Barriers to using technology to manage diabetes have included concern over the efficacy and reliability of the newest tools available. A recent study examines the clinical impact of new speech-recognition software that promotes the goal of modifying bolus doses by calculating a premeal insulin dose based on a voice description of the meal. These findings support the ongoing need to evaluate the efficacy of new technology applications through the use of robust scientific exploration to ensure confidence in both the patient with diabetes and healthcare providers responsible for managing their chronic disease.

Patients diagnosed with diabetes are continually afforded new, enhanced technologies to manage their disease. Not long ago, insulin pumps were considered among the “newest advances” in achieving blood glucose control; however, many patients are now using an insulin pump as a routine part of their diabetes treatment. Among the challenges faced when using the pump is the requirement to correctly calculate insulin bolus doses to compensate for elevated blood glucose following a meal. Although modification of the calculated bolus dose should be based on the type and amount of food consumed (estimating the amount of carbs, protein, and fat in each meal), most patients do not make such modifications.

Establishing stable blood glucose using appropriately modified boluses has become easier with readily available smartphone applications. The barriers to using of any new technology include the uncertainty about the reliability of computer software replacing the human factor. A new study published earlier this month aims to test a system supporting determination of modified bolus doses based on a voice description of a meal and is among the few studies that uses the strength of a randomized cross-over trial to evaluate the findings.

Foltynski et al evaluated the bolus calculator VoiceDiab, which consists of a smartphone app with three remote servers for automatic speech recognition, insulin dosage calculation, and text analysis. Forty-four people with type 1 diabetes treated with continuous subcutaneous insulin infusion were randomly assigned to either the VoiceDiab-supported group or an unsupported control group (in which human calculation was used to define bolus doses). After a 14-day washout period, patients from the supported group were switched to the unsupported group and those in the unsupported group were switched to the supported group.

The researchers noted that there was a significant difference between the supported and unsupported groups in the percentage of patients with 2-hour postprandial blood glucose within the 70- to 180-mg/dL range (58.6% vs. 46.6%, respectively; P = 0.031). The authors concluded that the technology evaluated in their study improves postprandial glucose control without causing either hyperglycemia or hypoglycemia and, therefore, the technology may be useful in treating patients with diabetes on intensive insulin therapy.