Thursday, October 06, 2005

the problems with labs

I have nothing against labrador retrievers, or dogs in general, they're not the "labs" I find problematical. I'm referring to the labs, or laboratory reports, that list a bunch of incomprehensible words and acronyms, along with a bunch of equally incomprehensible numbers, that delineate exactly what was found in a blood sample.

You'd think that someone with as many serious and complex medical issues as me would appreciate the simplicity and completeness of the blood test for diagnosis. I am profoundly grateful that we can monitor my cancer through such a simple thing, but I'm still not too keen on the labs themselves. There's a whole medical culture built up around labs that just seems slightly off, to me. Herewith are the problems I have, and some thoughts on improving the situation.

Labs are often ordered at the wrong time. If you have to make an appointment to see your doctor for a specific disorder that will be diagnosed and/or treated via a bloodtest, it makes infinite sense to have the results of the test in hand before you go in for the appointment. How many times do you go see the doctor, only to have him order the labs, and then have you make another appointment a week later? How frustrating is this, when you already had to wait a week or two to get in for the first appointment?

I realize that there are situations where the doctor must see you first before determining what tests to order. However, there are myriad situations where that requirement does not hold. If your GP is referring you to a specialist, for example -- the preliminary guesswork should be out of the way. Or if, like me, you have a chronic condition that requires periodic monitoring. Get the labs done a week or so before your appointment, and it will make your appointment that much more productive. Ask about getting labs done beforehand whenever you make an appointment that might possibly require labs! The doctor's office can fax or mail you a lab slip, or just call the order down to the lab.

Labs are hard to get if you're a "tough stick." I have teensy-tiny little veins that just won't come up, or will roll or blow out if not handled just right. I have blood drawn frequently at different facilities, so I always tell the tech who is doing the draw these important facts, and they invariably draw me using a "butterfly," a smaller needle that regulates the vaccuum pressure from the tube so the vein won't collapse when the tube is attached. A couple of tips: relax your fist when the needle's going in; if your muscles are contracted, it will make the vein walls stiffer and the needle is more likely to bounce or glance off. Also, stay hydrated before a draw, even a fasting draw will usually allow as much water as you want -- but if you have a question about it, ask to be sure. If you're dehydrated when a blood draw is ordered, start drinking water immediately, because even a little bit will help.

Labs don't give you all the information you need. I have in front of me a copy of my most recent labs. The three columns of interest are labeled "test," "result," and "limits." The first column is the name of the test; the second the results, which is usually the amount of the first column found in the sample; the third column gives the reference range for "normal" values. That's all well and good, and certainly a ton of information is conveyed in the typical "CBC" (complete blood count) test. However, having survived an advanced education at a science-driven university, I have a couple of questions that I have never seen answered on any lab report:
What's the margin of error? Anyone who's had a decent high school science course knows that no measurement is perfectly precise, and that the amount of the mistake, the margin of error, is determined by the tools or methods used to make the measurement. We typically hear the term margin of error with respect to public opinon polls, and it's included with poll results so poll readers can make a valid interpretation. If two candidates poll at 42% and 46%, but the margin of error is +/- 4%, then all the poll is showing is that the candidates are tied. Back in the world of blood tests, the amounts of "stuff" being analyzed seem vanishingly small, but the margin of error is still relevant. Take, for example, my WBC count: 4.1 x10E3/ul; the reference range is 4.0-10.5. 4.1 is sitting at the very bottom of the reference range; 0.2 ul less, and I'd have been flagged as deficient, and my doctor would perhaps be ordering up something-or-other for me to try and increase my immune capability. With a measurement on this scale, a margin of error of +/- 0.2 ul would actually be pretty good; it's likely to be even higher -- but it's never reported. So how many people are really outside the reference ranges, but not being treated, because the margin of error is not considered?

Forget the reference range: what is the median, or most common value? This is the one that really kills me. Take my WBC count again, "normal" at 4.1, although clearly at the bottom of the 4.0-10.5 "normal" range. What I want to know, but am never told, what is the WBC count of a healthy 42-year-old woman? That range is pretty big, but I've seen worse; a friend is having her son's growth hormones tested, and the reference range was 50-300! Her son's was 52, and she figures that probably explains why he hasn't been growing, but that 52 is still "normal." The example nearest and dearest to my heart, of course, is TSH, or thyroid stimulating hormone, which until recently had a reference range of 0.5 to 5.0. However, more than 95% of the euthryoid population has a TSH between 0.4 and 2.5, and most thyroid patients feel best when their TSH is maintained right around 1.0. Isn't that a significant data point that should be considered when comparing an individual's results?

It seems that, to the typical practitioner, the distribution of values throughout the range is constant -- a straight line -- whereas the most common distribution of values would most likely be more like this:

If I'm feeling crummy and my values are at one end of the range or another, it seems to me that I might feel better if treatment could bring my values back to that central "X."

Finally, and perhaps the worst thing about labs: Doctors diagnose lab results, not patients. It has happened to me, it has happened to countless friends and aquaintances -- I know there's something not right with me, but "the labs are all fine, there's nothing wrong," the doctor proclaims. Perhaps we're patronized with, "You're just stressed, you need to relax." The last time "the labs were fine," I had cancer. The time before that, I was becoming thyrotoxic because I had low serum ferritin and adrenal fatigue.

Two things to keep in mind, if "the labs are all normal": first, "within range" is hardly the same as "fine", since we can easily see that the references ranges do not reflect the optimal value that a healthy person would have. Second, "the labs are all normal" is not the same as "all the labs are optimal." When my doctor tested me and declared me "normal", he was looking at my thyroid hormone levels, not my serum ferritin or adrenal hormones. Sure, what he looked at was fine, but he wasn't getting the complete picture.

Patients with chronic diseases or difficult diagnoses should actively manage their labs and question the easy interpretations. Ask questions if you don't like what your doctor is telling you, especially if you're hearing "your labs are fine," but you are not. Always get copies of your test results! If I had not requested a copy of my labs last spring, my thryoid cancer would have gone undiagnosed for who knows how long. Doctors manage hundreds of patients, but you only have to manage a handful of doctors. So do it!

Here, by the way, is a site that explains a lot of the typical blood test terms. It's pretty technical, but a good reference, anyway.

1 comment:

cathyf said...

Wow! Are we twins? I've gone off on rants about error bars and sample sizes, and everybody just gives me that look like I have when my husband tries to explain quantum mechanics to me... You missed the most infuriating thing, though -- if they don't find anything, you never hear from anyone.

I'm a computer programmer, and so diagnosis is something I do day in and day out -- for like a quarter century now. When you diagnose a software bug, you collect symptoms. You think. You come up with something it might be. You figure out a way to test it. It's not it. You collect some more. You think some more. Another idea, another test, not that, either. You just keep going through the collect/idea/test loop over and over, until all of a sudden, there it is! (The rush you get at that point is addicting, and keeps you coming back day after day to fight again.)

But that's software. When it's someone's life you're dealing with, somehow everything is much more relaxed. You make an appointment with the doctor. Months later, your appointed time arrives. You give a careful accounting of your symptoms. Heaven help you, though, if you leave something out because you had no way of knowing it was significant. The doctor will never ask you about anything. You know how the usual slobbering complement of a doctor goes? "He listens so well. He answers all my questions." Forget it, buddy, if I've got a question I'll go find an encyclopedia -- what I want is a doctor who asks questions. I had one, once. I was thoroughly interrogated, and he actually listened to my answers and asked more questions based upon previous answers. Unfortunately I moved away and I've never found another one.

Also, heaven help you if there is more than one thing wrong with you. Because you'll get all those symptoms all mixed up, and it will be your fault that the doctor can't figure out what's wrong with you.

Ok, so now you've said your piece, and the doctor decides on a possible diagnosis and a test. So off you go to the lab. Now computer programmers don't expect to get it on the first try, but doctors live by different rules, evidently. The test is done. Weeks later your insurance company sends you one or more explanation of benefits showing that they are paying the doctor and the lab. So you know they did it. After a few more weeks, you finally call the office. First you explain who you are to the receptionist. Then the nurse (always a different one from the one you dealt with at the appointment) comes on, and in that why-are-you-bothering-us voice tells you, "uh, all the results were negative."

That's it. No cycle of trying different tests for different diagnoses. You got one chance, and that was it.

So you think... Well, there is an excellent chance that it's nothing. Or nothing serious anyway. Even if it's some horrible fatal disease, there is a chance that there is nothing they can do to make it better. And you consider... If you call back and ask for another appointment, you'll have to go through the whole thing again. Emotionally exhausting, time-consuming. If it's hard, you'll have to do it over and over and the whole staff will become seriously pissed at you because they can't figure out what's wrong with you and it makes them feel bad. So you think, "if I just blow them all off and go away, what's the worst that can happen? I'll die. Is that so bad in comparison?"

cathy :-)