A primer for proficiency testing.
Imagine two hospital laboratories running the same serum specimen yet reporting out digoxin levels that lead to a decrease in dosage at one hospital and an increase at the other. That can happen, but it isn't as likely as it once was. When such extreme variation in test findings was first documented by Belk and sunderman in the late 1940s, it profoundly changed laboratory medicine.
These disturbing differences in quality prompted various professional and Federal agencies to design programs challenging laboratories to analyze unknown specimens. Thus proficiency surveys were born. Laboratories could now evaluate their testing precision and also compare their performance with that of labs across the country. In addition, the data contributed to selection of the best methods, reagents, and instruments for a particular workload.
I and seven other consulting medical technologists in our organization have helped numerous laboratories cope with quality control, proficiency testing, preventive maintenance, and continuing education. Our organization is a group of 25 pathologists that provides pathology and laboratory coverage at well over 100 hospitals--large and small--in the Midwest. Some of the pathologists are based at one of the group's four reference laboratories in Sioux Falls and Mitchell, S.D.; Des Moines, Iowa; and Mankato, Minn. Others have offices at client hospitals.
Most of our time is spent on the road, accompanying each pathologist on visits to 10 to 15 smaller hospitals per month. The attitude toward proficiency surveys, I learned, varied greatly from one institution to the next.
Some laboratories conscientiously ran the tests and then performed an in-depth evaluation of the results. In contrast, I once discovered a survey stuffed in a drawer where it had remained unopened for six months. Between these two extremes were a number of laboratorians who struggled--they didn't know how to handle the survey or had trouble finding time for it.
Although the proficiency process is a necessary evil to some, many others find it a rewarding educational experience. One of our client labs reviews the results at its CE meetings and discusses any necessary corrective action. A supervisor at another likens proficiency surveys to a road map, guiding the selection of instruments and methods and facilitating the evaluation of current procedures.
The survey's comparative data can act as a laboratory consumer report. For example, if one test consistently outperforms another across the country, a laboratory may well decide to switch methods. The data can also spotlight basic instrument malfunctions and provide powerful ammunition for convincing administrators to retire an antiquated analyzer.
Proficiency surveys are available from several organizations, including the College of American Pathologists, the American Association of Bioanalysts, and many state agencies. Such specialty groups as the American Association of Blood Banks offer proficiency tests for their particular disciplines. Several manufacturers have surveys for their instruments or procedures; two common programs cover urinalysis and blood gas testing.
The 1985 JCAH manual states: "Proficiency testing services utilized for each discipline of the clinical laboratory must equal or exceed the requirements of pertinent Federal, state, and local licensing and regulatory agencies with respect to variety and frequency of testing and to criteria for satisfactory performance."
Some supervisors shop around for a combination of surveys that will save their laboratories money and still cover all areas of testing. It is especially important to include all tests. Often a laboratory will perform very well on basic procedures but overlook its more esoteric or infrequent tests. Whenever a lab adds a new procedure, it should routinely upgrade its proficiency profile.
In general, lyophilized serum makes the best specimen for surveys. Its advantages include stability, batch preparation, and easy shipment and handling. It does require reconstitution, which introduces an extra chance for error.
Liquid serum has been widely used in Europe, and stabilized cell samples for hematology have performed well in the United States. The main drawback is the possibility of deterioration in transit, especially in hot weather.
Inspect the contents of a survey package immediately upon arrival and note the type and condition of the specimens. Look for cracked caps and check volumes, because leaks can interfere with accurate results. If a hematology specimen appears to have hemolyzed, make a note. You will be glad of the documentation if results are off by a wide margin.
It is also wise to stamp the date of receipt and note the postmark. A long delivery delay signals possible problems, as does a very short due date (surveying agencies generally allow at least a week for turnaround). If the specimens arrive on Monday morning, find out where the proficiency package spent the weekend. The hospital's receiving department could be contributing to the delay.
Weather conditions can also upset delicate specimens. One supervisor, puzzled by nonsense results, discovered that the specimens had been in a heated storage room over a four-day weekend. A long trip in the back of a delivery truck on a summer day will cause similar problems.
Assuming the specimens check out, pay special attention to the survey instructions. A careful reading can prevent errors in reconstitution and in filling out report forms. If questions arise, telephone the agency for clarification. You also can report instances when a specimen is compromised or a test just isn't working properly. Your call, in fact, may be one of many. A last resort is to compare notes with other laboratories. Relying too heavily on a second opinion, however, negates the educational value of proficiency testing.
The purpose of these surveys is to provide a means of evaluating both test and staff performance. For this reason, survey specimens should be analyzed exactly like actual patient specimens. Technologists who routinely perform certain tests should process those survey specimens as well. Figure I, a worksheet devised to coordinate each technologist's proficiency performance and continuing education, gives supervisors an overview of survey activities in the lab. It can also help target inservices and guide assignment of surveys to staff members.
Rotate surveys among all technologists, and try to pinpoint problem areas. Resist the temptation to give surveys only to technologists who always get the right answers and to let others avoid proficiency testing altogether. As a compromise, you might let the star performer act as a control for less experienced laboratorians, giving everyone a chance at participation without jeopardizing the laboratory's score.
Some laboratory sections may find it helpful to develop specific instructions for their technologists. For example, a microbiology supervisor might offer the following tips:
* Process the survey specimen exactly as you would an actual patient test.
* Follow instructions carefully.
* Dilute specimens specifically according to the instructions.
* In addition to normal procedures, do a Gram stain on diluted broth for comparison with subsequent notes and to insure that suspected organisms are actually present.
* Use both coagulase and catalase to double-check for staph.
* Make sure you save all Gram stains.
* Note all procedures done on each survey specimen.
The type of ongoing evaluation shown in Figure I enables you to check past performance and identify and correct persistent problems. If a medical technologist's performance is unacceptable in two blood bank surveys in a row, something is definitely wrong. If necessary, go through the entire procedure with the technologist to find where the problem lies.
After the specimens are analyzed, the results should be carefully checked and recorded. Since clerical errors are a major source of erroneous results, it is crucial to recheck calculations, verify results, and proofread the report.
Next, collect all the data needed to interpret the results, and photocopy everything that will be submitted to the proficiency agency. Fill out and file worksheets, results of the standards and controls, and the names of those performing the work. Such documentation is an invaluable backup should the agency make a clerical error of its own.
Above all, get the results in on time. Surveying agencies place a great deal of emphasis on prompt turnaround. And don't discard the specimens after sending the results out. Chemistry and special serology specimens will keep indefinitely when frozen and give the laboratory a built-in quality control program. If discrepancies show up in the survey report, you can always shaw the specimen and rerun the test. The specimens can also serve as unknowns for students and new employees and for test runs of new methods or instruments. Supervisors can use them to spot-check staff performance.
All laboratories should strive to develop a written protocol for proficiency testing. The format in Figure II provides directions for the entire documentation and evaluation process: dates, survey performers, and corrective action taken for results falling outside 2 SD. When problems are investigated, record all areas evaluated and the resulting remedies.
Perhaps the most important aspect of follow-up is getting the technologists involved. In too many labs, those performing the proficiency testing never learn how they fared. Both technologists and the lab would benefit from their input in the survey evaluation.
I favor distributing and discussing the proficiency testing data sheet at laboratory meetings. This way, the entire staff shares the findings and participates in the investigation of errant results. Other labs prefer to post the results and note what type of corrective action followed.
The basic steps in the evaluation process are outlined in Figure II. Once the results falling outside 2 SD are identified--many proficiency reports flag the outliers--checks can then be made for clerical or transcription errors. Review the instructions and make sure that all directions were followed accordingly.
It's also a good idea to check the reconstitution of the survey specimens by examining both the pipets and diluent used. Perhaps the pipet delivered 48 ml instead of the required 50 ml, or maybe it needs a new tip.
Attempts should also be made to evaluate quality control for an appropriate time frame. That is, don't just look at the day the test was run. Go back a month or so and check for any change in ranges or other possible trends. Instrument problems and appropriate checks should also be noted.
One chemistry section was totally mystified because its quality control was right on target, and yet enzyme results were way out of range. After much sleuthing, we finally found that the instrument's temperature would hold constant for a while and then slowly start to drift. This is now part of the routine instrument check. Depending on the particular problem, the laboratory may choose to repeat survey tests, along with any other specimens in that original run. All results should be documented.
In general, I think it's a good idea to delegate the follow-up assignment to the technologist who originally performed the test. Supervisors should be available for guidance, of course, but otherwise should let staff members proceed on their own.
In the interest of teamwork, the various investigators should report their findings to the rest of the staff, preferably at the regular laboratory meeting. These sessions offer an excellent opportunity for laboratorywide continuing education, as the group decides on the most appropriate corrective course.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||differences in quality in hospital laboratories|
|Publication:||Medical Laboratory Observer|
|Date:||Jun 1, 1986|
|Previous Article:||Creating career choices: advice from an MT turned entrepreneur.|
|Next Article:||A medical technologist in central Africa; Peace Corps service in a hospital laboratory in Zaire included teaching medical technology in French and...|