This article is more than 5 years old. Information may no longer be current.
The finger-stick blues
A circumspect physical exam can make a big difference for the patient.
Click Here to Manage Email Alerts
A 94-year-old woman living in a nursing home was admitted to the hospital for severe hypoglycemia after a finger-stick glucose was found to be less than 20 mg/dL.
Routine point-of-care glucose measurements were being monitored at the nursing home because finger-sticks had indicated hypoglycemia during two recent hospital admissions for a urinary tract infection and pneumonia with sepsis, respectively. She had no history of diabetes or hypoglycemia before these recent episodes. No hypoglycemic symptoms were noted at the time of the severely low finger-sticks.
 Rifka Schulman |
 Ronald Tamler |
The patient’s past medical history was significant for hypertension, congestive heart failure, atrial fibrillation and dementia. The patient’s chart noted a history of an unspecified “lung disease,” and a CT report noted multiple lesions, possibly indicating lung metastases.
Her medications in the hospital included torsemide, cefepime, vancomycin and subcutaneous heparin. Metoprolol (given for rate control of her atrial fibrillation) was stopped on admission because of concern for hypoglycemia unawareness. The patient was unable to provide a complete review of systems or give a good medical history. According to her chart, she did not have a history of tobacco, alcohol or drug use. She did not have any known drug allergies. Family history was noncontributory.
On admission to the hospital, the patient’s oxygen saturation dropped to 70%, and bilevel positive airway pressure (BiPAP) was started for hypoxemic respiratory distress. Physical exam demonstrated an elderly, cachectic woman wearing a BiPAP mask with an impaired mental status. Vital signs included: blood pressure 140/92 mm Hg; pulse 82 bpm; and temperature 37.3·C. Her heart sounds were normal and lung sounds were decreased bilaterally. She had 2+ bilateral pitting edema.
During the course of her hospital stay, the patient had multiple episodes of finger-sticks as low as 10 mg/dL or less, and she was started on an infusion of dextrose 10% in water (D10W) at 100 cc/hour. Serum glucose measurements were all normal or elevated. Other laboratory data included a creatinine of 1.1 mg/dL; mild elevations in all liver function tests; decreased albumin of 2 g/dL; and morning cortisol of 25.5 mcg/dL.
The patient’s edema worsened, and she continued to require BiPAP. An endocrine consult was called. On the day of the consult, the patient’s finger-stick glucose measurement was 33 mg/dL and a nearly simultaneous serum glucose was 246 mg/dL.
 The patient’s
cyanotic fingertips, next to her daughter’s normal fingers.Photo courtesy of: Rifka
Schulman, MD |
What is the most likely explanation for this patient’s low blood glucose levels on finger-stick, and what is the next best step in her management?A. This patient most likely has an insulinoma. Check a C-peptide, insulin and beta-hydroxybutyrate level during the next hypoglycemic episode.
B. Check the patient’s roommate and nurse for hidden insulin vials because she may be a victim of surreptitious insulin administration.
C. The patient needs to be ruled out for adrenal insufficiency. Check a 250-mcg cosyntropin stimulation test.
D. Regardless of the etiology, the patient requires more dextrose to avoid brain damage from severe hypoglycemia. Increase the D10W to 200 cc/hour.
E. Perform a finger-stick glucose with a different glucometer and check for causes of pseudo-hypoglycemia.
CASE DISCUSSION:
Correct answer: E
The most likely diagnosis is pseudohypoglycemia (E). Although the patient’s impaired mental status makes it difficult to discern classic symptoms of hypoglycemia or Virchow’s triad (symptoms, documented hypoglycemia and resolution of symptoms with euglycemia), the disparity between finger-stick glucose and serum glucose from venipuncture should raise suspicion. The first step should be to make sure the point-of-care glucometer is working and maintained properly. However, in this case, a closer exam of this patient with congestive heart failure and hypoxia from severe lung disease caused us to hold our breaths: The patient had noticeable cyanosis of the finger tips (See photo). The finger-stick glucose measurements from the cyanotic fingers were artificially low due to impaired digital microcirculation leading to local increase in glucose consumption, and thus not representative of systemic hypoglycemia. This explains why simultaneous serum glucose values were always normal or high. In fact, as the cyanosis improved, her finger-stick values did as well.
It is unlikely that this elderly woman with multiple comorbidities, including malnutrition and liver dysfunction, also had an underlying insulinoma (A) or was being subjected to surreptitious insulin administration (B). Adrenal insufficiency, although indeed a risk factor for hypoglycemia, was unlikely, given the robust morning cortisol value (C). Increasing the rate of dextrose infusion would have likely worsened this patient’s edema and pulmonary status and caused hyperglycemia because the finger-stick values are not correct (D). The patient’s family and nursing home were advised to no longer check finger-sticks, and the infusion of D10W was discontinued.
Rifka Schulman, MD, is a clinical fellow, and Ronald Tamler, MD, PhD, MBA, is assistant professor, both in the division of endocrinology at Mount Sinai School of Medicine, N.Y. Dr. Tamler is also an Endocrine Today Editorial Board member.
Disclosures: Drs. Schulman and Tamler report no relevant financial disclosures.