Instituting a multiple-site urinalysis QC program.
At Island Medical Laboratories on Vancouver Island in the southwest corner of British Columbia, our quality assurance needs kept expanding with us. As we grew over a period of years from six to 26 labs, the facility ultimately included a central reference lab, seven analytical branches, and 18 collecting stations. o Too much diversity. When some of the laboratories joined their own ways of doing things. The growing distances fostered overly diverse methodologies. Only with a revamped quality assurance program, we decided, could we resolve the difficulties posed by logistics and the discrepancies that had consequently sprung up. Our solution will be discussed in this article.
Inconsistent methods. Urinalysis was done in the reference lab and analytical branches. At the collecting stations, only limited analysis for glucose and ketones was performed. The number of specimens tested at our sites each day differed widely-from 10 in the smallest branch to as many as 200 in the reference laboratory.
Two labs used semiautomated urinalysis strip readers. The majority of branches read test strips manually. Separate strips for testing glucose and ketones were read manually in the labs that tested specimens from patients undergoing glucose tolerance tests. Light microscopy was used for sediment examination in all our analytical labs except the main lab, which had a phase microscope. Centrifugation speed and spinning time varied from one location to another. One location left 1.0 to 2.0 ml of sediment after removing supematant from spun urine; another left 0.2 to 0.5 ml. This discrepancy could, of course, drastically alter the values of the constituents of microscopic sediments.
Outdated manual. The urinalysis manual used in most of our locations was incomplete and in need of updating. The result was test methodology that differed from one lab to another. Because specimen handling and collection were poorly defined, the accuracy of results was variable. Poor collection procedures. Although the collection of specimens for culture was carefully controlled, collection for routine urine chemistry analysis and microscopic examination varied dramatically. it was not unusual for our technologists to receive specimens in pickle jars, juice or shampoo bottles, tomato soup cans, and other unusual containers. These specimens had often been collected hours before reaching our labs. Even after arrival they frequently sat unrefrigerated too long before being tested. Inadequate follow-up. Labs weren't given procedures for following up out-of-control results. We had no rules for determining when abnormal specimens would be referred for review by the pathologist. The performance of confirmatory tests, such as those for bilirubin, protein, and glucose, was haphazard. The manual failed to define protocols for maintaining microscopes, refractometers, and urinalysis strip readers; therefore, this work was not performed uniformly. Moreover, we had no formula for evaluating individual technologists' ability to recognize microscopic elements.
Limited QC efforts. For quite a while before we thought of updating our urinalysis quality control program, only two QC measures were routinely being used at our labs. First, a different factitious urine specimen was circulated to several sites each week. Second, technologists at our two largest labs tested their own ability to recognize microscopic elements by examining the 35-mm transparencies provided in the quarterly surveys by the College of American Pathologists.
Our artificial test samples, prepared in advance, never contained bilirubin or urobilinogen, but often included whole blood and contaminated urine. Because these elements are unstable, test results varied from one location to another, depending on when the test specimen had arrived. The collecting stations, whose technicians performed only limited urine glucose and ketone testing, were not included in this program.
We knew a great deal of work lay ahead
* Research. A formal literature search by author Page uncovered few useful articles about urinalysis quality control. Standard textbooks of laboratory medicine were more helpful, as were materials on urinalysis procedures from manufacturers of diagnostic reagents.
Authors Page and Leadbeater conducted a telephone survey to learn about the practices of local hospitals and other laboratory groups. Page's experience was especially invaluable in this regard, as our group described in a previous article published in MLO ("Fill That R & D Position Now," April 1988). She had the background and time to pull together the information we needed from the abundant resources that were now at hand.
* Soliciting advice. The employees of each branch were asked to share ideas for the new program. Leadbeater made site visits and communicated extensively with branches via electronic mail. Using the staff's valuable hints and information, we were able to standardize the procedure for microscopic examination of urine, simplify the report form to one that physicians could read more easily, and update our protocol for telephoning test results.
We presented our tentative QC program changes to our pathologists and chief technologists at the next research and development meeting. At about the same time, we circulated copies of the program to lab supervisors for comment. When all suggestions had arrived, we incorporated the best of them into the final plan. Only then did we begin to schedule inservices.
* Education first. Technicians at many of the smaller branches, where only glucose and ketones were tested, had never participated in QC measures for urinalysis. In visits to those branches and to any others about to participate in such a program for the first time, Leadbeater reviewed the details with the staff in person. Finally, with education complete at all branches, we set a date for implementing the program, of which Figure I provides an overview. 9 Program components. Our logical course was to apply quality control to the most commonly measured variables in urinalysis. Ten of the 12 parameters of our program (Table 1) were chosen to correspond with the number of standard reagents contained on the strips we were using. (Urine color was assessed visually; proficiency at microscopic determination was maintained by reviewing the 35-mm CAP slides.)
Our manual, readily available at all locations, includes the following sections, in this order: uniform instructions for specimen collection and handling, standardized testing procedures, criteria for instrument calibration and maintenance, standardized methods for test result reporting, outline of QC procedures, and criteria for deciding when to send an abnormal specimen to the central lab for pathologist's review. To facilitate updates, we kept to a simple format-printouts in red loose-leaf folders.
* QC specimens. Once a day, technologists at all seven analytical labs make a solution for testing by reconstituting urinalysis control strips in distilled water. Once a week, reference technologist Leadbeater prepares factitious specimens from a "recipe" that simulates eight of the 12 urine parameters. Sodium chloride is used to test specific gravity, for example, and dextrose to test glucose (Table 11). The specimen is split and sent to all locations, where it is tested within hours. Once a month, Leadbeater sends all sites a commercial abnormal urine control that includes leukocytes, urobilinogen, and bilirubin.
To promote consistency, all staff members are asked to handle QC specimens as though they had been collected from unidentified patients. The lot number and expiration dates of the urinalysis strips are noted on each report. With this step we make sure only fresh reagents are used and are able to trace a bad strip to its lot number.
*9 Results computed. Our computer department designed a program that delineates the range of normal for every parameter we are testing. After the reference technologist receives results at the main laboratory, she or her designee enters them into the computer. The program determines tile mean for each test and flags any results that stray too far from our standards.
We now use a consistent procedure for handling out-of-control values. First, the reference technologist immediately discusses the situation, by phone or electronic mail, with the supervisor of the lab, to ascertain whether all procedures were followed as outlined in the manual. The items to be rechecked might include reagent expiration dates, correct timing of tests, and instrument maintenance and calibration. If necessary, the reference technologist visits the lab to discuss the problem. Any necessary remedial education is provided at this time with the pathologist's assistance.
Leadbeater (or a staff technologist) prints out QC results to be circulated, posted, and stored in a permanent file. In a quarterly summary of weekly results from all locations, she assesses trends and identifies any problem areas. The summary is retained for the accreditation inspections held every five years by a team from the Diagnostic Accreditation Program of the College of Physicians and Surgeons of British Columbia and the British Columbia Medical Association.
The CAP slides, previously reviewed at only two sites, are now circulated to all of our labs that perform urine microscopics. Leadbeater tabulates and compares the results from each location. This use of 35-mm slides is helpful, since we currently assess proficiency at urine microscopics in no other way. o Benefits and drawbacks. Now that our revamped program has been in place for over a year, we can review its advantages and plan improvements. We are pleased to be using standardized procedures for collection, handling, and analysis at all branches. Our newly incorporated daily, weekly, and monthly QC checks insure that any deviation will be detected and corrected rapidly. Our staff is proud to produce top-notch results far more consistently than before. Finally, it's comforting to have a comprehensive quality control program that virtually assures high marks during accreditation inspections.
The distances between our facilities still cause problems, including inevitable delays. We found it too expensive to use a daily QC specimen other than the urinalysis control strip. Furthermore, we have not always been happy with the quality of commercially available abnormal QC specimens. Some, for example, are difficult to interpret because of their color. We continue to search for products of better quality.
* Continual upgrades. In today's competitive climate there is always the need to improve. The semiautomated instruments we plan to introduce in all branches should provide greater consistency in interpreting urinalysis strips and speed specimen processing. Budget permitting, we hope to place phase microscopes in all locations where urinalysis is performed.
In addition, we are considering new protocols that would allow us to reduce or eliminate unnecessary routine microscopy. It is inefficient in cost and time to perform routine microscopy on every urine specimen that comes in. Depending on what our physicians suggest, we may reduce microscopy or increase our reliance on semiautomated equipment. o Tightening bonds. Our group of labs must remain physically apart by virtue of the great distances in our territory. Our program, however, keeps us in close touch. With its assistance, we are better able to work as one. 5
|Printer friendly Cite/link Email Feedback|
|Author:||Leadbeater, Ann; Page, Norma; Brigden, Malcolm|
|Publication:||Medical Laboratory Observer|
|Date:||Jan 1, 1991|
|Previous Article:||Lab salaries make altruism still a vital component.|
|Next Article:||You can learn a lot from your new employees.|