 Dr. Westgard |
|
|
|
There's an ongoing interest in alternative QC procedures, especially for point-of-care applications where traditional statistical QC is difficult to implement due to the cost involved, the training required, and the ongoing support needed to update the QC statistics and maintain appropriate QC charts. Manufacturers have been pushing for acceptance of "electronic QC" as an alternative approach and the government and accreditation organizations appear to be getting on board. Laboratory and point-of-care managers need to be clear about the pluses and minuses of this approach.
Electronic QC involves performing an electronic check on an instrument in place of the analysis of a control sample. An electrical signal is substituted for the signal that would normally be generated by a sensor. This is easy to do for certain types of instruments such as those having electrode sensors, where a voltage from a battery (or other stable electrical source) might substitute for the voltage from the sensor. This electrical signal is then measured to see if the right value is obtained on the instrument readout.
The advantages are the simplicity and low cost. It's quick and easy to do and doesn't use up any test cartridges or reagents. Furthermore, the instrument often is programmed to give a Yes or No decision on the acceptability of the value, therefore it takes little training and no understanding of statistics to utilize electronic QC.
Electronic QC isn't new! We've always had photometer checks, filter balance checks, cuvette checks, blank checks, and a variety of electrical voltages that are monitored by the analyst or by the instrument itself. These checks have been part of the overall quality system and used together with traditional statistical QC.
The total testing process includes preanalytic steps (such as specimen acquisition and handling), analytic steps (such as sample manipulation, sample processing, measurement of a signal from the sensor, and instrument readout), and post-analytic steps (recording of test results). A problem or error in any one of these steps can invalidate the results of the whole testing process.
QC provides a way to check whether this testing process is working properly. However, different QC procedures may check different steps of the process. While it is possible to utilize a different QC check for every individual step in the process, it is much more efficient to have a single QC procedure that can check all of the steps in the process. If there's a problem, then it's useful to dissect the process to pinpoint the specific step that's the source of the error.
One way to assess the usefulness
of alternative QC procedures is to determine which steps in the
total testing process are monitored by the different procedures.
For example, the accompanying figure illustrates the total testing
process for blood gas measurements and identifies the steps that
can be monitored by different types of QC procedures, such as
electronic QC, traditional statistical QC, tonometered blood,
patient sample replicates, and patient sample comparisons.
Electronic QC can monitor for the performance of the instrument readout device. Traditional statistical QC can monitor the readout plus the sensor steps and some of the sample steps (how many depends on the matrix of the control material and whether the handling and manipulation of the control sample mirror the steps for the real patient sample). Tonometered blood can monitor the entire analytic process. Replicate measurements on a patient specimen can expand the monitoring to include many of the preanalytic steps. Patient sample comparisons, where the measurements are made by different instruments, provide the most complete monitoring and even allow some checking on the postanalytic steps involved in recording the test results.
Analogies are never perfect, but
they can still be helpful for seeing a problem from a different
perspective. Consider the problem of quality controlling the drivers
of automobiles. The accompanying figure portrays the Total Driving
Process in a parallel manner - pre-driving, driving, and post-driving
parts of the process. Within the "driving" part of the
process, there are substeps for operator skills, vehicle functions,
and motor checks. You can start the engine and monitor the gauges
on the dashboard to see if the motor is functioning okay. You
need to test drive the vehicle to be sure it steers okay and that
the brakes work. Your driving skill is monitored by traffic cops
who observe the operation of the vehicle. Pre-driving factors
are monitored by the initial licensing of operators and also by
the periodic renewal of licenses. In addition, the procedure for
periodic license renewal can take into account the record of traffic
violations, accidents, etc.
In the Total Driving Process, monitoring the gauges on the dashboard will only tell you that the motor itself is running okay. The motor is certainly a critical part of the driving process, but it's not the complete process. The motor can work okay, but the steering and the brakes can still be bad, the driver can lack the skill to operate a car with a straight shift, the driver's eyesight might be bad, and he may not know which side of the road to drive on. The driving process can be filled with problems even when the motor is working okay.
In a parallel manner, electronic QC can tell you if the readout is working okay, but it doesn't provide any confirmation that the rest of the process is okay - particularly whether or not the operator is performing properly. Electronic QC is useful for monitoring an essential part of the process, but it does not monitor the complete process. Therefore, electronic QC is not a substitute or alternative for traditional statistical QC, but rather is useful as part of a system of QC procedures.
It's essential to know that the readout part of the device is working okay, particularly in point-of-care applications where the device is portable and might be dropped on the floor or bounced off the wall. It's an essential check, just like it's essential to check whether the oil pressure and temperature of the engine are okay and whether there's gas in the car everytime you get ready to drive. BUT, electronic QC is not sufficient to assure that the test result is correct and reliable. Electronic QC needs to be used in conjunction with other QC procedures to monitor the quality of the total testing process.
The issue of what steps are being monitored has somehow gotten lost in the discussions between industry, government, and the accrediting agencies. There now seems to be acceptance of electronic QC as a substitute for traditional QC if the electrical signal is available in a form that can be charted, instead of just a Yes/No indicator. To focus on the form of the signal is missing the whole point! Most cars don't have gauges for oil pressure anymore because a light can alert the operator to that problem more effectively (and most drivers don't really understand the measurement of oil pressure). Putting gauges back in those cars won't really improve the quality system. Likewise, displaying the results from electronic QC on a control chart doesn't expand the steps of the process that are being monitored!
Electronic QC needs to be understood as part of a system of QC procedures that are to be used to monitor the Total Testing Process AND to isolate problems to certain steps of that process. The frequency of use of the different QC procedures is expected to vary. Electronic QC should be performed very frequently, whereas traditional statistical QC and operator proficiency can be monitored less often. Each laboratory and testing service needs to decide the appropriate frequency for applying the different QC procedures. Because statistical QC can be very effective for monitoring operator proficiency, statistical QC is still a very big part of the quality system. However, it is the overall system of QC procedures that is necessary to assure the quality of the test results.
