What's in Your Blood?

I am sure that you have seen all those credit card advertisements with the tag line: "What’s in you wallet?"

I was thinking about this a couple of days ago when I heard about the husband of a friend of ours who had been in a really nasty motor vehicle accident which left him with several broken bones. After a few days he started to develop problems with his breathing and the family noticed that he was bruising and hemorrhaging from his scars. Yet some of his lab values appeared to be "normal." Fortunately, I was able to offer some advice, and his recovery is now back on track.

But the case reminded me that many physicians have not been taught the value of looking at laboratory values as a whole. Too often we see computer-generated sheets that highlight the "abnormal" values. Yet this is not quite right. First, however good the equipment, laboratory values are never totally accurate. There is always an associated measurement error. This is like those political opinion polls where the expert tells us that the error is + 3%. So if candidate A has 47% and candidate B has 50%, you cannot tell who is winning. Maybe candidate A really has 50% and candidate B really has 47%. Let me give you two clinical examples. I have seen a young doctor who wanted to give someone a blood transfusion because their hemoglobin had fallen by one gram, without realizing that the lab error was + 0.5. So when I asked him to repeat the measurement, there had actually been no change. And I have seen countless people get worried because someone’s temperature has gone up by a degree or two, even though this is within the normal measurement error, and is, by itself, quite meaningless.

And that leads me to the second point: an elevated temperature, or a low hemoglobin, is simply a guide; it is one part of the whole gestalt that we use for diagnosing and monitoring illness. It is wrong to treat a lab value.

Third, when we look at lab values, we are interested in the overall pattern, rather than one result here or there. So I might have a sheet of results in front of me without a single highlighted value, and yet be able to discern a pattern that points toward a specific illness. An individual’s blood chemistry can tell an eloquent story: if you know how to read it.

A fourth point is that for many laboratory tests, there is no such thing as "normal." We use what are known as "reference ranges." Blood levels that have been measured on a lot of apparently healthy people, that are lumped together and used as average reference value for that population. Any deviation is regarded as "out of range," but it does not necessarily mean that it is abnormal. If we go and do another reference range in China or Africa, we may get entirely different results. I once dealt with a large number of patients from some islands in the Caribbean who had two apparent problems: many of them has been misdiagnosed with a condition called acromegaly, a condition in which adults begin to produce large amounts of growth hormone. It can be a nasty illness, and it was apparently responsible for the death of Earl Nightingale, one of the early leaders of the personal development movement. The thing was that amongst these islanders, elevated growth hormone levels were normal and caused no ill health. Second, many ran high levels of an enzyme called creatine kinase that is commonly used to diagnose myocardial infarctions or heart attacks. So we could not use the enzyme for diagnosis in these people unless we saw an extra big change. And that’s an important reason for measuring hormones, chemicals and temperatures: not just to get one-off values, but to monitor change. Once again bearing in mind that any measurement will likely have some level of inaccuracy associated with it.

This last point about reference ranges is exceedingly important. I have seen countless people who have been diagnosed with "subclinical" thyroid or adrenal problems because the treating clinician was not familiar with the interpretation of endocrinological investigations. In a later entry, I am going to discuss why many endocrinologists have become a little skeptical about so-called illnesses like "adrenal exhaustion" and "sub-clinical hypothyroidism."

Technorati Tags: laboratory tests,  medical diagnosis, disease