Research Gap Identification

We developed an analytic framework (Figure 1) to identify potential populations, insulin delivery and glucose monitoring methods, and outcomes gaps from the 2012 evidence report. As indicated in the analytic framework, we focused on the population of patients having type 1 or type 2 diabetes, with subpopulations defined by age categories. We did not include pregnant women as a separate category in this report because we thought it would require a separate process (and a different group of stakeholders) to adequately assess the needs for future research in pregnant women having type 1 diabetes, type 2 diabetes, or gestational diabetes (the evidence report did not include gestational diabetes and did not find any studies on pregnant women with type 2 diabetes). In the analytic framework for this study, we included the types of interventions that are currently being evaluated for delivering insulin continuously or for monitoring glucose continuously, even though we did not have any studies about the artificial pancreas or the reactive low glucose suspend pump in the original evidence report. In the original evidence report, we excluded studies of interventions that were no longer used in clinical practice, but we did not specifically exclude these new interventions.

We then searched the results and discussion sections of the evidence report, using the analytic framework, to identify potential research gaps. A Web-based assessment tool was populated with the identified research gaps specific to the induction of remission and the maintenance of remission. For each research gap category, an optional, free-text field was provided for stakeholders to identify gaps not listed in the assessment tool.

Criteria for Prioritization

The stakeholders were provided with a copy of the Executive Summary of the 2012 evidence report, were asked to read the full Executive Summary for familiarization of the findings, and were asked to independently identify the highest priority gaps for future research among individuals with type 1 diabetes and with type 2 diabetes within each of the following categories: populations, insulin delivery methods, glucose monitoring methods and outcomes. First, we asked the stakeholders to rate the three highest priority age-based populations with type 1 diabetes and with type 2 diabetes. Second, we asked stakeholders to rate the highest priority insulin delivery method and the highest priority blood glucose monitoring method for each age stratum of each diabetes type. Finally, we asked the stakeholders to rate the three highest priority outcomes for each age stratum for each diabetes type. Stakeholders were given the opportunity to add insulin delivery methods, glucose monitoring methods, and outcomes not on the list as an ‘other’ free-text option.

Engagement of Stakeholders, Researchers, and Funders

Stakeholder Engagement

We used a modified Delphi method to identify and prioritize existing gaps in the published literature as it pertains to insulin delivery and glucose monitoring methods in seven steps across four phases (Figure 2). The Delphi method involves iterative rounds of responses by group members, providing aggregated feedback about other members’ responses until consensus is reached. For each round, we used a Web-based assessment tool (SurveyMonkey™, Palo Alto, CA), with the list of the research gaps. Consensus among stakeholders was pre-defined as agreement in responses of 50% or higher in three or more options for each category of Future Research Needs. Categories that did not achieve 50% or greater consensus among the stakeholders on three or more options in phase 2 were returned for the stakeholders, with their compiled feedback from phase 2, to reprioritize.

Figure 2

Outline of key steps for identification and prioritization of Future Research Needs for insulin delivery and glucose monitoring methods.

We developed an analytic framework (Figure 1) to identify potential populations, insulin delivery methods, glucose monitoring methods, and outcomes gaps from the 2012 evidence report. We then searched the results and discussion sections of the evidence report, using the analytic framework, to identify potential research gaps. A Web-based assessment tool was populated with the identified research gaps specific to the induction of remission and the maintenance of remission. For each research gap category, an optional, free-text field was provided for stakeholders to identify gaps not listed in the assessment tool.

Research questions were developed based on feedback from stakeholders that achieved consensus during the second and third rounds. The stakeholders were presented with their compiled feedback from the second and third phases along with the research questions developed. They were asked to provide feedback on the clarity, utility, study design feasibility and priority of the research questions.

Research Question Development and Research Design Considerations

Research questions were developed based on feedback from stakeholders that achieved consensus during the second and third rounds. The stakeholders were presented with their compiled feedback from the second and third phases along with the research questions developed. They were asked to provide feedback on the clarity, utility, study design feasibility and priority of the research questions.