Inappropriate use of hemoglobin A1c test?

[Nunavut Boy]

Our lab has recently purchased a small-volume hemoglobin A1c/microalbumin analyzer. Since obtaining the ability to perform A1cs in house instead of waiting a week for results to come back from Ottawa, I would estimate that demand for this test has increased 1000%.

I was taught that A1c testing was for MONITORING KNOWN DIABETICS to see if they're keeping their glucose under control, not DIAGNOSING new diabetics. I was taught that diagnosing new diabetics was done one of two ways:
-FBS greater than 7.0mmol/L
-2hr OGTT glucose greater than 11.1mmol/L


In our remote communities, it is nurses who do all of the diagnosing and blood test ordering. It is these communities in particular who I think are abusing this test.

We also have no pathologist here or any pathologist we consult with. This is a bad idea in my opinion, but I am not the boss. Many times have I been consulted as an 'authority' on laboratory science.

So, the following are the questions I would be asking my pathologist, if I had one.

-Since the A1c test is 10x more expensive than a glucose test, can we reject any test that comes in that is not from a known diabetic? How is this handled in other hospitals?

Wow. That is one long, rambling OP with very little payoff. I apologize to anyone who has gotten this far.

[Dr_Paprika]

By and large, I'd agree with you. Diabetes is, in practice, suspected by asking about its symptoms, finding sugar on a urine dip or having an elevated random blood sugar. In practice, it is usually diagnosed after a fasting serum glucose. The OGTT is a high maintenance test, although a little more accurate, I think most clinicians reserve it for borderline cases especially in pregnancy.

The Hb1Ac test might be of value for some borderline cases of diabetes, but if no diagnosis of diabetes NOR glucose intolerance exists than it probably isn't economical.

[Nunavut Boy]

Yes, I am glad I seem to be thinking properly on this subject. But, do I (as a lab tech) have any authority to deny tests, even if we have no pathologist?

[KarlGauss]

HbA1c (and related) tests are NOT currently recommended for the diagnosis of diabetes (http://care.diabetesjournals.org/cgi/content/full/26/11/3160). In addition to the point made in the link, another reason is that although it is a specific test (i.e. a high value rules in diabetes), it is not a sensitive test (i.e. some people with diabetes will have a normal level). Screening tests should be relatively inexpensive and sensitive (i.e. you want to detect as many cases as possible by screening, even if it means making a few 'false positive' calls).

[Qadgop the Mercotan]

My colleagues have pretty much nailed this one. I just wanted to share how I diagnose Type II diabetes, which is pretty much the accepted standard. And in my setting, I diagnose a lot of new cases, 4 or 5 a month.

1) If they have a random serum glucose over 200 mg/dl with symptoms.
2) If they have a fasting serum glucose which exceeds 126 mg/dl on 2 different days.
3) If they have a postprandial glucose (2 hours post meal) over 200 mg/dl.

Meet any one of the 3 criteria above, and it's diabetes type II.

And you observe that there's no need for a HgbA1C in this diagnostic plan! Nor is a GTT really needed either.

But a US lab tech doesn't have the leeway to reject the order. Don't know if it's different north of 60, or in the rest of Canada.

[Nunavut Boy]

No, technically I don't have the leeway to change orders, but in practice, because we don't have a pathologist to consult with on these things, someone has to do it.

I'll let it be my boss. I don't get paid enough to lose my license over this kind of stuff.

I think what I will do is write a paper introducing the new test, what it is used for and what its limitations are and fire that off to all the communities. Perhaps this will decrease some of the orders.

We shall see! Thank you all for the info.