This article is more than 5 years old. Information may no longer be current.
How to determine who or what is normal
Click Here to Manage Email Alerts
When we order lab tests on patients, we expect to get back a result that gives us the value for the patient together with the normal range for the test in question. If the patient’s result is outside the range, the information is usually provided in a bold font together with a notation, “H” or “L.” All that information is correct, but how meaningful is it to you or the patient?
Increasingly, and appropriately, the information is shared with the patient at the same time it is available to you. Many patients will consider a result identified as H or L to mean that something is wrong with them, and many will turn to the Internet to look up all the causes for the abnormal result. It is critical that we as clinicians fully understand how to best interpret the result and respond appropriately.
Most often, an abnormal result will confirm your clinical suspicion that prompted you to order the test in the first place, and the patient is to some extent anticipating that their test result will be outside the normal range. It is when the result is marginally outside the normal range that confusion and concern arises.
The first issue to consider is the relevance of the normal range to the patient. Most lab reports do not use the term normal but the more correct term, “reference interval.”
How often are you aware of the population from which the reference interval is established? Not very!
How often is the reference interval the same in different labs? For most common tests, there is remarkable uniformity with excellent oversight by appropriate regulatory bodies. For less common esoteric tests, uniformity from lab to lab is less likely.
How often is the population from which the reference interval is established representative of your patient? Unless you have details of how the reference interval was established, you have no idea! Stop for a moment to consider how many “normal” volunteers would need to be recruited to encompass the mix of patients in your practice with respect to age, sex or ethnicity, just to mention the most obvious variables. Next, consider the effort and expense of recruiting this reference population. Ideally each volunteer should have a complete history and physical exam. Should only patients who are not taking any medication be included, or should the lab painstakingly make sure that each and every medication in every volunteer has no effect on the assay result? To the extent possible (practical?), all labs go out of their way to make certain that the reference population is as representative and as “clean” as possible but must remain conscious of the practical limits.
The second issue to consider is the degree to which the result is abnormal. Let’s consider an assay where the upper limit of the reference interval is 10.0 mg/dL. A result of 9.9 mg/dL will not be flagged, but a result of 10.1 mg/dL will be reported as “H.” The patient who gets the 9.9 result is happy, the patient with the 10.1 will be less so.
Is there really a difference between 9.9 and 10.1? There is no easy answer to this, because it will vary from assay to assay. There is noise in every assay, and the true result may not be 9.9 ± 0.0 or 10.1 ± 0.0. It would be impractical for the lab to report the method imprecision of each analyte, but that does not absolve us from being aware that such imprecision does exist and that in some assays it is quite large.
Finally, not every patient with disease will have a lab value outside the reference interval and the patient with a value of 9.9 in the above example may indeed have the clinical situation in question. In most circumstances there will be corroborating studies you can perform in addition to repeating the original test. Examples in endocrinology include normocalcemic primary hyperparathyroidism and subclinical hypothyroidism, and examples exist in most areas of medicine.
I fully support patients learning their lab results ASAP once available and equally support patients looking into possible causes of an abnormal result. That simply makes it more important for us to be aware of the limitations surrounding lab results.
