and Medicaid Services (CMS) has incorporated key elements of risk management from CLSI EP23 into the new CLIA interpretive guidelines that offer a QC option called an Individualized Quality Control Plan (IQCP) (7). The CMS changes were launched in January 2014 and have a 2-year educational period. Beginning in 2017, laboratory tests, including POCT, will have two options for defining the frequency of QC for moderate- and high-complexity tests: either two concentrations of liquid QC each day, or developing an IQCP.
Inspectors will be checking that the laboratory's IQCP is comprised of three parts: a risk assessment (RA), a Quality Control Plan (QCP), and a Quality Assessment (QA). In the RA, the laboratory identifies and evaluates potential failures and errors in a testing process. The QCP is the laboratory's standard operating procedure that describes the practices, resources, and procedures to control the quality of a particular test process. The QA plan is the laboratory's policy for the ongoing monitoring of their IQCP (7).
IQCPs will be valuable to laboratories that use unit-use devices and instrumentation with built-in control processes. While CMS will be enforcing the new CLIA interpretive guidelines and IQCPs on only moderate- and high-complexity laboratory tests, any laboratory will find risk management principles useful in defining their weaknesses and reducing errors in their testing processes.
The first step in developing an IQCP is collecting information about the test and conducting a risk assessment. How will the test result be utilized in patient care? This defines the clinically acceptable tolerance for analytical performance, bias, and precision. Take glucose, for example. Use of a glucose result for diagnosis of diabetes requires tighter performance than use of glucose tests for managing insulin dosage. These differences in clinical use limit glucose meters to management rather than diagnosis or screening purposes. Laboratories should also have an understanding about who will conduct the test and where the test will be analyzed—for example, in a laboratory setting, at the bedside, or in a mobile ambulance, each with different environmental conditions and operators.
Sites with clinical staff or more frequent staff turnover may have higher risk of errors and require additional training or supervision compared to sites with experienced medical technologists. Laboratories should also collect information from their accreditation agencies about their standards for QC requirements.
Finally, laboratories will need information about the test systems themselves and how internal control processes work. Good sources of information include the test package insert and device owner's manual. Understanding the test limitations as well as manufacturer recommendations for use can help laboratories minimize use under conditions that may increase the risk of error.
Performing Risk Assessment
Risk assessment identifies potential hazards—failures or errors—that can occur at any step of the testing process. The risk assessment process takes into account preanalytical, analytical, and postanalytical processes. To assess risk, a laboratory maps its testing process by stepping through each part of the procedure to look for weaknesses: from order to sample collection, transport, processing, analysis, result reporting, and communication of results. The laboratory also should consider the sample, reagents, operator, test system, and environment as potential sources of error.
CLSI EP23 provides a fishbone diagram of major sources of error to consider when conducting a risk assessment (Figure 1). For each of the identified hazards, the laboratory should develop an action plan that details how that risk will be handled. In some instances, the manufacturer's internal control process may address the risk. Take, for example, barcoded reagents that prevent use after the package expiration date. Barcoding is a control process that minimizes the possibility of using expired reagents. While barcoding doesn't absolutely prevent the error from occurring (one can never achieve zero risk), the likelihood of this error is reduced to a clinically acceptable level.
For other hazards, the risk of error may be unacceptable and require the laboratory to take additional actions. For example, an analyzer might have clot detection to reduce the probability of releasing a falsely decreased test result, but a laboratory could emphasize training, collection technique, specimen mixing, and monitor phlebotomists for frequency of clotted specimens as additional control measures. The selection of how each laboratory addresses its identified hazards creates the individuality of the QC plan. The summary of all identified hazards and the laboratory-specific actions to address each risk become the laboratory's IQCP.
Once developed, the laboratory should monitor the effectiveness of its IQCP. Benchmarks of quality can trend the frequency of failed QC, error codes from internal control processes, repeat testing, physician complaints, and any unexpected events. When the laboratory identifies a trend, it should determine the cause of the problem and take corrective action to prevent recurrence. Once corrected, the laboratory should reassess the risk to determine if a particular hazard was missed during the initial risk assessment, if a specific error occurs more frequently, whether the laboratory action is not as effective as predicted, or if missed errors have greater patient harm than thought. The outcome of the risk assessment will determine whether the laboratory needs to take additional steps to mitigate this hazard and whether the IQCP should be modified. In this manner, the laboratory has created a continuous improvement cycle of identifying, assessing, addressing, and monitoring risk.
Focusing on the Right QC
The primary objective of IQCPs is not to reduce the frequency of analyzing liquid QC, but rather to ensure the right QC to address a laboratory's specific risks and ensure quality test results. In the context of POCT, laboratories should incorporate both internal and external control processes. Each device is unique, operates differently, and offers specific control processes engineered into the test. And since no single control process can cover all potential risks, a laboratory's QC plan must incorporate a mix of internal controls and traditional liquid QC.
Each test will require a specific IQCP, because devices are different and present unique risks. However, a single risk assessment and IQCP could cover multiple tests conducted on the same instrument, provided the IQCP factors in the differences unique to each analyte. For instance, a single IQCP for a chemistry analyzer could cover all tests conducted on that analyzer, since instrument operation, risk of error, and functionality of control processes is shared amongst all analytes on the same analyzer. Specific tests, like potassium, present unique consideration for hemolysis, and that risk could be added to the general IQCP covering the analyzer. This process will simplify a laboratory's risk assessments and efficiency in developing its IQCPs.
IQCPs will benefit laboratories in a number of ways. Laboratories using unit-use devices will define the optimum frequency of liquid QC in conjunction with the manufacturer's control processes. For unit-use blood gas and coagulation devices, laboratories can be more efficient by analyzing QC for lots of reagents using a subset of devices rather than every device available, since the chemistry of the test is in the unit-use cartridge—not in the device, which acts only as a volt-meter or timer. For molecular arrays and labs-on-a-chip, analyzing liquid QC across each reaction may be less effective than controlling the processes of greatest risk, such as quality and amount of sample, viability of replicating enzyme, and temperature cycling.
In conclusion, no device is foolproof and errors can occur anywhere in the testing process. Recognizing the conditions that could lead to errors and outlining the necessary actions to avoid them is the basis of developing an IQCP. Risk management and the principles of an IQCP should not be an entirely new concept to the clinical laboratory as most laboratories already recognize the potential for errors and take steps to prevent and detect errors that could ultimately harm a patient. By adopting an IQCP for POCT, laboratories can make certain that patients receive the highest quality of care, with faster turnaround times that do not compromise the accuracy of results.
- Bonini P, Plebani M, Ceriotti F, et al. Errors in laboratory medicine. Clin Chem 2002;48:691–8.
- O'Kane MJ, McManus P, McGowan M, et al. Quality error rates in point-of-care testing. Clin Chem 2011;57:1267–71.
- Lippi G, Guidi GC, Mattiuzzi C, et al. Preanalytical variability: The dark side of the moon in laboratory testing. Clin Chem Lab Med 2006;44:358–65.
- Plebani M. Does POCT reduce the risk of error in laboratory testing? Clinica Chimica Acta 2009;204:59–64.
- International Organization for Standardization (ISO). Medical devices – Application of risk management to medical devices. ISO 14971. Geneva, Switzerland: ISO 2007.
- Clinical and Laboratory Standards Institute (CLSI). Laboratory quality control based on risk management; approved guideline. CLSI document EP23-A. Wayne, Pennsylvania: CLSI 2011.
- Centers for Medicaid and Medicare Services (CMS). Individual Quality Control Plan (IQCP) for Clinical Laboratory Improvement Amendments (CLIA) laboratory nonwaived testing. http://www.cms.gov/Regulations-and-Guidance/Legislation/CLIA/Downloads/IQCP-announcement-letter-for-CLIA-CoC-and-PPM-labs.pdf (Accessed June 2014).
Sarah Njoroge, PhD, is a clinical chemistry fellow at Vanderbilt University School of Medicine in Nashville, Tennessee. Email: [email protected]
James H. Nichols, PhD, is a professor of pathology, microbiology, and immunology, medical director of clinical chemistry, and associate medical director for clinical operations at Vanderbilt University School of Medicine in Nashville, Tennessee. He was the chairholder of the CLSI EP23 Document Development Committee. Email: [email protected]