December 13, 2010

Robin Hood and Diabetes Meter Accuracy

WARRING:  This may be boring. In fact it may only be the first a series of boring posts. Sorry about that. I’ll do what I can to keep it moving hence all the Robin hood references. I am going to try put it into a larger context of care quality... aka happiness in Sherwood. 

That said parts of this are gonna be as dry as the deserts of the crusades from which Robin returns. Feel free to skip over and see what Kerri is writing about I’m sure it will be more interesting. Come to think of it - I may join you.

Accuracy in the context of quality care to me means the measurements needed to obtain great glucose control - kinda wordy - happiness in Sherwood sounds better. It means testing enough to manage blood sugar levels. It means those test are a good measure of the actual blood glucose. It means there has been education on effectively using the blood data. It means actually acting from that data and education. It means there is good food nutrition information, regular activity and insulin and insulin delivery that is predictable, stable and reliable even if your diabetes is unpredictable, unstable and unreliable.

In short happiness in Sherwood is a myth.

Diabetes care is complex as this bit points out:

The first step in accuracy is to test often enough to get actionable numbers and then to (and this is the really sophisticated part) act on them.  Lee Dubois speaks to the challenges of getting enough strips via medicaid. Even more drastic Jeff Hitchcock points to a Washington State study that suggests that type 1 kids don't need more than one strip a day. 

WilI and Jeff draw attention to the simple fact that access to strips is a critical part of accurately managing blood glucose.  I don’t care if those strips are as spot on as a lab reference Yellow Springs Instrument if there isn’t sufficient number to do the testing needed manage diabetes there can’t be quality care. Robin Hood needs arrows for happiness in Sherwood.

Once a patient has enough strips they need to get results from the meter that contribute to quality care. Many of us in the diabetes world talk about meter accuracy. Those conversations are frequently based on often repeated assertions that amount to whipper down the lane lack instead of being based on a solid underpinning. Accuracy is enough of an issue that the meter companies produce white papers to clarify their accuracy positions. I have read a lot of these. Typically 2 - 4 pages with a few charts and graphs and a bunch of footnotes. Sadly they are typically not the solid underpinning we can base informed conversation on. My view is the papers exist to make users feel comfortable that a given meter is accurate without any actual context. I may not go as far as to say they are Prince John’s minstrels singing his praise but it is a very tempting bit of thematic puffery to toss in - so there it is.

I have however found some good pieces. Thanks Google Scholar. One of the best that I have read is by Barry H. Ginsberg, M.D., Ph.D. (M.D. and Ph.D - that is like Doctor Squared!) Factors Affecting Blood Glucose Monitoring: Sources of Errors in Measurement is a nice little ten page run through the forest of meter accuracy issues. If you are interested in the topic this is a great place to start - unless you are looking for some kind of a score card of which meter is best. It is a great primmer on meter accuracy, it is not a betting form for the big archery competition Prince John is sponsoring. (Hint: Double down on Mysterious Stranger!)

Think of your meter as an archer at that competition. If it shoots a bunch of arrows at the target that are evenly distributed up down left and right, on average it is on mark but not precise.  Said archer it isn’t getting any attention from Maid Marion.

If all the arrows hit a spot all clumped together in the lower right side of the target - well it is precise but not accurate and still not getting any love from yonder fair noble woman.

When Robin Hood... err.. Mysterious Stranger, steps up, his arrows are all smack in the bulls eye, accurate, precise and for show, this is Marion's favor we were talking about - showing off matters, splitting the prior arrow right down the middle.

We want our meters to be Robin Hood. Barry (can I call you Barry? No! Ok... Sorry.) Dr. Ginsburg writes of four specific sources of influence that can throw off our archer’s aim. These source of meter inaccuracy are: variances with the strips, physical factors, patient issues and pharmacological problems. I strongly encourage you to read the good doctor’s paper and not solely trust the woefully inadequate analogies I am about to draw for these four horsemen of error. Here goes and remember these are the Saturday Matinée version:

Strips: Just as Robin Hood needs some serious quality control in his arrows to shoot straight, strips need to have their component parts straight and true. Strips need the right amount of the magic enzymes, in the right size little lollypop pool for blood and enzymes to react in and the tiny electrodes in the pool need to be the right size too. Olde Robin’s arrows may warp as they get schlepped around the forest so temperature, humidity, altitude, exposure and well you know the drill just about everything can effect the conditions inside that little strip.

Physical Factors: Just as the strip can be influenced physical factors aka temperature, altitude oxygen etc so can the test process itself. Mr. R. Hood has to shoot through the air and that air influences the arrow’s trajectory. Yes altitude and temperature have an effect. Alternative site test when it is cold can see even more of a lag to actual BG than normal condition.

We the Patient: We play a roll in accuracy. The good Doctor squared has a nice little chart that includes the impact of Chips Ahoy residue on BG. That is one of the more humorous thing I have seen in a scholarly discussion of meter accuracy and yes  washing hands has an impact. You think Robin Hood could shoot straight when he had grease from a slab of mutton on his fingers? There is more too, patients can miscode just as the outlaw may mistakenly use a duck feather arrow when he thought he was aiming with a goose feather arrow.  In addition to patient/archer technique a blood check may have variations due to the cellular composition of the victim’s blood. For example the red blood cell count aka hematocrit may be off. That can mess up tests. I am struggling for an analogy to the band of merry men in Sherwood forest for that one - how about: If they haven't poached one of the King’s steers recently they feel a little anemic and can't shoot. I do now I am struggling to understand why my kid’s meter doesn’t read when her hematocrit is off but I know it doesn’t. The honorable Dr suggests a GDH-PQQ or WaveSense meter and we have switched to the former as the later is the one that wasn’t working. 

Pharmacological: In short if Robin is on drugs it may impact his arrow accuracy. Various medications can influence how our meters work.

Strips, physical factors, patients and pharmacology - YDMV.

In summary, and I will be personally shocked if anyone reads this far, accuracy is about facilitating quality care. That means first and foremost sufficient checking done and acting on those checks to inform care.  So checks need to be sufficiently precise to make for quality care. Robin Hood is a myth but a serviceable metaphor for the many things that can make our meters vary. After watching the Matinée here at YMDV it is a good idea to read something more substantial.

No comments :

Post a Comment