menu
Courtesy of the US National Institutes of Health
Regulatory agencies have traditionally required trials be double-blind for good reason. Open-label trials, in which the researchers and patients know which treatment is being administered, often expose researchers and patients to biased and flawed results. But because of the way insulin is made and administered, double-blind trials in this patient population have risks as well. Achieving the appropriate insulin dose in a specific patient is a delicate balancing act. Too much insulin results in hypoglycemia, with symptoms ranging from discomfort to dangerous and even life threatening consequences. However, if the insulin dose is too low, sequela from hyperglycemia, such as headaches, difficulty concentrating, blurred vision, and fatigue (weak, tired feeling), may occur. So in insulin studies, the safest choice is often to have an insulin titration algorithm within the protocol. 

That adds safety for patients, but it creates obstacles as dosing insulin is often an art as much as a science. Investigators should be expected to generally follow the insulin titration algorithm. However, they should also be given the leeway to go outside the algorithm when it’s felt to be medical necessary, without fear of committing a protocol violation. 

Knowing that fear of hypoglycemia by both subjects and investigators is a common reason subjects get close to goal rather than achieving goal, an insulin titration committee is one way to ensure patients receive appropriate insulin doses within studies. An insulin titration committee, separate from the clinical study team, who oversees insulin dosing on preset criteria can decrease bias, especially is open label studies. 

When we run studies for insulin development, an insulin titration committee greatly increases the chance patients achieve the insulin dose that is exactly right for their needs, which has the potential to improve trial outcomes, and patients’ quality of life. And ultimately, that’s the goal of clinical research.