Medicare Strikes Again

Last week I received a nasty notice from a Walmart pharmacy, to which I had sent a prescription for blood sugar test strips for a slender, 72-year-old woman that I saw recently, referred by her primary care physician for diabetes. Her diabetes was poorly controlled. She has multiple myeloma and has been on various treatments, including high-dose steroids intermittently.

Her condition had been diagnosed as type 2 diabetes by her primary care physician, given that she was an elderly patient and had been on steroids. Yet she was on insulin alone, unlike most patients with type 2 diabetes who are on multiple oral meds or non-insulin injectables. She had been on these in the past, had no benefit from any of them, and was very sensitive to insulin (that is, she required very low doses of insulin to adequately control her sugars and to avoid becoming hypoglycemic), the antithesis of standard type 2 diabetes, but classic for type 1 diabetes (which is due to insulin insufficiency, rather than to insulin resistance as in type 2 diabetes).

I had measured a C-peptide level, a surrogate for insulin levels that reflects insulin produced by the patient's own pancreas rather than the exogenous insulin injected by the patient. The C-peptide level was very low, indicating insulin deficiency, type 1 diabetes, explaining her poor response to type 2 meds and her insulin sensitivity. I therefore diagnosed her with type 1 diabetes. While this is unusual for an elderly patient to present with type 1 diabetes, it is hardly rare, and I have seen many cases of this particular misdiagnosis. It is a common experience for Endocrinologists.

I had tried to get the patient to use a continuous glucose monitor, which is standard care in type 1, as well as in type 2 diabetes patients on insulin (it should be for type 2 patients not on insulin, as use of these devices alone improves sugars rather dramatically, just by providing continuous information to the patient on ambient blood sugar levels, a biofeedback-type process). I had even talked to her about using an insulin pump. She declined both of these and wanted a refill on her test strips. These I sent in for her.

Whereupon her Walmart pharmacy sent me the aforementioned nasty note. The gist of it was that I had sent an invalid diagnostic code, and therefore the prescription would not be filled. The code that I had sent was E10.65, which indicates type 1 diabetes with high sugars. The note informed me that her prior prescription, sent in by her primary care physician, was E11.65, which designates type 2 diabetes with hyperglycemia. Since my diagnosis was not the same as the original diagnosis, my diagnosis was deemed invalid.

I called to talk to the pharmacist about this. It was of no avail for me, as I was only allowed to talk to a tech, who simply reiterated that the code I had sent was not valid. I advised her that the code was indeed entirely valid, under which Medicare coverage provides three test strips a day. (Never mind that the patient was testing as much as 5 or 6 times a day—this is to her credit and good care, as her diabetes is quite brittle, and sugars unpredictable, and it is dangerous for her to test less, given the possibility of hypoglycemia, which can be catastrophic. Medicare deems that testing three times a day is sufficient for everyone. Hence the coverage of testing three times daily for type 1 patients as well as type 2 patients treated with insulin. As far as Medicare coverage is concerned, it makes no difference if you send a code for type 1 or type 2 diabetes for a patient on insulin. The pharmacy and Medicare both know she is taking it—they have the prescriptions to confirm it—and not just insulin, but multiple daily injections of insulin, the most intensive type of insulin management short of insulin pump use.)

So I re-sent the prescription under a diagnosis code that specifies only long-term current insulin use, without specifying type 1 or type 2 diabetes. I don't want to send the prescription under a type 2 diagnosis code, as that would be false coding, but Medicare seems to insist on false coding and remains unwilling to accept a different code despite that code being incorrect.

This behavior of Medicare is a first in my experience. I wonder if all the focus on fraud and abuse in these government programs has led leadership, such as Dr. Oz in the case of Centers for Medicare & Medicaid Services (CMS), to implement AI filters to detect fraud. If so, my prescription didn't get past the filters. One thing Medicare may be focusing on is the possibility that physicians are illegitimately changing codes to get more test strips covered (quite a picayune concern, if that's the case) and attempting to forestall such fraud. So perhaps I am guilty of Medicare fraud for providing a correct diagnosis code that differs from the primary care diagnosis code.

I do tend to believe that the ethos of Dr. Zeke Emanuel holds sway at CMS. He is the one who famously said that we shouldn't have to pay for medical care for anyone over age 70, on the grounds that they no longer are useful to society. (How is that for pushing granny off a cliff?) Soylent Green redux. So perhaps I can be excused for believing CMS is seeking new ways to deny care to Medicare beneficiaries, even perhaps by targeting their doctors for fraud.

Or, giving the benefit of the doubt, perhaps Dr. Oz just didn't set the AI filters quite right to spot Medicare fraud, and a few tweaks are all that is needed to avoid erroneously excluding benefits.

Please excuse me if I seem to believe that Medicare is actually trying to kill my patients. I could be wrong. Still, I would advise anyone championing Medicare for All to be careful what they ask for. As if any such person would listen.