| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Special Reports |
Departments of
1
Laboratory Medicine and
2
Medical Education, University of Washington School of Medicine, Seattle, WA 98195.
a Address correspondence to this author at: Department of Laboratory Medicine, University of Washington School of Medicine, Box 357110, Seattle, WA 98195-7110. Fax 206-548-6189; e-mail mastion{at}u.washington.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: This study describes the development and implementation of a computer program, Urinalysis-ReviewTM, which periodically tests competency in microscopic urinalysis and then summarizes individual and group test results. In this study, eight Urinalysis-Review exams were administered over 2 years to medical technologists (mean, 58 technologists per exam; range, 44–77) at our academic medical center. The eight exams contained 80 test questions, consisting of 72 structure identification questions and 8 quantification questions. The 72 structure questions required the identification of 134 urine sediment structures consisting of 63 examples of cells, 25 of casts, 18 of normal crystals, 8 of abnormal crystals, and 20 of organisms or artifacts.
Results: Overall, the medical technologists correctly identified 84% of cells, 72% of casts, 79% of normal crystals, 65% of abnormal crystals, and 81% of organisms and artifacts, and correctly answered 89% of the quantification questions. The results are probably a slight underestimate of competency because the images were analyzed without the knowledge of urine chemistry results.
Conclusions: The study shows the feasibility of using a computer program for competency assessment in the clinical laboratory. In addition, the study establishes baseline measurements of competency that other laboratories can use for comparison, and which we will use in future studies that measure the effect of continuing education efforts in microscopic urinalysis.© 1999 American Association for Clinical Chemistry
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the US, microscopic urinalysis is considered a moderate complexity task. Therefore, personnel who perform microscopic urinalysis must be assessed for competency. A competency assessment program for a moderate complexity test should have the following characteristics: (a) it should be performed and documented at least two times annually; (b) it should identify problem areas for individuals and provide remedial retraining in those areas; (c) it should be applied uniformly across the laboratory; and (d) it should be easy to administer, taking a minimal amount of technologist and supervisory time, at the same time providing a maximal amount of information regarding competency of both individuals and the entire laboratory.
The standard approaches to competency assessment for microscopic urinalysis are exams based on viewing urine specimens through a microscope or exams based on printed photos or projected slides. These traditional approaches have several drawbacks. The microscope-based approach is difficult because specimens that demonstrate the desired urine sediment structures are not always available, and when available, the specimens are difficult to preserve for use by a large group of technologists. In addition, the variability of microscope equipment can make it difficult to administer an exam uniformly across a large laboratory or group of laboratories.
An adequate collection of photos or slides can demonstrate all the relevant urine sediment structures. However, it is time-consuming to develop a photo or slide library and expensive to purchase one. In addition, it is difficult to obtain adequate brightfield photomicrographs of urine sediment structures that have low contrast, such as hyaline casts and bacteria. Finally, photos or slides cannot simulate microscope techniques that are sometimes necessary to identify urine sediment structures. These techniques include moving the microscope stage and using polarization or phase contrast microscopy.
Although clinical laboratories are required to assess the competency of their technologists in microscopic urinalysis, the results of these assessments are not published routinely. Therefore, the average competency of technologists who perform this test is unknown. Proficiency testing results are available, but it is difficult to derive general conclusions about technologist competency from these results.
For the last 6 years, the University of Washington Department of Laboratory Medicine has been developing computer programs that teach the interpretation of image-based laboratory tests and monitor the competency of individuals and laboratories that perform image-based laboratory tests [for an early review, see Ref. (5)]. The underlying premise of this work is that computers can overcome some of the problems associated with traditional methods of instruction and competency assessment. Our previous work covers a variety of image-based tasks including peripheral blood smears (6), gram stains (7)(8), ova and parasite exams(9), identification of fungi (10), electrophoresis of serum and other body fluids (11), and immunofluorescence assays for anti-nuclear and other autoantibodies(12)(13)(14)(15).3
Urinalysis-TutorTM (distributed by the University of Washington, Seattle, WA and by Bayer Diagnostics, Tarrytown, NY) systematically teaches the microscopic examination of urine sediment, and its development, implementation, and educational effectiveness has been the subject of a previous publication (16). This study focuses on a complementary computer program, Urinalysis-ReviewTM (University of Washington). Urinalysis-Review is a competency assessment program that tests the ability of laboratory personnel to identify and quantify urine sediment structures and keeps track of individual and laboratory performance over time. Here we describe the program and the results obtained when we used the program in our laboratory over a 2-year period. The study demonstrates the feasibility of using computer programs as an alternative to traditional methods for competency assessment, and it establishes baseline measurements of competency that other laboratories can use as a basis for comparison.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The images in the exams were collected from fresh urine sediments that were prepared in the clinical laboratories at the University of Washington Medical Center (Seattle, WA) and the Harborview Medical Center (Seattle, WA). The images were captured using a digital video microscopy system operated by Optimas image analysis software (Optimas). The hardware components of the system included a light microscope (Olympus model BH2; Olympus), a color CCD camera (Javelin Chromachip II model JE3462RGB; Javelin Electronics), an 80486 computer (Gateway 2000), a video imaging board (MVP-AT; Matrox Electronic Systems), and a 13-inch closed circuit television monitor (Sony) that was used to display the images. The imaging board converted the analog camera signal into a digital image that could then be saved and edited. When necessary, digital image editing was performed using Adobe PhotoShop (Adobe Systems). Editing could include contrast and brightness adjustment, color correction, and noise reduction. The goal of image enhancement was to make the images appear nearly identical to images viewed through a high-quality microscope.
Urinalysis-Review is written in Microsoft Visual Basic for Windows® (Microsoft). The program runs under Windows 3.1, 95, 98, or NT, and the minimal computer requirements to run the program are an 80486 computer running at 33 megahertz, four megabytes of RAM, and two megabytes of hard disk storage per exam. The minimal display resolution is 640 x 480, 256 colors; the optimal display resolution is 800 x 600, 256 or more colors.
Urinalysis-Review is distributed to the laboratories in our academic medical center and to client laboratories. There are currently two modes of distribution of the program. The first is a subscription service (University of Washington, Seattle, WA) in which participating laboratories receive a new exam on a floppy disk every 3 months. The second mode of distribution is through a special release of Urinalysis-Tutor (Bayer Diagnostics), which incorporates a version of Urinalysis-Review that has four exams.
A schematic diagram of Urinalysis-Review is shown in Fig. 1
. Urinalysis-Review is designed for two general classes of
users. The first class of users includes technologists or other
individuals who are taking the exam. At the start of the exam, the
examinee logs in his or her name and then is presented with 10
image-based, multiple-choice questions. There is no exit from the
program until the exam is completed. The second class of users includes
the laboratory supervisor who interacts with a control panel that
displays results and program settings. The supervisor accesses the
control panel using a password. The control panel can display the
results of individuals and the entire laboratory for the current exam,
any previous exam, or cumulatively on all exams to date. The control
panel also has a program options menu that can change the password to
the panel, back up the program's statistics file, or export the file
for use by other programs, such as a spreadsheet, database, or
statistics program.
|
For each exam, 9 of the 10 questions require the identification of one
or more urine sediment structures. Each structure identification
question has 34 possible answers, which are listed in Table 1
. For an answer to a question to be scored as correct, the
examinee must identify all structures that are present and must not
select any structure that is absent. The large majority of the
structure identification questions are on unstained urine. Some
structure identification questions allow the examinee to change the
microscope configuration from brightfield microscopy to polarization or
phase contrast. After the examinee completes the question, the
correct answer and a detailed explanation are presented.
|
An example of a structure identification question, with its associated
answer, is shown in Fig. 2
. The question (Fig. 2A
) involves the identification of
transitional epithelial cells, white blood cells, hyaline casts, and
bacteria. The user navigates through the 34 possible answers using the
"Next" button and selects the structures that are present using the
mouse. The screen in Fig. 2A
was captured when the display showed the
cell types that can be selected. If the "Next" button were chosen
from this screen, the list of possible casts would then be presented.
After casts, the order of presentation is normal crystals, abnormal
crystals, and organisms/artifacts. When a structure is selected, the
name of the structure turns from white lettering to green, and it is
then listed in the "Your answer(s)" field. In Fig. 2A
, this field
is blank, which indicates that no structures have been selected. After
choosing the answers, the examinee selects the "Done" button,
and a screen (not shown) in which a list of the correct answers
is displayed next to the user's answers. If these two lists are not
identical, the user is told that the question was not answered
correctly, and then a screen is presented in which the correct answers
are demonstrated using text and arrows pointing into the image (Fig. 2B
).
|
One of the 10 questions requires the quantification of a urine sediment structure. Each quantification question simulates the movement of the microscope stage so that multiple microscope fields can be viewed and counted. After the question is completed, the correct answer and an explanation are presented.
An example of a quantification question is presented in Fig. 3
, which tests the examinee's ability to quantify white blood
cells. In Fig. 3A
, the user is asked to quantify (on a scale of 1+ to
4+) the white blood cells present, assuming the microscope is set at
high power (total magnification, x400). The user simulates moving the
microscope stage, using the arrows displayed below the image. The white
square in the center of the arrows represents the entire microscope
slide, and the movable gray box within the square represents the
current field of view, which is shown in the image above the arrows.
After an answer is selected, the user is told if the answer is correct
(not shown) and then shown the explanation for the correct answer (Fig. 3B
). The explanation for the correct answer expresses the answer both
on the semiquantitative scale and as cells per high power field.
|
study population
The primary examinees in this study were medical technologists,
all with a Bachelor of Science degree in medical technology, who work
at the University of Washington Medical Center. The study was performed
from May 1996 to May 1998. To ensure that there was no fragmentation of
the database of exam results, Urinalysis-Review resided on only one of
the several computers in the clinical chemistry laboratory. A new
Urinalysis-Review exam was installed on the computer every 3 months,
and the study encompassed eight consecutive exams. The technologists
were required minimally to take a Urinalysis-Review exam twice per
year. However, the majority of technologists took all eight exams.
Overall, a mean of 58 (range, 44–77) different technologists took each
of the exams.
The official departmental policy regarding use of the program as a competency assessment tool is stated below:
Each technologist who is routinely scheduled on the urinalysis bench must complete a designated Urinalysis-Review exam twice each year. Results are recorded in five major areas: cells, casts, normal crystals, abnormal crystals, and organisms/artifacts. Quantification skill is also assessed. Each technologist is given a score for each area as well as a composite score for the entire exam. Acceptable performance on each exam will be an overall score of at least 70% and a score of at least 60% in each individual area. Any technologist who obtains a score below the stated criteria is required to complete the appropriate portion of the Urinalysis-Tutor as retraining. The date that retraining is completed is recorded on the exam performance summary log.
overview of exams used in the study
The eight exams used in this study had a total of 80 questions
consisting of 8 quantification questions (1 per exam) and 72 sediment
structure questions (9 per exam), which required the identification of
one or more structures. The eight quantification questions consisted of
four questions about red blood cells, three about white blood cells,
and one about granular casts. The 72 structure identification questions
covered 134 structures (mean, 17 per exam) consisting of 63 examples of
cells, 25 casts, 18 normal crystals, 8 abnormal crystals, and 20
organisms/artifacts. For the 72 sediment structure questions, 68 (94%)
were on unstained urine, 4 (6%) were on urine stained with either the
Kova variant of the Sternheimer-Malbin stain (3 of the 4 stained
urines; Hycor Biomedical) or a Sudan IV fat stain (1 of the 4 stained
urines; JT Baker). In addition, 8 (11%) of the 72 sediment structure
questions allowed the examinee to switch back and forth from
brightfield to polarization microscopy, and 1 (1%) allowed the user to
switch back and forth from brightfield to phase contrast microscopy.
The structure identification questions encompassed the majority of
microscopic findings in urine, and many structures, especially each
specific cell type, were covered multiple times.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. For this exam, the quantification question was about granular
casts. The nine structure identification questions had a total of 19
urine sediment structures to identify, consisting of 9 cells (3
questions containing white blood cells, 2 containing renal tubular
cells, 2 containing squamous epithelial cells, 1 containing red blood
cells, 1 containing transitional epithelial cells), 2 casts (1 hyaline,
1 renal cell), 2 normal crystals (1 calcium oxalate, 1 triple
phosphate), 2 abnormal crystals (1 bilirubin, 1 cystine), and 4
organisms/artifacts (3 bacteria, 1 yeast).
|
Like all the figures presented in this section, Fig. 4
is a computer
screen captured from the exam results section of the program. The text
at the top of the screen in Fig. 4
shows that the exam date is December
1996 and 53 examinees took the exam. For the 53 people to correctly
identify the 19 structures in the exam, they would need to make a total
of 53 x 19 = 1007 correct responses. Of the 1007 possible
correct responses for sediment structures, the group scored 809 (80%)
correct responses. There were 53 x 9 = 477 possible correct
cell responses, and the group scored 341 (71%; blue column in graph,
first column from left) correctly. The group had 86 missed cell
identifications. Missed cell identifications are defined in the program
as errors of commission in which a cell that was not present was
identified as being present. There were 53 x 2 = 106
possible correct responses for casts, and the group scored 76 (72%;
green column, second column from left) correctly. There were 53 x
2 =106 possible correct responses for normal crystals, and the group
scored 104 (98%; turquoise column, third column from left) correctly.
There were 53 x 2 =106 possible correct responses for abnormal
crystals, and the group scored 93 (88%; red column, fourth column from
left) correctly. There were 53 x 4 =212 possible correct
responses for organisms/artifacts, and the group scored 195 (92%;
magenta column, fifth column from left) correctly. There was one
quantification question; therefore, there was 53 x 1 = 53
possible correct answers for the group, and the group scored 37 (70%;
yellow column, sixth column from the left) correctly. The menu at the
bottom left of Fig. 4
shows the exam date field, which indicates that
these results pertain to the December 1996 exam. Results for a
different exam or for all exams cumulatively can be selected using the
mouse. The field in the bottom center of Fig. 4
indicates that these
results are group performance on the Dec 1996 exam. The other option,
"Performance over time", can be activated only when "All Exams"
has been selected from the exam date field. There are two buttons
located in the lower right of the screen. One is a help button, which
leads to instructions on how to interpret the display, and the other is
a control panel button, which leads to the general menu for the results
section of the program.
March 1997.
Fig. 5
shows the group performance (n = 56 examinees) on the
March 1997 exam. For this exam, the quantification question was about
red blood cells. The nine structure identification questions had a
total of 20 urine sediment structures to identify, consisting of 10
cells (4 questions containing red blood cells, 3 containing white blood
cells, 1 containing renal tubular cells, 1 containing squamous
epithelial cells, 1 containing transitional epithelial cells), 3 casts
(1 hyaline, 1 waxy, 1 granular), 4 normal crystals (1 triple phosphate,
1 ammonium biurate, 1 urate, 1 calcium oxalate), 1 abnormal crystal
(sulfonamide), and 2 organisms/artifacts (1 bacteria, 1 yeast). The
most obvious finding in Fig. 5
is the low score on the abnormal crystal
(13 of 56, or 23%). The results on the normal crystals (143 of 224, or
64%) were just above the minimum competency point. The group
performance on cells (485 of 560, or 87%), casts (150 of 168, or
89%), and the quantification question (53 of 56, or 95%), although
suggesting room for improvement, were well above the 60% competency
threshold.
|
group performance: cumulative results on all exams, and performance
over time
A statistical summary of the ability of the University of
Washington Medical Center laboratory staff to provide proper
microscopic interpretations of urine sediment is shown in Fig. 6
, which shows the cumulative results on all eight exams taken
over a 2-year period. These exams contained a total of 80 questions on
urine sediment and covered the majority of microscopic findings in
urine, many of the structures being covered multiple times. For the
laboratory staff to get every structure identification question
correct, they would have to give 8019 correct answers regarding urine
sediment structures. Overall, the laboratory scored 6369 (79%) correct
responses for urine sediment structures. The performance on specific
structures was 84% (3143 of 3759) for cells, 72% (1102 of 1526) for
casts, 79% (842 of 1068) for normal crystals, 65% (293 of 448) for
abnormal crystals, and 81% (989 of 1218) for organisms and artifacts.
The laboratory scored 89% (414 of 465) on the quantification
questions.
|
The group performance over time is shown in Fig. 7
. This is the screen that is displayed when "All Exams" is
selected in the exam field and "Performance over time" is selected
in the graph options field. For this graph, the program combines all
urine sediment structures into one category. The graph indicates that
the overall performance of the University of Washington Medical Center
laboratory staff on urine sediment structures (blue line in the graph)
has remained relatively constant over a 2-year period. The performance
on quantification questions (green line in the graph) varied from 70%
on the December 1996 exam to 100% on the June 1997 exam, but it does
not show an upward or downward trend.
|
individual performance: specific exam, march 1997
For each individual technologist in the laboratory,
Urinalysis-Review can show the results of a single exam, results on all
exams cumulatively, and results over time. Fig. 8
shows the performance of an individual on the March 1997 exam.
Fig. 8A
is a one-screen statistical summary of the performance on this
exam. The fields at the top of Fig. 8A
show the exam date and the
technologist's name (which has been changed for publication). Either
field can be changed using the scroll bars and clicking on the desired
date and name. The results listed in the lower left of Fig. 8A
indicate
that the technologist was correct on six of the nine (67%) structure
identification questions and the one quantification question. The nine
structure identification questions contained 20 structures to identify,
and 16 of these (80%) were identified correctly. For the individual
structure categories, the technologist correctly identified 10 of 10
examples of cells, 2 of the 3 casts, 3 of the 4 normal crystals, and 1
of the 2 organisms/artifacts. The one example of abnormal crystals was
missed.
|
The panel of buttons in the lower right of Fig. 8A
allows the
supervisor to view the questions in the exam and the answers given by
each technologist. The incorrect responses are indicated in red. Thus,
in this example, the technologist was incorrect on questions 1, 5, and
7. (In Urinalysis-Review the structure identification questions are
questions 1–9, and the quantification question is number 10.)
Selection of question 7 leads to the screen presented in Fig. 8B
, which
shows the question and the answer given by this technologist. The
answer given was "Leucine crystal", and the correct answer was
"Sulfonamide crystal". Because this was the only abnormal crystal
on the March 1997 exam, this answer was scored as 0 of 1 abnormal
crystals correctly identified, and one missed identification because
leucine was selected when it was absent.
The ability to rapidly view the answers to questions in the exam also allows the supervisor to detect error trends. Thus, by toggling through the list of technologists taking the March 1997 exam, the supervisor could determine that leucine crystals were frequently confused with the sulfonamide crystals in question 7, that yeast and red blood cells were sometimes confused in question 5 (not shown), and that uric acid crystals and triple phosphate crystals were occasionally confused in question 1 (not shown). The supervisor can analyze errors for their clinical relevance and direct continuing education efforts toward reducing the most serious errors.
individual performance: cumulative results on all exams, and
performance over time
The cumulative results on all exams for an individual technologist
(whose name has been changed for publication) in our laboratory are
shown in Fig. 9
, which is a statistical summary of the ability of an individual
to provide proper microscopic interpretation of urine sediment. Fig. 9
has the same basic features of Fig. 6
except that the statistics are
for an individual and not a group. This individual took all eight
exams, which represents a total of 80 questions, consisting of 72
structure identification questions and 8 quantification questions.
Overall, there were a total of 134 urine sediment structures to
identify, and this technologist identified 115 (86%) correctly. The
best performance (100% correct) was on abnormal crystals (red column)
and the quantification questions (yellow column). This technologist
also scored 87% (55 of 63; blue column) on cells, 80% (20 of 25;
green column) on casts, 89% (16 of 18; turquoise column) on normal
crystals, and 80% (16 of 20; magenta column) on organisms and
artifacts. A comparison with the group cumulative performance presented
in Fig. 6
shows that this technologist had an above-average
performance. The individual cumulative results were also used to detect
technologists with below average performance, and these technologists
received additional continuing education.
|
The individual performance over time for the technologist in Fig. 9
is
shown in Fig. 10
, which is similar in appearance to the screen on group
performance over time shown in Fig. 7
. For this graph, the program
combines all urine sediment structures into one category. The graph
indicates that the overall performance of this individual has remained
relatively constant over a 2-year period. The performance-over-time
graph can also identify technologists who are improving or declining
over time, or it can be used to rapidly detect a particularly
problematic exam for a technologist. That exam can then be further
dissected to define specific areas for improvement.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
).
A second achievement of this study is the publication of 2 years of
urinalysis competency data from a large clinical laboratory (Figs. 6
and 7
). Despite using a detailed literature search, we were unable to
find similar studies. These baseline results are a starting point for
our laboratory as we attempt to measure the effects of educational
programs designed to improve urinalysis performance. In addition, other
laboratories could use these data for comparison as they implement
their own procedures for competency assessment and continuing
education.
Because competency assessment results for microscopic urinalysis are not published routinely, many clinical laboratory scientists use proficiency testing results to compare the competency of their institution with other institutions. Compared with the exams and results provided by Urinalysis-Review, proficiency testing is not as useful for assessing and comparing competency in microscopic urinalysis. This is because proficiency testing is not as comprehensive and proficiency testing specimens are not handled by all technologists in the laboratory. In addition, although proficiency testing specimens should be handled like any other laboratory specimen, many laboratories give these specimens special attention because of the importance of maintaining accreditation. Thus, proficiency testing tends to overestimate competency.
The group performance results presented in Figs. 6
and 7
are probably a
conservative estimate of our laboratory's competency because the
technologists analyzed the images without urine dipstick or other urine
chemistry results and without the ability to chemically manipulate the
urine with acid, alkali, stains, or other treatments. In addition, no
patient history was provided, and the program only occasionally allowed
the technologist to invoke polarization (11% of structure
identification questions) or phase contrast microscopy (1% of
structure identification questions) and did not allow changes in
focusing. Lastly, the display quality on the laboratory computer was
slightly inferior to the computer display used to develop the
questions. This might have caused some unfairness in a few questions
related to bacteria.
Thus, although a very skilled technologist should have been able to
identify the images in Urinalysis-Review without further aid,
additional information and a better computer display would have
improved performance. For example, many technologists confused the
sulfonamide crystal in question 7 on the March 1997 exam (Fig. 8B
) with
a leucine crystal. In actual practice, technologists might take a
variety of actions when faced with the crystal in Fig. 8B
. Because
leucine crystals are rare and are associated with liver disease, most
technologists would look for confirmatory evidence such as a positive
bilirubin on the urine dipstick. In addition, a technologist,
suspecting a sulfa crystal, might perform a sulfa confirmatory test and
call the clinic to ask if the patient was taking a sulfa drug. Another
technologist error we detected was confusion of yeast and red blood
cells. In this case, many technologists would have treated the urine
with dilute acetic acid, which lyses red blood cells while leaving
yeast intact, to differentiate the two. Similarly, several crystals
would have been identified correctly if the technologists knew the pH
of the urine and the solubility characteristics of the crystals. Some
technologists who missed hyaline casts would have identified them
correctly if phase contrast microscopy and changes in focusing were
available.
Despite its limitations, the study clearly indicates that the laboratory could substantially improve its ability to identify all categories of urine sediment structures. Specifically, we should focus on abnormal crystals because the laboratory performed most poorly (64% correct) on these clinically important structures.
There are several areas where Urinalysis-Review will be improved. As
discussed above, the program would be more realistic if it included
urine chemistry results, patient history, and allowed more microscope
manipulation. In addition, it would be useful if the technologist had
the option to chemically manipulate the urine with acid, alkali,
stains, or other treatments. Moreover, it would be useful if the
results section of the program could break down the five sediment
structure categories into subcategories so that specific errors could
be more easily detected. Thus, in addition to being able to score the
general categories of cells, casts, crystals, and organisms and
artifacts, the program should be able to rapidly identify the specific
structures (e.g., granular casts, renal epithelial cells, sulfonamide
crystals) that are problematic. This is important because not all
errors are equal. For example, there is an important clinical
difference between missing a normal crystal such as ammonium biurate
and missing an abnormal crystal such as leucine. With the current
version of the program, it is possible to determine specific structures
that are problematic, but this is a time-consuming task because it
requires dissecting the results of each exam by going through each
individual's performance (Fig. 8
). A more rapid display of the
specific structures that are problematic to the laboratory would
greatly help supervisors to identify and provide education regarding
the most clinically relevant errors.
In this study, we concentrated on results that are available using the
display options within Urinalysis-Review (
Figs. 4–10
). Another
on-going study involves analyzing the Microsoft
Access® database that underlies
Urinalysis-Review and collecting survey data from technologists who
have been using Urinalysis-Review. This detailed analysis will help us
more rigorously define the current state of competence in our
institution. For example, we will be able to compare and contrast the
performance of the two geographically distinct core laboratories in our
academic institution, the first step in ensuring that the laboratories
are performing at the same level. In addition, we will be able to
determine what specific structures are commonly confused with each
other and what clinically relevant errors are made most frequently.
Lastly, the addition of the survey results might help us identify
modifiable factors that affect competency in microscopic urinalysis.
In addition to a more detailed data analysis, we are pursuing other logical extensions of our work. One obvious extension is to repeat the study after enhancing both Urinalysis-Tutor and Urinalysis-Review. This will allow us to establish even more accurate competency data. A second project is an interlaboratory comparison based on collecting competency data from other laboratories that subscribe to Urinalysis-Review. A third project is to implement the program outside the core clinical laboratory as a competency assessment tool in point-of-care settings. In this case, the examinees are healthcare personnel outside the laboratory who perform microscopic urinalysis—for example, physicians, nurses, and physician assistants—and the supervisor could be the staff member who oversees laboratory testing.
Another area of interest is to determine competency for other common microscope-based laboratory tests such as peripheral blood smears and direct gram stains of body fluids. We have developed computer programs to accomplish this. The programs, PeripheralBlood-ReviewTM and GramStain-ReviewTM, function analogously to Urinalysis-Review, but our implementation of these programs in our laboratory would need to be more systematic and formal to yield the quality of results that have been presented here.
In addition to conducting further studies, our future plans include improving the distribution and content of Urinalysis-Review. In June 1999, the University of Washington will begin the twice-yearly distribution of the program on floppy disks.4 In addition, there is an enhanced release of Urinalysis-Tutor (Bayer Diagnostics) that contains a four-exam version of Urinalysis-Review. In the United States, this can provide 2 years of competency assessment.
In addition to the current modes of distribution, we would eventually like to make the program available over the World Wide Web via a subscription service. This has several advantages. Most importantly, it would allow laboratories to anonymously compare their competency assessment results with those of other laboratories and then set quantitative goals regarding improvement in microscopic urinalysis. In addition, an on-line service should enhance convenience because technologists could take exams from any computer with Internet access and the results would reside on a master database in our institution. Currently, the program lacks an easy method for combining results residing on multiple computers; therefore, to maintain the integrity of the database of results, our technologists must take the exams on the same computer.
In summary, we have developed and implemented Urinalysis-Review, a computer program that periodically monitors competency in performing microscopic urinalysis. Results collected over a 2-year period from our laboratory showed that our technologists correctly identified 79% of urine sediment structures, and this performance has been relatively constant over time. Our future effort will focus on improving the content and distribution of the program, and performing additional studies.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
4 The cost of the program is $100 per year.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  S. Kim, M. Reeves, and M. L. Astion Web-Based Method for Establishing National Competency Benchmarks in Fourteen Areas of Clinical Laboratory Services Clin. Chem., April 1, 2004; 50(4): 753 - 755. [Full Text] [PDF]
|
|
|
|
|
|
S. Kim, P. J. Henderson, C. Phillips, A. R. Orkand, E. Maddox, C. Bien, A. Smith, and M. L. Astion Web-based Competency Assessment System for Microscopic Urinalysis Clin. Chem., September 1, 2002; 48(9): 1608 - 1611. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
