Lab Test Ranges

[Luang]

Anybody have knowledge regarding the (often significant) reason hospitals/labs can have such differenes in their test ranges? It can cause confusion as i have just found out.

Here's a scenario for you: I very recently had a test carried out at St. Louis' for (x). The result came back at 51 with their lab range of (30-65), so would seem normal, so this part of me is ok...right?

However, when tested at Samitivej a few years ago for the same thing the result was 49 (even lower than now) but their range was (0-40) so considered High and not ok!

When i mentioned this to a doctor (not working at either of these institutions) that if i compare the recent (St. Louis) result with Samitivej's ranges then the result is still high, he said, 'no. no you cannot do that' and have to go with the range of the testing hospital. Ok, but what is confusing me is this, if i had chosen to be tested at Samitivej instead of St. Louis then my result would be exactly the same (51) would it not? (they use same units). And therefore considered above their range and High.

So, with one hospital i am fine and healthy but with the other i am not and sick?! What am i missing here?

A confused Luang

P.S. Sheryl and FBN.........Help!

[FBN]

Testing biological samples is not exact and even within the same lab, a test done a week later may have a slightly different result. The biological environment is continually changing within the "normal" parameters. To be able to judge whether the results are indeed abnormal, the specific test needs to be known. Some tests are based on a direct measurement and some on an indirect measurement. Indirect will obviously have more inherent inaccuracy associated.

Hospitals use different analytical methods on different lab equipment so values my change and also reference values, that is why they will publish the reference value valid at the time of testing for certain tests. On any one machine, a variance of 5-10% in the value (depending on the test) of the same test is accepted.

Individual tests should also be evaluated as part of the complete picture, not one test alone.

[Sheryl]

You also need to check what unit is being used for the denominator, as some values and ranges may be per ml, dl (deciliter) or liter etc.

If your example is literal, it is a bit puzzling. Different cut-offs for high and low values are common, but it would be unusual for one hospital to have a range starting at 0 and another at 30. Would help to know what the lab test was.

In addition I would add that the normal ranges are set taking into consideration the prevalence of certain conditions in the local population. for example, alpha-fetoprotein (AFP)is a "marker" for liver and other cancers. in the West, this would not be part of standard check-up, but because Asians have a comparatievly high incidence of primary liver cancer (rare in the West), it is often included in Thai packages and, again because of the high incidence of liver cancer in the Thai population, the upper end cut-off point is often set much lower than it would be in the west. The normal range is anything less than 10 nanogram/ml, but some Thai hospitals use less than 7. This is not because people with levels a bit over 7 necessarily have a problem (most don't) but because, in the Thai population, further evaluation of such people will succeed in identifying a few cases of early liver cancer. It would not make sense to apply the same cut off to other populations, but reference values are set for the majority.

Which is why results always need to be reviewed in the context of the individual patient, especially borderline results.

[ronnight]

Also need to remember that the normal ranges are really two standard deviations from the mean, so it is normal for about 1 out of 20 people to have a result out of normal range but the result is normal for that individual

[jimmym40]

All info above is true. I was told once that it also depends on the technology of the equipment used in performing tests. Maybe the reference ranges come with the equipment, and that is what is published or given to the patient. Every test result there is a possibility of different readings from day to day. I guess the best thing to do if you are worried or way out of balance, is have the test done at several different places.

[Pacificperson]

There is another factor in determining "normal." It is the levels at which medical opinion says that intervention may be required. Think of cholesterol levels. A large percentage of the population has cholesterol levels above 200 or whatever the current mark is, so if your going on the 95% distribution, then these are "normal", but medical opinion says that these levels put you at risk. So, the results come back as outside the normal range. It is a heads-up to the docs to pay attention.

[Sheryl]

There is another factor in determining "normal." It is the levels at which medical opinion says that intervention may be required. Think of cholesterol levels. A large percentage of the population has cholesterol levels above 200 or whatever the current mark is, so if your going on the 95% distribution, then these are "normal", but medical opinion says that these levels put you at risk. So, the results come back as outside the normal range. It is a heads-up to the docs to pay attention.

That's what I was getting at in my 3rd paragraph above..and that the cut-off for a "heads up" is very much influenced by the type of problems common among a particular patient population.

In the case of tests which are markers of organ damage or cancer, it is also useful to know how far from the norm a value is likely to be in the presence of serious disease. In many instances, serious disease would yield values 10, 100, even thousands of times the normal range; in such instances a borderline value or mild elevation is of much less significance than for indices where serious disease corresponds to values just a few points above the norm. Again, goes back to what test is under discussion.