Why Do kits Work Better or Worse in Some Laboratories;

Observations from Clearwater, 2001

By Joesph Waner,

The symptoms were familiar. It was the second day of the Symposium and I was
again enduring doubts about what my laboratory was doing and its future directions.
However, this year seemed particularly bad. Sunday’s panel discussions were more
annoying than usual because they provoked the question that I had successfully avoided
for years, namely, “why do commercial kits work so much better or worse in different
laboratories”? I would dismiss the issue if the laboratories in question were not
reputable, even excellent facilities. So, I would deal with it. First, however, it was
necessary to organize my thoughts, fortify my spirit, and come up with an approach.
Frenchy’s was open; corner table on the porch in the shade; umbrella optional; regular
grouper sandwich with a beer. Yes, a bowl of chowder would also be nice. Now, what
about those varying experiences with test kits!
This was a gathering of the best clinical virology laboratories in the country, so the
answers must be out there. Importantly, my own supervisor, Sarah, was attending this
year. Consequently, I would appear informed when I approached friends and
colleagues.
“Sarah, this is snap quiz for your own good, to see how well you respond outside of the
laboratory environment. Review for me our procedures for dealing with respiratory
specimens, particularly as they may apply to using the rapid kits”. Sarah politely began
with “as you know” and went on to detail the litany of our procedure. Now armed, I
attacked the corridors of the hotel and the aisles of the poster sessions.
Naturally, I approached representatives of the most prestigious laboratories first.
“Gordian, what are you doing about rapid testing for respiratory viruses”?
“We just got in the next generation of LightCycler. We’re looking in realtime
simultaneously for the quantity of mRNA of the principle structural proteins of 8 viruses
as well as for their genomes. A computer determines the relative contribution of each
nucleic acid. For any virus detected, the full copy is produced by PCR and the relevant
genes sequenced for drug resistance:.
Fighting a panic induced urge to flee, I asked, “What sort of specimen collection do
you recommend for this technique”?
“Specimen”, responded Gordian with a look that would have been identical if I had
asked for an explanation of string theory. He continued, “I think there is a box on the
requisition slip to check if it’s a respiratory specimen. But the nucleic acids are
concentrated from all specimens received by our specimen preparation group. Why do
you ask”?
I wanted to respond with some falsehood about not getting our LightCyler yet, but
opted for the truth. “We’re a laboratory that does immunofluorescence and have always
emphasized getting cells, in particular the types of cells that respiratory viruses replicate
in. It seems that if you do that, you get the antigens of the infected cell and the nucleic
acids”. He seemed curious, so I continued. “Our collection procedure targets the
nasopharynx”. I paused, preparing to explain where that was. “A Dacron swab is used
to loosen mucous and to gently abrade the cell layer. A saline wash is then used to

harvest the material. The wash is put into a couple of mls of transport medium and sent
to the laboratory on wet ice. Storage is in the refrigerator until testing, which occurs
within 12 hours, usually sooner. In older patients who object to the wash, the swab is
left in the back of the nasopharynx for about a minute and placed in transport medium.
Throat swabs are bad specimens and we discourage using them. We can view the cells
for numbers and type and make a judgement about quality of specimen. This approach
also gives us consistency of collection whether for fluorescence, rapid EIA, or culturing”.
Gordian paused before responding. For a moment I thought I had a moral victory.
Then, “you know, I always have been interested in cells too”.
“Nice talking with you Gordian. Have a good meeting”. A shattering start and
Frenchy’s was not a viable option for solace.
Not to be deterred, I decided to scale down on my interviewees. “Prudence, how are
you? I’m curious about what your experience is with those rapid tests for respiratory
viruses”.
“We just love them. That’s all we use now”.
“What kind of sensitivity and specificity values do you get”?
“Good, we get a lot of positives in the winter and not many in the summer”.
“Interesting. “But how do they compare with isolation in cell culture in your
laboratory”?
“We don’t do cell culture”.
“Immunofluorescence”?

“No, gave that up when we went with the kits. Since they were rapid tests too, we
didn’t want to duplicate our efforts”.
“So, you didn’t compare them with other more conventional techniques to get an idea
of how they behave in your laboratory under the conditions of your institution and patient
population”?
“No need to, the company gives you all that data in the brochure”.
“Nice talking with you, Prudence. Have a good meeting”. I considered asking about
specimen collection but lacked the courage.

“Bill, how are you? Do you use any of those rapid tests for respiratory viruses”?
“Have to, outpatient and emergency room made us do it”.
“What do you think”?
“They’re no good. When we’ve backed them up with culture, there’re a lot of false
positives and false negatives”.
Promising. After determining that specimen collection was adequate or, at least
understood, I continued. “How do you deal with the work flow from specimens coming in
at all times and your laboratory having to do tests immediately”?
“We just leave the kits on the table and when a specimen comes in, we deal with it.
The technologist can do other things while the incubations are going on. If it goes a few
minutes longer or if you have to cut it a little short, no matter, you know the manufacturer
has built in those contingencies. Besides those tests are not that good anyway”.
“But Bill, the companies all make specific recommendations about storage and
handling of reagents, incubation times, as well as specimen requirements. They pay for
independent studies to make those determinations. The better ones include that data in
their inserts and even provide references of independent studies that were published”.
“There are inserts”?
“Nice talking with you Bill, have a good meeting”.

I finished my research over dinner with friends. The lack of standards regarding
specimen collection, characteristics of individual laboratories, and neglect of scientific
rigor all seemed to be involved in producing the varying results. Strange that millions of
dollars were involved, thousand of clinical tests performed each year, and there was no
consensus as to the quality of the rapid tests. And we would be revisiting the issue
again next year in the corridors and poster sessions if not in panel discussions. I
concluded by being satisfied that I knew the performance characteristics of the tests
used in my lab and that this would be a reasonable approach for most laboratories.
“Yes, a piece of key lime pie would be nice”. It’s a great meeting.