Multivariate models of health-related utility and the fear of hypoglycaemia in people with diabetes

Aim: The aim was to statistically model the degree of fear of hypoglycaemia experienced by people with diabetes, and then model the resulting change in health-related utility associated with differing severity and frequency of hypoglycaemia. Methods: The study used pooled data from two previous postal surveys among subjects with confirmed diabetes conducted in Cardiff, UK (n = 1305 responses). The fear of hypoglycaemia was characterised using the Hypoglycaemia Fear Survey (HFS [eight question worry sub-scale only]), and health-related utility using the EQ5D(index). The data were then analysed using univariate and multivariate analysis. Results: Following detailed preliminary analysis, a two-stage approach was used since fear was important when estimating the EQ5D (index). Fear was then modelled as a function of the severity and frequency of hypoglycaemia while controlling for other factors such as diabetes-related complications. Each severe hypoglycaemic event resulted in a change of 5.881 units on the HFS. One or more symptomatic hypoglycaemic events over the same period results in a corresponding change of 1.773 units on the HFS. A 1 unit increase on the HFS results in a 0.008 unit decrease on the EQ5D(index). Conclusion: While controlling for other factors, the fear of hypoglycaemia was an important determinant of health-related utility. The magnitude of fear of hypoglycaemia was associated with the severity and frequency of hypoglycaemia. Hypoglycaemia was associated with a considerable decrement in health-related utility as a function of increased fear. Measures should be taken to minimise the severity and frequency of hypoglycaemia.