Wednesday, July 2, 2008

Lab Work

I wish I had learned more in medical school.

I did learn, what should have been a very valuable lesson, quite early on during the laboratory medicine part of our pathology course. What I was taught was:

Don't order any investigation where the result (positive or negative) will have no influence on the management of the patient.

With that knowledge in hand out I went into the world.

I learned another thing along the way.

If after talking to and examining the patient, you have not the foggiest what could be wrong, no lab test is going to help you.

Now if you read this blog, you will know that collecting blood for the lab was one of my most favourite parts of internship. Through sheer stupidity I actually used elective time to do orthopaedics thinking I would actually learn something useful. Now on ortho, every admitted patient (and back then there was very little day surgery and no same day admission) whether he was 19 or 90 got a laboratory panel of 24 tests known as the SMAC. This required 5 tubes of blood to be collected. So on my first or second day, I asked the resident, "why do we do so many tests on apparently healthy patients?". He looked me in the eye and said, "Anaesthesia wants them". Naturally all these patients were admitted after 1500 when the lab blood collection went home, which left the blood collection to the interns.

Of course "normal" results are in fact the range that 95% of healthy assymptomatic patients fall into. This means that 5% of otherwise normal patients will have an abnormal result to a test. If you order 24 tests, there is an absolute certainty that at least one will be abnormal. These abnormal tests have to followed up on which of course means more blood work.

Now did "Anaesthesia" really want them. Apparently not as I learned a few years later when I read the guidelines for routine lab work. Here are the current ones:

Now I was able to find these in about 30 seconds because I know where to look for them. If I had googled pre-operative blood work it is little more complicated but by simply looking up the guidelines of your national anaes. society, it should be quite easy. Or you could ask your anaesthesia department.

But why do I even give a shit?

1. I pay taxes

Most healthcare is publicly funded in Canada. A significant amount is publicly funded in the US. I want my tax dollars spend on treating patients not on lab work! As an aside, much of the blood work in my province is done by a private company. Do you thing they have any interest in reducing lab work?

2. Healthcare is a zero sum game

Even in the US, there is a finite amount of money that can be spent on healthcare. That means money spent in one area is money that will not be spent in other areas. Every dollar spent on unnecessary lab work is a dollar that could be spent on something useful like chronic pain.

3. It holds up things.

How often has your day in the OR been disrupted by a cancellation or postponement due to an abnomral lab test that shouldn't have been ordered that has absolutely no bearing on the patient's ability to tolerate surgery. Just the same, you have to cancel the case, or wait while the test is repeated or the necessary follow-up tests are done. Many years ago I saw a patient who had been cancelled a month ago because of an abnormal gamma GT that someone had ordered pre-op. It was elevated so she was cancelled, went for the million dollar work-up which was of course normal and was re-booked. Of course somebody ordered the gamma GT again which was still elevated. Remembering from my time in family practice that this test is a marker for alcohol abuse (or use?) I asked her if perhaps she might have had a glass of wine the night before her testing and lo and behold she said yes. I put her to sleep and she survived despite her abnormal lab test.

4. It delays necessary lab work.
A classic example is the PT/PTT. Nobody would deny the benefit of these tests in following response to anti-coagulant therapy. There are also important in the management of an evolving coagulopathy or in monitoring whether a patient has been off his anti-coagulants long enought to do surgery or stick a needle in. Except...when you order the stat PT/PTT you really need, it is going to be queued up behind all the "baseline" PT/PTTs that have been ordered. Therefore you are going to get back your PT/PTT long after you have already bit the bullet and given the FFP your patient may or may not have needed.

5. Nobody looks at it or does anything about it
Sadly, that is the case. We order all this stuff and it gets filed in the chart and nobody looks at it. This includes abnormal lab work which frequently isn't followed up on. Chest X-rays are the classic which are often ordered pre-op and reported on post-op. As I found out as a resident, actually going to X-ray and looking at the CXR is no help. The one and only time I tried that, the tech at the film library laughed at me. Actually at our hospital, CXRs are available on-line now if I could just find the time to fill out the stupid form, make up a password etc. EKGs of course go off to be reported by the cardiologist which means they sit on someone's desk until look after the surgery.

6. Surgeons actually think lab work is a substitute for a proper history and physical.
The last time I cancelled somebody as medically unfit, the surgeon's whine was, "But I ordered a cardiogram". Further people think that lab work is a static thing, like the fact that the patient's K was normal on admission, means his K is still normal after he's been vomitting for 3 days.

7. Politicians know we order too much lab work and use it as a stick to beat us on the head when we complain about lack of funding in other areas. "The ER is full of patients waiting to be admitted?. That's because you doctors order too many blood tests".

8. It leads to more unnecessary testing and interventions

I have a personal story here. My first born was born at term after an uneventful labour and was normal size for gestation. Despite this somebody decided he needed to have a blood glucose done. It was something like 3.0 which would be low for an adult but normal for a full-term neonate (because of this episode the normal range for neonates of 2.5-3.5 which of course I had to know for my recently completed written exams has stayed in my mind). Of course the lab reported that number as low using the adult range, so the first thing the nurses did was to feed him some glucose and water. This meant when my wife woke up a few hours later with swollen boobs, our son didn't want to feed. It also meant he got an extra heel prick the next day to see if his blood glucose was still "abnormal".

No comments: