THE Patient State Analyzer (PSA 4000; Physiometrix, North Billerica, MA) is an FDA-approved device that quantifies regional electroencephalographic data and converts them into a linear scale (“score”) from 1 to 100. This score has been shown to correlate with depth of anesthesia. 1In this report, I describe a case in which the device was instrumental in alerting the anesthesiologist to hypoglycemia.
A 60-yr-old, 98-kg, white man presented to the operating room for a laparoscopic resection of an insulinoma. His past medical history was significant for multiple episodes of altered mental status, associated with confusion and disorientation, over a 3 to 4 month period. On each occasion, his blood glucose concentration was 30–40 mg/dl. He experienced neither seizures nor coma during these periods of profound hypoglycemia. His only medication at the time was fluvastatin (a cholesterol-lowering drug). The initial diagnostic workup revealed an inappropriately high concentration of insulin after a prolonged fast and an increased concentration of C peptide. Magnetic resonance imaging and computed tomography scan of the abdomen showed no abnormality. Despite this preoperative inability to localize the source of excess insulin, the patient was scheduled to undergo an exploratory laparoscopy, with possible resection of an insulin-secreting tumor.
Other significant past medical history included occasional gastroesophageal reflux disease and mild multiple sclerosis, causing minimal right-sided weakness. Significant laboratory findings on the morning of surgery included a blood glucose concentration of 61 mg/dl and a potassium concentration of 4.8 mEq/l. His heart rate was 82 beats/min, and his blood pressure was 135/75 mmHg.
On arrival to the preoperative holding area, the PSA 4000 array was placed on the patient's head, and 30 ml sodium bicitrate was given orally. He was brought into the operating room, where intravenous access was established, and an infusion of 5% dextrose in 0.9% sodium chloride was started at 100 ml/h. The device was activated, and the patient's initial PSA value was 97. The standard American Society of Anesthesiologists monitors were applied. Metoclopramide, 10 mg, and midazolam, 2 mg, were administered intravenously. After preoxygenation, a modified rapid sequence induction with 2 mg/kg propofol and 0.25 mg/kg cisatracurium was performed while maintaining cricoid pressure. The trachea was intubated approximately 90 s later.
After induction, the patient's blood pressure and heart rate remained within normal limits, and his PSA score decreased to 19. Anesthesia was maintained with 50% oxygen, 50% air, 0.2% isoflurane, and neuromuscular blockade. Several minutes later, his score increased to 38, as the effect of the induction dose of propofol diminished. His heart rate and blood pressure remained unchanged. On insertion of the urinary catheter by the surgeons, the patient's heart rate increased to 90 beats/min, and his blood pressure increased to 160/88 mmHg. Fentanyl, 250 μg, was administered intravenously. This resulted in a decrease in both heart rate and blood pressure to their baseline value. However the patient's processed electroencephalography scan remained essentially unchanged. An arterial catheter and a second intravenous catheter were placed.
Over the next 15 min, during preparation and draping of the patient, the PSA value gradually decreased to 17, and the isoflurane was discontinued. Despite this adjustment, his electroencephalography value remained essentially unchanged. During this time, there were no significant changes in heart rate or blood pressure, and the patient remained normothermic. An arterial blood sample was drawn, the analysis of which revealed a blood glucose concentration of 40 mg/dl, an arterial oxygen tension of 171 mmHg, and an arterial carbon dioxide tension of 32 mmHg. Intravenous dextrose, 12.5 g, was given rapidly. Approximately 2 to 3 min later, his PSA value increased to 48, and his blood glucose concentration was 235 mg/dl. The isoflurane was restarted at 0.2%, while the surgeons continued to prepare and drape the patient. Several minutes prior to incision, the isoflurane was titrated up to 1.9%. The patient tolerated the incision, and the isoflurane was gradually decreased. The patient remained stable throughout the procedure and was maintained with 50% oxygen, 50% air, and isoflurane between 0.9 and 1.5%.
An infusion of 10% dextrose in water was administered intraoperatively to maintain a blood glucose concentration between 100 and 200 mg/dl. Frequent blood gas analyses were performed to direct this goal. Laparoscopy revealed a tumor of the pancreatic islet cells. After enucleation of the tumor, the patient's blood glucose concentration increased to 359 mg/dl. The 10% dextrose in water and 5% dextrose in 0.9% sodium chloride infusions were discontinued, and an infusion of a balanced salt solution was started. At the conclusion of the surgical procedure, the patient awoke, the trachea was extubated, and he was transferred to the postanesthesia care unit. In the postanesthesia care unit, his blood glucose concentration stabilized, and he was transferred to the inpatient ward later that evening. He was discharged home on the third postoperative day. The patient denied any intraoperative awareness or memory.
There has been much controversy over glucose management during the resection of an insulinoma. 2–4Although a continuous intravenous infusion of glucose is not absolutely indicated, it is commonly used to achieve the goal of maintaining a blood glucose concentration greater than 50–60 mg/dl. 4–6Frequent intraoperative blood glucose measurements are necessary. 7
Hypoglycemia can cause permanent cerebral damage. 5The signs of hypoglycemia include hypertension, tachycardia, sweating, change in mental status, and even bradycardia, which may be secondary to an increase in vagal tone. 6Many of these signs are masked by general anesthesia. 4,7Those available are nonspecific and have multiple etiologies. In this patient, the heart rate and blood pressure were unchanged. The only indication of possible hypoglycemia was his low PSA score. His heart rate and blood pressure remained at their baseline. The PSA score of 17 reflected a very deep anesthetic state, despite an end-tidal isoflurane of zero. The patient's blood glucose concentration at that time was 40 mg/dl, and, at this concentration, he was known to be symptomatic. Another indication that the score was affected by blood glucose concentration was its rapid increase to 48 following treatment of hypoglycemia.
The PSA 4000 monitor utilizes regional quantitative electroencephalographic data. When a general anesthetic is administered, there are regional changes in electroencephalographic activity that occur regardless of the anesthetic used. 8These regional electroencephalographic changes can then be quantified. A score of 100 corresponds with the awake state, whereas a score of zero corresponds with complete electroencephalographic suppression. Hypnosis during general anesthesia is reflected by a score of 25–50.
In this case, the patient's score decreased immediately on induction of anesthesia from 97 to 19. His score promptly increased to 38, indicating an adequate depth of anesthesia. However, during the period of preparation and draping, his score began to decrease to very low levels without any measurable end-tidal volatile anesthetic. The end-tidal isoflurane had been stable at 0.0%, but his score continued to decrease to well below 25. Hypoglycemia was considered as a possible cause, and the patient's blood glucose concentration was checked, confirming this etiology. When a bolus of glucose was given, the score promptly increased. Of note, hypertension and tachycardia, common signs of hypoglycemia in the anesthetized patient, were not seen. Whether these changes in vital signs would have ultimately occurred is not known.
In this case, a hypoglycemic patient had significant quantitative electroencephalographic changes, as reflected in the PSA 4000 score, while all vital signs remained within normal limits. Further studies are needed to determine if there are consistent electroencephalographic changes seen in the hypoglycemic patient, and whether these changes routinely precede alterations in hemodynamic variables.