Objective: We reviewed the depth and length of hypoglycaemia in a cohort of patients undergoing Insulin Tolerance Tests (ITTs). We evaluated the safety of the test, its reproducibility and explored factors that might predict the optimal dose of insulin.
Methodology: ITTs were performed at a teaching hospital in the south of England between 2005 and 2010.
Result: The major indication for ITT was non-functioning pituitary macro adenomas. 76% of the cohort was hypoglycaemic (<2.0 mmol/l) for 60 min or more. The nadir plasma glucose (NBG) ranged from 0.14.6 mmol/l and correlated significantly with basal plasma glucose (BPG) (r=0.56; P<0.001), Insulin dose (r=0.27; P<0.001) and weight (r=0.21; P<0.01). 24 patients received an insulin dose exceeding 0.15 IU/kg body weight. These patients were characterized by higher weight (mean 93 vs 86 kg) and BPG (mean 5.7 vs 4.8 mmol/l) compared with the rest of the population.
Using multiple regression analysis, the factors determining nadir blood glucose were plasma glucose (b=0.56; P<0.001 20% contribution) and weight (b=0.14; P<0.05 2% contribution).
Only one patient had an adverse effect during the test. He developed unstable angina and needed coronary artery by-pass surgery following a finding of 3 vessel coronary artery disease. The mean NPG and insulin dose in this patient was comparable with the rest of the population.
Discussion: The limitations of the ITT include the perceived risks of prolonged hypoglycaemia. Our data shows that it is a relatively safe test but that hypoglycaemia can be prolonged and unpredictable in a significant proportion of patients.
We have also shown that baseline plasma glucose and the patients weight predict the nadir plasma glucose.
We therefore propose that a better way to avoid unnecessarily prolonged hypoglycaemia is the use of an insulin and glucose infusion with bed-side plasma glucose analysis.