Demonstrating the burden of hypoglycemia on patients' quality of life in diabetes clinical trials: measurement considerations for hypoglycemia.

Objectives: To evaluate the association between hypoglycemia and health-related quality of life (HRQoL) in the context of a clinical trial using both an objectively confirmed and a patient-reported measure of hypoglycemia.

Methods: During a phase III, double-arm, randomized study, patients completed the short form 36 health survey (SF-36), a generic HRQoL questionnaire, at baseline and at weeks 24, 52, and 104. The objectively confirmed measure of hypoglycemia was based on a combination of plasma glucose measure and presence of hypoglycemia-related symptoms. The patient-reported frequency of hypoglycemia was defined as the following item: "How often have you felt that your blood sugars have been unacceptably low recently?" The association between hypoglycemia and HRQoL was evaluated in intent-to-treat patients (N = 3059) by using repeated-measurements analyses, with SF-36 scores used as explained variables and baseline SF-36 score, age, sex, country, time, and either number of objectively confirmed hypoglycemic events (0, ≥1) or patient-reported frequency of hypoglycemia (continuous variable 0-6) as explanatory variables.

Results: During study duration, less than 6% of patients experienced at least one objectively confirmed hypoglycemic event and about half the patients reported unacceptably low blood sugars "none of the time." The association between the number of objectively confirmed hypoglycemic events and HRQoL was not statistically significant, while the patient-reported frequency of hypoglycemia was statistically significantly related to all SF-36 scores (P < 0.001), except physical functioning; patients reporting greater perceived frequency of hypoglycemia had worse HRQoL.

Conclusions: Using a patient-reported measure of hypoglycemia in the context of a clinical trial could enable the burden of hypoglycemia for patients to be demonstrated.