One of the more difficult things to deal with in medicine – I think – is borderline lab values. A huge component of most medical decisions relies on the patient’s latest labs; often the decision is almost perfunctory. But what about times when lab values are just barely positive? For example, a normal blood pressure is 120/80. What if a guy has 122/80? Is that hypertension? Should I put this disease on his problem list…there for insurance companies to see and thus charge him more for life or health coverage?
Most lab values can be manipulated to some degree. The entire field of biofeedback starts from this premise. In biofeedback models, the lab values can be changed by effectively willing your body to change it’s own parameters. If effect, you can think yourself to a lower blood pressure, lower cholesterol, more hair…maybe even bigger sexual organs I suppose. Go for it. Imagine big, big gonads if you want. See what happens.
Anyway, some people consider this mere quackery – I think it will be considered a fad in a few years – but some swear by it. And yeah, I’m sure there is some bridge across the Cartesian mind-body quagmire we’ve been wallowing in since Modernism awoke, but I’m not sure biofeedback is that bridge. Aside from whatever volitional input we may have on our bodies, there is the fact that all lab tests have margins of error associated with them. You hear about “sampling” error in all these perpetual polls in the presidential primaries. Obama leads Clinton 47 to 42 percent…but the margin of error is plus or minus 8 points. So Obama doesn’t lead anybody. He and Clinton are tied. Many of the good polling companies actually do some real statistics to come up with these numbers, but it doesn’t really matter much because none of us voters really care. All we care about is who’s gonna win the stupid thing? Who’s the prez gonna be? And, (for me) WHO’S FINALLY GOING TO PUT AN END TO DYNASTIC, BOURGEOIS RULE IN AMERICA?
Sorry, this is a blog, so I can digress and feel only a little bad. The points are that lab values are manipulatable both physiologically and statistically. The statistical manipulation – the margin of error, for example – exists because determining something like the amount of glucose molecules in a sample of blood isn’t exactly easy. There’s 2 major ways to do it, and a 3rd less-popular way as well. Each way is different from each other, so if you determine glucose values using Type 1, and then use Type 2, you have to convert one to the other’s value to even determine if they’re equivalent. The conversion is a math equation, not observational science.
Why does an average doc like me care about all these details? Well, these facts play out in my mind when I get a lab value that is juuuussst barely out of range. A positive value means the patient has a new disease that can affect their ability to get health insurance (a quaint little reality of this American health system I admire so much), but also can affect their care. One little lab value can determine if a person suddenly needs to take expensive medicines – brand-name cholesterol-lowering meds are a couple hundred bucks a month, for example. Or it may mean that they now qualify for surgical exploration to find a possible tumor. Some people have had things cut off of them – like breasts – only to find that they didn’t have cancer at all. This is called the false-positive rate and is one of the biggest reasons why I try very hard NOT to do even little tests on patients unless it really looks like they need it. Breast self-exams are a classic example of dangerous false-positive tests that lead to sometimes catastrophic interventions occasionally for no reason. **Disclaimer on this one: major controversy exists about whether or not breast self-exam are good or bad. This was just an example. Keep doin’ em if you do ’em.**
When pregnant, one of the things we screen for is gestational diabetes (we call it GDM). This disease only shows up during pregnancy, and often vaporizes just after the baby is born. It more resembles Type 2 DM than Type 1 in that it won’t really kill you in the short-term, but deranged glucose values give the patient headaches and lots of nausea and makes them feel pretty much miserable. The problem is that high glucose in the mom can lead to huge problems with the baby. First of them being that the baby can be, in fact, HUGE. But they can also end up obese in life. They usually end up with REAL diabetes. Being so big, there are risks for the delivery that aren’t pretty too.
So, we need to avoid GDM. We screen for it by giving the mother an oral glucose test at 28 weeks. She drinks this sugary solution – absolutely disgusting…I’ve tried it – and then testing her blood sugar levels an hour later. This is a classic screening test: It has lots of false-positives, but fewer false-negatives. It’s reasonably good for catching GDM. But because of the fasle-negatives, we do a second test to try to filter some of them out, which is basically the same test with more sugar solution and blood tests every hour for 3 hours. People who have out of range glucose values for both tests, probably have GDM and need treatment which can include testing glucose values 3-5 times a day (rather painful, a bit bloody, and not cheap). The might also get insulin shots, based on the glucose values. They are usually sent for nutrition counseling to teach them how to eat basically no-carb foods. They also get regular ultrasounds every month or less to determine if the baby is getting too big, which may lead to C-section. And while we’re on the topic of false-positives, ultrasound becomes less and less accurate for determining weight as the pregnancy progresses. By the end of the 3rd trimester, the US could be off by as much as 2 lbs! An alarming US can easily lead to surgery when in fact the baby is of normal weight.
So, my patient had a 1-hour glucose test of 199. The cut-off is 150. That’s obviously positive – she qualified for the 3-hour test easily. The first test of the 3hr is a fasting glucose test (arguably the most important) and she was well under the limit. Her second test, at one hour, was also well below the threshold for GDM. Third test – at 2 hours – also negative. Then, at 3 hours, her test was 164. The cut-off was 160. She has a positive value. She may very well have GDM.
You be the doc. What would you do?
Here’s some possibilities:
A.) Bring her back for another 3h test, since they can be falsely positive (and, actually falsely negative). In other words, they aren’t always accurate.
B.) Give her the full-monty: regular US, daily glucose monitoring, insulin, diet counseling, weekly visits, evaluation for c-section
C.) One or two elements of option B – a partial intervention, so to speak
D.) Tell her the test is “fine” and leave it at that. Effectively, regard the test as negative, and tell her this.
E.) Bring her in and just test her glucose twice a week. Treat her if one of those values is out of range.
**I’ll describe what I did in a week or so.**