Do Clinical Trials in Diabetes Measure the Wrong Things?

The lack of significant progress in treating diabetes has led some experts to question whether clinical trials (which are, after all, the major factors in deciding optimal treatment regimes) measure results which are actually important to patients. Mayo Clinic researchers decided to an analyze the design of clinical trials to see what proportion of them were focused on effectiveness with respect to what they called ‘patient-important outcomes’, such as illness, pain, function, and death. They published their results recently in the Journal of the American Medical Association.

Public-available data-bases were used to identify over 2,000 registered clinical trials, but these were narrowed to about 1,000 using eligibility criteria. Of these, half – in fact, 436 - were randomly selected for full analysis. The trials were categorized according to their measured outcome:

Physiological changes - e.g. insulin levels, C-reactive peptide levels

Surrogate outcomes – e.g. cholesterol levels, kidney function tests

Patient-important outcomes – e.g. illness, pain, loss of function, death

The distribution according to these 3 categories was 16%, 61%, and 18%. Patient-important outcomes were noted as being of primary or secondary importance in only 46% of trials.

Clearly, the outcomes of most (61%) of these trials cannot be considered as immediately relevant for their participants. The trials in both the surrogate and physiological changes groups will not provide adequate information on the effect of interventions on patient-important outcomes.

This problem is pointed up by the results of some large studies on the intensive treatment of diabetes to achieve a given low level of the hemoglobin A1c (HbA1c). Two of these have just been published online in the New England Journal of Medicine, while I have commented recently on a British study that found that concentrating on blood sugar levels in type 2 diabetics may not be as productive as one would hope. Clearly we need to refocus attention on patient-important outcomes in finding the best treatment for diabetes.