When I started my career in laboratory medicine, we tested any fluid that was handed to us, for any analyte requested by the doctor. We did this for a number of reasons that we thought were good ones including that the doctor is a medical professional who knows what he wants and needs, and that the test results will help diagnose and treat the patient. We were trying to be helpful. Along the way though, laboratory professionals have come to understand that testing like this may not provide accurate results and may be doing more harm than good.
Now days, CLIA has clearly mandated that if the manufacturer of an FDA-approved assay system has not validated that system for a specific fluid type, the lab must perform that validation before testing and reporting results on that fluid type.
This is sometimes a hard rule to explain to the medical staff who have been trained in medical school to order such things as amylase on peritoneal fluid to look for pancreatic injury, or glucose on nasal fluid drainage when a CSF leak is suspected. And these doctors often have literature references for what they wish to have measured, although in general the references are not recent. I have a three-pronged approach to the explanation I give doctors as to why I won’t analyze the sample they sent me.
First, and probably most importantly, I cannot guarantee the accuracy of the result. Matrix effects are real and a test designed for serum will not perform the same on urine. Similarly, a test designed for serum and urine will not perform the same on a pharmacy preparation or an ascitic fluid sample. The result I provide if I test that sample could very well be wrong.
Secondly, I have no way to interpret the results of the test on an un-validated fluid type. There are no established reference intervals that allow us to determine the meaning of the result we’re providing. For example, who knows how much glucose is normally present in nasal drainage? I would assume no one knows, because why would you measure it in normal nasal drainage, and for that matter, what constitutes normal nasal drainage? Thus if I test that unknown sample for the analyte requested, I’m providing a possibly inaccurate result that is uninterpretable. And the physician is going to treat the patient on the basis of that result. In most cases, the physician changes his or her mind at this point in the discussion.
However, if that isn’t enough, I bring out the big guns. The agencies under which the lab operates forbid me from analyzing this sample for this analyte unless I validate the sample fluid type in my lab using the stringent validation criteria described in CLIA. This validation would take a considerable amount of time and resources and enough patient samples to set a reference interval.
If a doctor would still like to be able to order that test on that sample type after the discussion, I request that the doctor be involved in the validation process. First of all, I will want to know that enough of these tests will subsequently be ordered that the time and effort spent to validate the test will be worth it. In addition, the doctor will need to collect sufficient numbers of that sample type to allow us to perform an adequate validation and reference interval study. Also, QC material with the same matrix as the fluid type will need to be used and may have to be made in-house as it’s generally not commercially available. Biannual proficiency testing for that fluid type will need to be performed also, with internal PT developed for it. All of these considerations mean that the number of body fluid types and analytes we have validated is small, but we do not analyze un-validated fluid types. We will often try to locate a reference lab who does analyze them for a doctor. When that fails we will try to help the doctor find an answer to his medical question through use of other, legitimate tests.
-Patti Jones PhD, DABCC, FACB, is the Clinical Director of the Chemistry and Metabolic Disease Laboratories at Children’s Medical Center in Dallas, TX and a Professor of Pathology at University of Texas Southwestern Medical Center in Dallas.
Lablogatory 