In some industries, quality is an excellent sales tool. In others, it’s a matter of life or death. Naturally, products and services within the medical field fall into this latter category. Subsequently, health care products, devices, and services tend to be highly regulated, with laboratories and lab testing no exception. Guidelines include those from government organizations, such as CLIA (which are administered by CMS and FDA), and accrediting agencies, such as COLA and the College of American Pathologists (CAP).
“CLIA and COLA require clinical laboratories to have a written quality control program,” says COLA representative Anita Alvare. Elements must include the number and type of controls to run for each test, the frequency with which to run controls for each test, the proper handling of all control materials, how the lab will determine whether the controls are acceptable, the corrective actions that should be taken when they are not, and how quality control will be documented and reviewed.
This year, that program can include equivalent quality controls, or EQC, which replaces the general CLIA/COLA requirement that calls for two levels of external control each day of testing. This means that laboratories can establish a program, as suggested by the manufacturer, which may provide for running far fewer controls. Alvare notes, however, that EQC may only be instituted after performing a successful qualifying study. “And COLA only allows two of the three CLIA options for EQC,” Alvare says.
Fewer controls mean less reagent, technologist time, and documentation—which translate to fewer costs. However, James O. Westgard, PhD, who runs a Web site dedicated to laboratory quality control, has expressed concerns that the EQC method is not only inadequate to ensure quality, but dangerous in that it leads to a “false sense of confidence.”1 Regulating agencies like COLA, however, believe that manufacturers have tried to address these concerns during the development of their products.
“In theory, EQC is adequate for stable test systems and very important in laboratory settings such as point-of-care. But in application, it is up to the user to ensure its validity,” Alvare says.
The user is also responsible for ensuring that whatever quality control method is employed, it is actually completed. “From the lab’s perspective, it is challenging to ensure that all staff consistently follow the quality control program, as written,” Alvare says. Similarly, accrediting agencies can find it difficult to educate laboratory staff on how to incorporate emerging technology into a QC program.
Education at all levels is key to overcoming these challenges. COLA offers several educational products on quality control to help laboratories meet regulatory requirements and improve quality. Individual laboratories also can provide educational opportunities, including in-service meetings. “Staff should understand the relationship between proper instrument maintenance, calibration, personnel competency, and quality control. All must work together to ensure quality,” Alvare says.
Technology helps too, providing a means to perform and track quality control. Alvare notes that a “carefully selected” laboratory information system (LIS) can help to understand and analyze QC results; apply rejection rules and flag controls that are out of range; plot, analyze, and store graphs; review inserted comments; and store 2 or more years of QC documentation. Other technologies exist to help with more specific needs.
External Quality Assessment
Randox, headquartered in Crumin, UK, offers programs that can assist laboratories with their external quality assessment, or EQA, programs. Typically, clinical labs in the United States are required to complete proficiency testing (PT) three times per year. “This means they are comparing themselves to other laboratories only every 4 months or so,” says Alison Canning, BSc (Hons), Randox’s global product manager (QC/EQA). Problems may therefore go unnoticed for a significant period of time.
Randox’s 24/7 is an Internet-based program that permits labs to compare their internal control results against those of other labs running the same controls on the same instruments at any time—24/7. It calculates the (internal) mean and standard deviation, displays Levey-Jennings charts, and applies performance rules based on Westgard Rules. “Plot graphs show how a laboratory compares to a selected peer group. The lab can then use this information to identify whether long-term biases are developing and take action more quickly,” Canning says.
If a lab fails a PT, it is required to determine what went wrong and to fix it, as well as show evidence of its investigation and the corrective action. Two failures and the testing of that analyte is suspended, resulting not only in the additional time and work to investigate and make adjustments but the resulting associated losses, including the impact on revenues and reputation.
By establishing between-laboratory comparability, a lab can better identify the accuracy of assigned values and ranges. The program permits a lab to make comparisons as often as it wishes. A LIMS integrator makes the process even easier by removing the need to perform data entry. “The LIMS integrator pulls data from the LIS system automatically into 24/7,” Canning says. This feature, according to Canning, is one of the most requested.
Antibiotic Disk Susceptibility
Because of the reduction in data entry-related errors associated with LIS use, many software programs for the clinical lab are designed to easily integrate. The BIOMIC V3 from Giles Scientific Inc, Santa Barbara, Calif, is no exception. The automated antibiotic susceptibility and identification system features a direct two-way connection with the LIS. “Virtually all of our users are connected to their LIS, and what that does is eliminate all typing and transcription errors,” says David Gibbs, PhD, president of Giles Scientific.
The system automates the reading and interpretation of human CLSI antibiotic disk susceptibility tests of bacteria and yeast and a variety of commercial identification panels, incorporating digital imaging. The system reads the results, displays them on screen, checks them electronically against rules through the expert system, and, if necessary, offers the users steps to take to resolve an unusual result.
This last aid helps to ensure that inaccurate results are not entered into the LIS where the physician may see them and act on them.
“The digital imaging removes the variability to improve quality and consistency,” Gibbs says, adding that the system can help to overcome challenges such as being inexperienced or color-blind.
“The system is not a blind system like many others, where you stick a reagent or kit into it and walk away. With BIOMIC V3, the image of each test result is on the screen in front of the technologist or microbiologist, who can see the results and make adjustments if necessary,” Giles says.
The system has evolved over more than 20 years, incorporating feedback from users and new standards from the industry. It has more than 2,500 rules based on CLSI documents, as well as published recommendations and standards from the UK, France, and Germany. The ability to access and print these rules, as well as easily determine the next steps when a result is odd, can have a great positive impact on a lab’s quality. The BIOMIC V3 can help a lab to have a standard, well-established QC routine that can be easily verified. “The purpose of QC is to spot problems and fix them before you report the results. If you have a good, standard system with good record keeping, it’s easy to go back and examine problems,” Gibbs says.
Problems can result from some of the most elemental tasks, such as pipetting. In the worst-case scenario, errors introduced in the lab through pipetting can lead to a misdiagnosis or a missed diagnosis, and can result in serious patient consequences. “Even in the absence of patient impact, poor pipetting can cause failed quality control, which requires troubleshooting and corrective action,” says George Rodrigues, PhD, senior scientific manager of ARTEL, Westbrook, Me.
As sample aliquot sizes decrease, the accuracy of a pipette—or rather its inaccuracy—will have a greater impact on the result. Rodrigues has noted a trend toward smaller aliquot volumes that has coincided with increased automation (including within molecular diagnostics) and more expensive reagents. “We are getting requests for the ability to calibrate smaller volumes, down to the single microliter range. People would like to go even smaller, but they lack confidence in their liquid handling ability,” Rodrigues says.
In some instances, this lack of confidence is well deserved. Rodrigues notes that about 60% of pipette-related error is due to improper use. “It turns out that the pipette itself is responsible for only about 40 percent of error,” Rodrigues says.
When the problem is the user, ARTEL can offer training, both in calibration and pipette use. For the pipettes themselves, ARTEL offers pipette calibration systems. ARTEL PCS [pipette calibration system] works with single-channel pipettes down to the smallest volume (0.1 mL, according to Rodrigues); ARTEL MVS [multichannel verification system] calibrates multichannel pipettes using radiometric photometry; and the ARTEL Pipette Tracker manages the process. “[The Tracker] helps customers to schedule and keep track of their quarterly calibration, automatically compare the results to pass-or-fail tolerances, and maintain an electronic record of compliance,” Rodrigues says.
The systems work on-site in the laboratory, ensuring that pipettes are calibrated in the same environment in which they will be used. Some standards, such as ISO 17025, require the recording of environmental conditions, according to Rodrigues. “All calibrated equipment is susceptible to environmental factors—some to a greater extent and some to less—with temperature, humidity, and barometric pressure the three most common factors recognized,” Rodrigues says. He notes that because of their design, pipettes—particularly air-displacement devices—are more susceptible to environmental variation.
Pipettes are regulated by two sets of standards, according to Rodrigues: the American Society for Testing and Materials, which recommends quarterly testing and calibration; and CLSI, which is drafting new guidelines expected also to make a quarterly testing recommendation. “We feel this is a good place to start. It will depend on how the pipettes are used, how frequently they are used, and how rough the treatment is,” Rodrigues says. Some labs will adjust the frequency depending on their experience and quality objectives.
Controls can ensure a system is operating properly, but only if they are meaningful. Ralf Schoenbrunner, PhD, vice president of research and development with AcroMetrix, Benecia, Calif, says control materials should have at least two basic properties. “They need to be commutable—behave like real clinical samples on different systems—and they need to provide traceable results,” Schoenbrunner says.
Independent controls can provide verification that is independent of the measurement system. “Traceable and commutable materials ensure that results for patient samples will be comparable independent of the medical laboratory that produced the result,” Schoenbrunner says.
Manufacturer-provided kit controls, however, tend to be assay and lot specific, are often not expected to mimic human samples, and do not offer independent checks and balances. “These controls are really not designed for comparing performance across systems or labs, and so don’t really provide the trending information that would be valuable,” says Dave Petrich, AcroMetrix’s director of quality systems.
AcroMetrix offers products to help laboratories validate the performance characteristics of their analyte-specific reagent or laboratory-developed test. “Labs are required to perform calibration verification, and they need to use materials that are independent and separate from controls,” Petrich says.
Control materials are available in the areas of molecular/nucleic acid testing and immunoassays, including EBV, EV, CF, CMV, HBV, HCV, HIV, HPV, HSV-1, HSV-2, and MRSA. “All of our materials are now in compliance with ISO 17511, which allows reporting of traceable results,” Schoenbrunner says.
EDCnet, which stands for Electronic Data Collection—Internet-based reporting, provides the means for laboratories to compare their results with other labs running the same lot of controls. The program detects QC result trends early and features built-in validation rules, customized screens, Levey-Jennings charts, vertical scatter comparison graphs, vertical bar graphs for the evaluation of assay performance, and tabular reports for storage of hard-copy reports.
The value of comparison increases when labs run standardized tests, though typically they must use manufacturer-recommended (and often manufacturer-provided) materials. Use of standardized QC material (“traceable and commutable,” Schoenbrunner adds) would improve consistency between laboratories and ensure that proper medical decisions are made, according to Petrich. “When you compare results across labs and get different results, that actually affects medical decisions,” Petrich says. With patient lives at stake, quality is critical and therefore “job one.”
Renee Diiulio is a contributing writer for CLP.
- Westgard JO. EQC, AQC, and FDA QC clearance: From control to compliance; from compliant to complicit. January 2008. http://www.westgard.com/essay121.htm. Accessed February 16, 2008.