This article is more than 5 years old. Information may no longer be current.
What are the odds?
When a patient’s blood glucose levels are too good to be true.
Click Here to Manage Email Alerts
A 60-year-old Hispanic woman with insulin-requiring type 2 diabetes, which was diagnosed 27 years ago, and a history of kidney transplant presents for follow-up. The patient said she occasionally has hypoglycemia at lunch, which she treats with juice. The patient did not bring her glucometer to the appointment. However, she did bring a handwritten glucose log.
 Brett Ives |
 Ronald Tamler |
Other medical history consists of diabetic retinopathy, osteopenia and hypertension. The patient does not smoke or drink. She is taking insulin glargine 10 U at night, insulin lispro 5 U with breakfast, 8 U with lunch and 10 U with dinner. She is also taking prednisone 5 mg per day, mycophenolate mofetil, atenolol, losartan, amlodipine and cyclosporine.
The patient is in no acute distress. She is 62 inches tall, 161 lb, blood pressure is 128 mm Hg/42 mm Hg and heart rate is 62. The remainder of the exam is unremarkable except for a holosystolic murmur, which has been known for years.
An HbA1c recorded a month earlier was 9.2%.
Which recommendation for blood glucose management will be of greatest benefit to the patient?
A. Increase insulin glargine to 15 U per day to better meet basal insulin needs.
B. Decrease lispro with breakfast to 3 U to prevent lunchtime hypoglycemia.
C. Make sure the patient is using a glucometer with a memory function and ask her to ensure the time stamp is correct and to return it along with the meter 1 to 2 weeks later for further analysis.
D. Investigate for hemoglobinopathy.
E. No further action is required because the patient’s blood glucose levels have improved drastically in the past month.
CASE DISCUSSION:
Answer: C
This is a challenge that many of us have encountered in our practice: A patient with blood glucose levels that do not match the HbA1c. Although there are legitimate reasons for discrepancies, such as difficulties with coding the glucometer or rapid glycosylation, it appears that this patient’s stated blood glucose levels may not correlate with what was on the meter’s display, if she measured at all. The odds of a blood glucose level ending in “0” are one in 10. What are the odds that the patient may be experiencing almost exclusively round numbers?
This case, although particularly illustrative, is by no means an exception: There are data to show that as many as 50% of patients in diabetes clinics keep inaccurate blood glucose logs or do not measure blood glucose. A blind decrease in prandial (B) or increase in basal insulin (C) may lead to worse dysglycemia.
In this case, the best management is to ask the patient to return with the glucose meter and make sure it records the blood glucose values. Many meters are now downloadable, which enables visual analysis of the data together with the patient. Depending on the individual patient, a direct confrontation with the observation may be of benefit. However, it may also alienate the patient or lead to a “perfection of technique” when it comes to making up blood glucose values in their log book.
Brett Ives, MSN, APRN, CDE, and Ronald Tamler, MD, PhD, MBA, are both of the division of endocrinology at Mount Sinai School of Medicine, N.Y.