Toward Development of Psychosocial Measures for Automated Insulin Delivery

Abstract

The INSPIRE study working group launched its initial workshop in February 2015 to facilitate collaboration among key stakeholders interested in automated insulin delivery (AID) systems and the psychosocial outcomes of individuals who may use these new technologies. Specifically, the INSPIRE team’s goal is to facilitate measure development assessing the psychosocial factors associated with AID systems. A second working group was held to foster exchange among key stakeholders in AID system development. Patient, health care provider, engineering, industry, academic, regulatory and payer perspectives were presented. The INSPIRE working group will continue to serve as a platform to encourage open dialogue among all stakeholders with the aim of facilitating technology that offers minimal user burden and maximum benefit from both a psychological and physiologic perspective.

On June 5, 2015, at the beginning of the 75th Annual Scientific Sessions of the American Diabetes Association, the INSPIRE Study Group (Insulin Delivery Systems: Perceptions, Ideas, Reflections and Expectations) convened their second open session, inviting the diabetes community to participate in developing a uniform group of psychosocial measures for use with automated insulin delivery programs underway globally. The first meeting was held at the 8th international meeting of Advanced Diabetes Technologies and Treatments for Diabetes (ATTD) in Paris in February 2015. While glycemic control (HbA1c, time in range, avoidance of hypo- and hyperglycemia, etc) remains a fundamental outcome for any automated insulin delivery system, self-care behaviors on the part of patients (and family members) will continue to be fundamental (and necessary) components for assessment in any approved system; thus a battery of patient-reported (and family-reported) outcomes (PROs) will need to be developed and deemed acceptable by all pertinent stakeholders. This second meeting included presentations by a number of these important constituencies.

Jill Weissberg-Benchell PhD, CDE, introduced the goal of the INSPIRE team to the group, highlighting the key stakeholders in the development of measures assessing the psychosocial aspects of automated insulin delivery systems (also known as AP systems). Specifically, Dr Weissberg-Benchell reviewed the importance of capturing the unique needs of children, teenagers, emerging adults and adults of all ages, as well as parents and partners/spouses. The importance of capturing the needs and concerns of health care professionals, industry, payers, and regulators was also reviewed. The ultimate goal of INSPIRE is to develop reliable, valid, and consistent measures that can be used by all AP systems to assess the psychosocial factors associated with this new technology.

The Patient Perspective

Dana Ball from the T1D Exchange and Aaron Kowalski PhD from JDRF presented information to the group about the patient perspective. Mr Ball discussed some of the reasons that people with diabetes are interested in AP systems, including the emotional burden of the disease and the possibility that such systems might reduce burnout, fears of hypoglycemia, and fears of long-term complications. The enthusiasm for such systems can be seen by the fact that GLU had over 50 000 hits on their discussion of the bionic pancreas. Mr Ball also discussed the importance of considering the availability of such systems when they are commercialized. Availability can be understood from both the perspective of affordability (including out-of-pocket costs) and from the perspective of health care professionals’ (HCPs) willingness to support patient access. Concerns about eligibility criteria were also raised by Mr Ball—will there be a “subset” of patients that are “appropriate” for such systems, or will all interested patients have equal access to the commercialized AP systems?

Patient concerns about the safety and reliability of the systems was raised, with the point being that if the systems are not reliable and patients cannot trust them, then they will not be successful products. Issues related to the patient’s perspectives regarding the wearability of AP systems were also discussed by Mr Ball. Considerations regarding the comfort patients experience while using such systems was raised, as was consideration regarding the number of insertion sites needed. In keeping with the theme of comfort and wearability, Mr Ball raised the importance of the intuitiveness and ease of use of AP systems to meet the needs of patients. Finally, Mr Ball discussed the need for patients to be able to have access to their data and share access with loved ones and HCPs as they choose. Issues related to ease of uploading data and the platforms that can support data access were discussed as well.

Dr Kowalski picked up the discussion regarding patient perspective on automated insulin delivery systems by pointing out that what users want, what health care professionals want, and what payers want may be different. Specifically, he suggested that users want decreased burden, HCPs want improved glycemia, and payers want decreased cost. He encouraged the workshop members to consider all perspectives when developing measures that can assess outcomes and inform interventions. Dr Kowalski also raised the important balance between glycemic control and daily burden for the person with diabetes. With respect to burden, he raised the issues of time spent interacting with therapy (including time spent bolusing, pushing buttons, resetting basal rates, frequency of blood sugar checks, frequency of site changes), and the burden of alerts/alarms. He reminded the group that “added value” can mean different things to different stakeholders and all must feel that AP systems add value. He suggested that engaging in a workflow analysis of automated insulin delivery systems versus current sensor augmented pumps systems may provide additional data to inform uptake and retention of these systems.

The Health Care Provider Perspective

Anne Peters, MD, presented her thoughts from the health care provider perspective. Dr Peters raised the concern that many HCPs find technology to be overwhelming, and that they often do not receive the necessary training to master the technology that they are expected to support. She raised the concern that the increased time necessary to care for patients (time spent face to face with patients as well as time spent negotiating for coverage of AP systems with payers and time spent completing forms for insurance companies) is rarely paid for, takes up more clinical time, and could decrease physician availability to meet the needs of their patients. Dr Peters expressed concern that this increased, non-reimbursed time may not even lead to improved patient outcomes. She also raised the concern that increased time spent discussing and interacting with technology during the clinical encounter may negatively impact patient/provider relationships. With respect to patient outcomes, Dr Peters wondered if the automated insulin delivery systems are addressing the multiple areas of diabetes care that HCPs see as problematic, for example, achieving benchmarked clinical outcomes associated with “pay-for-performance” metrics. Dr Peters encouraged AP teams to develop training curricula and training programs that can be disseminated to providers so that they are able to master the technology. She wondered if such training would be mandated by the ACGME, as mandated training will naturally lead to increased comfort with these systems, and lack of training will impede clinician’s comfort in recommending AP systems’ use. Dr Peters also wondered where the technical and customer support for the systems will come from, and expressed the hope that each AP system will have its own team to provide such patient support.

The Engineering Perspective

Ed Damiano, PhD, spoke about the engineering challenges inherent in designing systems aimed at automated glucose control. Dr Damiano expressed the belief that the key question system developers need to address is whether or not the new technology will make life easier for people who live with diabetes. In addition, automated glucose control systems need to be designed so that they are robustly adaptive to all possible patient needs in addition to meal times and exercise, including growth and development, intercurrent illness, hormone fluctuations, and technology failure. Dr Damiano suggested that the key concern around system failure is when the CGM system has an error or fails. He states that all automated insulin delivery systems must have a mechanism for handling those events safely.

The Industry Perspective

Marie Schiller offered the industry perspective on automated insulin delivery systems. She stressed the importance of language when discussing such systems, recommending that developers clearly describe what their systems can do so that patient expectations are realistic. For example, will the glucose control system be fully automated or semiautomated? Ms Schiller similarly encouraged AP developers to focus on the impact of these systems. Specifically, data focused on how such systems impact the patients and their families can help stakeholders determine the best methods for capturing loyalty and repeated use of particular systems. Finally, Ms Schiller encouraged AP system researchers to determine which outcomes (other than HbA1C) matter, and to clearly explain why these outcomes are important.

The Payer Perspective

Mr Aaron Davis offered the payer perspective. He reminded the audience that there are multiple payer types, and that the needs of both commercial payers and government payers (Medicare and Medicaid) must be considered. Consistent with Ms Schiller’s message of the importance of language, Mr Davis underscored the importance of clearly communicating the key benefits of AP systems to the payer groups with data and key outcomes that they can understand. As AP systems are a unique and different platform than either pumps or CGM, AP developers have the unique opportunity to teach payers what outcomes are important in AP system use. Mr Davis suggested that framing outcomes in a cost/benefit manner will be helpful. For example, discussing the reduction in time spent in hypoglycemia saves health care dollars, which therefore makes the benefit of the AP system worthwhile.

The Regulatory Perspective

The workshop’s last presenter was Stayce Beck, PhD, from the FDA, who offered the regulatory perspective. Dr Beck reported that the FDA has worked closely with almost all of the automated insulin delivery system developers and believes in the importance of close working partnerships with these research groups. Dr Beck reviewed the FDA’s dedication to validated and reliable assessments of patient-reported outcomes and stressed the need for such data to be collected as part of the pivotal trials instead of being added at the end of the study, as an afterthought. Dr Beck reminded the audience that AP systems are class III devices, and are therefore subject to premarket approval, which is a process involving scientific and regulatory review for evaluating the product’s safety and effectiveness. Dr Beck stressed the fact that PROs can help the FDA when they look at the risk/benefit profiles of new technologies. Ms Beck encouraged AP system developers to gather data on whether or not patients value the benefits of the AP system more than the risks, and the methodology used to assess patient preferences must be clear, appropriate, reliable, and valid.

Summary and Future Directions

The audience expressed substantial interest regarding this topic of psychosocial measures for use with automated insulin delivery systems. There remains a need to consider the perspectives of the multiple stakeholders as the INSPIRE Study Group continues to make progress. The salient take-home features from this session include the following: It is important to build on Dr Kowalski’s point that the AP systems have to add value, and that added value will likely be calculated differently by patient, HCP, industry stakeholder, payer and regulator. It is also necessary to attend to Dr Damiano’s point about making life easier for patients and caregivers. Moreover, the systems offered from industry will likely provide a staged approach to automated insulin delivery, beginning with semiautomation; such products need to articulate their limitations along with their advances through realistic expectations. The measures under development by the INSPIRE study team should be able to assess whether or not the systems, indeed, are “making life easier.” It is similarly important to keep in mind that payers, just like regulators, need to be part of the conversations early on. This highlights the need to include both clinical and patient-reported outcomes in all assessments of AP device performance, with careful consideration given to cost-benefit analyses. Finally, to ensure adequate uptake of automated insulin delivery systems, one must consider HCP burden (Dr Peters’s points) and how to ease those burdens.

The next open meeting of the INSPIRE group will take place at the 9th international meeting of ATTD in Milan on Wednesday, February 3, 2016, from 12:30 to 2:00 pm. All are invited.