Increasing Awareness and Uptake of Connected Insulin Pens for Eligible Patients With Diabetes: A Quality Improvement Success Story Free

We are part of an outpatient network of clinics at Stanford Health Care, a large academic medical center within the suburban/urban San Francisco Bay area of California, with an ethnically diverse patient population. We have three practice settings: two adult primary care clinics and one adult endocrinology clinic. The primary care practices are internal medicine faculty practices; one has 20 clinicians and two pharmacists and the other has 11 clinicians and one pharmacist. The endocrinology clinic has five endocrinologists and two pharmacists. The demographics of the patient population served by these clinics include sex: male 41%, female 59%; race: American Indian 0.45%, White 45%, Asian 31%, Black 5%, unknown or other 19%; ethnicity: non-Hispanic 85%, Hispanic 11%, unknown 4%; and age: 18–49 years 51%, 50 to >90 years 49%.

Pharmacists at Stanford Health Care practice under a collaborative practice agreement with the clinicians as directed by the state of California’s Pharmacy Practice Act (1,2). Under this agreement, the primary care providers refer patients with poorly controlled diabetes, defined as an A1C >8%, to the pharmacists, who perform certain patient care functions such as initiating insulin or intensifying diabetes medications, based on a set protocol. Endocrinology providers also refer patients with poorly controlled diabetes or patients who may need assistance with starting new devices or medications. Because the pharmacists are certified diabetes care and education specialists, they can also provide comprehensive diabetes care management.

Before initiation of this quality improvement (QI) project, connected insulin pens (CIPs), sometimes called “smart” pens, were infrequently prescribed to our patients who were on multiple daily injection (MDI) insulin therapy. In the two primary care clinics, CIPs were not prescribed by either the primary care providers or the pharmacists. Before this project, the pharmacists in the primary care clinics had no training in initiating CIPs. A subset of pharmacists and endocrinologists were prescribing CIPs.

The American Diabetes Association encourages the use of diabetes technology to improve the lives and health of people with diabetes but also recognizes that the rapidly changing landscape of diabetes technology can be a barrier to adoption and implementation of devices (3). In our practice, we recognized an opportunity to educate clinicians and patients about CIPs to increase their use among people who might benefit from them. Compared with conventional insulin pens, CIPs improve adherence to insulin therapy regimens (4) and increase the availability of data for providers to streamline counseling strategies and optimize therapy plans (5).

This program focused on increasing the use of CIPs for people with type 1 or type 2 diabetes who were on MDI therapy. Initially, we prioritized individuals with an A1C >8%, but we later extended the program to all eligible individuals on an MDI regimen. When discussing CIPs, we included a diabetes technology attitude questionnaire to facilitate shared decision-making.

Because of prior work with the T1D Exchange Quality Improvement Collaborative (T1DX-QI), we recognized that CIPs were usually not being offered as an option to our patients on MDI insulin therapy. The data analyst on this project obtained baseline data on CIP use through the Stanford Research Repository (STARR) database. At Stanford Health Care, patient encounters in the electronic health record (EHR) system are available in STARR for use in research and QI projects.

For data reporting, we aggregated data for our three clinics. In May 2022, there were 2,660 adults with a diagnosis of type 1 or type 2 diabetes for at least 1 year who had at least one visit that month. Of these patients, 7% were on MDI insulin therapy, and none had a CIP on their medication list.

This was a 9-month QI project that began in May/June 2022 and ended on 28 February 2023. The intervention initially targeted eligible patients seen by the pharmacists involved in the project; it subsequently expanded to other patients within the primary care clinics and then to those in the endocrinology clinic. For this project, we used the Plan-Do-Study-Act (PDSA) model (6). Data were collected at the end of each cycle.

Our core team included a primary care physician, two clinical pharmacists working in the primary care clinics, a data analyst, and a post-baccalaureate intern trained as a medical scribe. The primary care physician was the project leader and had previously led a QI project for patients with type 2 diabetes as part of the T1DX-QI. Each member of the team had an interest in advancing diabetes technology use for patients.

Our team set up weekly virtual meetings to review each PDSA cycle and discuss new interventions. We coordinated with a CIP company representative to provide virtual CIP training for the primary care and endocrinology pharmacists. We provided an in-service to primary care clinicians to increase their knowledge about CIPs and encouraged both primary care clinicians and endocrinologists to refer patients directly to the pharmacists for CIP initiation and diabetes management. We implemented a prescription ordering workflow (Supplementary Figure S1) to follow up on all CIP prescriptions ordered and to keep patients informed about the status of their CIP, which was carried out by the intern on the project.

We developed a shared decision-making tool, which was a patient handout comparing insulin delivery devices (Supplementary Figure S2A and B). We also improved the efficiency of getting CIPs by incorporating various support staff such as our intern and local durable medical equipment (DME) representatives in troubleshooting insurance coverage issues, as depicted in our prescription ordering workflow (Supplementary Figure S1). We plan to continue this workflow through our pharmacy technicians, who already provide prescription troubleshooting support to the clinical pharmacists.

For our endocrinology clinic, we developed smart phrases, which are abbreviations that allow for longer phrases to be pulled into a patient note in the EHR system. Smart phrases were used in communication with patients who were eligible candidates for CIPs. Phrases included background information about CIPs, instructions for setting up their CIP, and information about sharing their CIP data with the clinic. We also developed workflows to streamline the process of obtaining CIP reports from patients after they started using a CIP; these included using smart phrases describing instructions for emailing or faxing reports to the clinic before visits.

• Plan. For all patients with diabetes who were on MDI insulin therapy, were seen in the endocrinology clinic, and had an A1C >8%, we sent a message to their primary endocrinologist asking if they would be appropriate candidates for a CIP. If the endocrinologist felt that a patient would be a good candidate, an EHR message on behalf of the endocrinologist was sent to the patient describing the QI project and the CIP. These messages invited patients to make an appointment with the pharmacist if they wanted to learn more. An in-service training was given to the endocrinologists about the QI project, and they were invited to discuss CIPs with their patients and refer interested patients.

• Do. We did this for 4 weeks.

• Study. After 4 weeks, we were able to complete outreach to 47 patients with a 23% response rate (13 patients) and a 13% CIP start rate (6 patients). Half of the patients who were started were contacted directly by their endocrinologist or pharmacist.

• Act. Given the increased success when providers discussed CIPs directly with their patients, we modified the approach by focusing more on providing education to providers and staff. We also streamlined the method for obtaining patient CIP data to increase provider and staff support and satisfaction.

After 9 months, we observed an increase of 4% in the number of patients using MDI insulin therapy who started using a CIP (Supplementary Figure S3). The chart showed three shifts determined by control chart rules (7). The first shift occurred during the baseline period of January to May 2022, the second shift was during June to September 2022, and the third was during October 2022 to January 2023. There was an increase from 1.52 to 2.86% from the baseline period to the second shift and a second increase from 2.86 to 5.40% from the second shift to the third shift.

We found that patients were more receptive to discussing a CIP if this topic was initiated by the provider who helped to manage their diabetes rather than through a secure patient message in the EHR system from someone with whom they had no prior relationship. In the endocrinology clinic, patients were receptive to having another provider—the pharmacist—start their CIP if they were initially referred to the pharmacist by their endocrinologist.

Using a shared decision-making tool that compared CIPs to other insulin delivery methods helped patients understand whether starting a CIP would be right for them (Supplementary Figure S2A and B).

We will continue to provide in-service trainings to other primary care providers about the utility of CIPs and to encourage referrals to our pharmacy clinics for starting patients on a CIP. We will also continue to train our pharmacy providers on ways to use the shared decision-making tool to help determine whether their patients with diabetes could benefit from a CIP. We will also start to incorporate CIP discussions into our new patient visits.

We encountered several obstacles in ordering CIPs and insulin cartridges. There was a lack of awareness among community pharmacies about CIPs, and they often required clarification on prescription orders. Pharmacies often wanted clarification between insulin cartridge versus CIP orders. We adapted to this need by adding specific comments in the orders, with extra clarification. Another ordering obstacle was that patients with Medicare insurance required that their CIP prescription be sent to the DME company directly but their insulin cartridge prescription go to their local pharmacy. This requirement caused confusion for both local pharmacies and patients.

Having engagement from different staff members allowed us to gain more knowledge on bridging the gap between the pharmacy and prescriptions being sent. We created a workflow for our intern to follow up on prescriptions (Supplementary Figure S1). Our intern called local pharmacies and followed up weekly or as needed until patients were able to obtain their prescriptions. This step helped us troubleshoot insurance coverage issues in a timelier manner. We also found that working with our local DME representatives helped us better navigate coverage issues.

Our first attempt at increasing shared decision-making around using a CIP was through administration of the Diabetes Technology Attitude (DTA) survey (Supplementary Figure S4). The intent of the survey was to gauge interest in diabetes technology, which we hoped would lead to a discussion about CIPs. This survey was sent to patients through secured patient messaging in our EHR portal or administered through virtual visits with clinicians. We had a 61% response rate for completing the survey, and although we found this tool to be helpful in introducing the topic of diabetes technology, it was not an efficient way to discuss CIPs specifically. We stopped administering the DTA survey after 2 months.

Our second attempt at increasing shared decision- making was through outreach to patients through secured patient messaging using a smart phrase with a description of CIPs. We found the description to be helpful but still had low patient engagement.

Our final attempt at increasing shared decision-making was through the creation of an infographic comparing CIPs to other insulin delivery methods (Supplementary Figure S2A and B). This infographic was shared with patients during visits or sent through EHR patient messages. We felt that including images for patients to review was a more helpful way for patients to participate in shared decision-making.

Before this project, we did not have a standardized way to document shared decision-making in progress notes. With each patient encounter at which CIPs were discussed, we incorporated a smart phrase into our progress notes to document shared decision-making.

Patient outreach about CIPs primarily occurred through EHR messages or through direct provider contact in the clinics. We initially sent outreach to our own pharmacist patient panels and then expanded to other providers’ panels (Supplementary Figure S5). We found patient outreach messages were best received if they came from a provider with whom patients had a prior relationship. Overall patient interest was very low in the patient panels of providers who were not directly involved in this QI project.

We found training patients for a CIP was best done face-to-face because of dexterity concerns and questions related to CIP app setup. Many patients needed assistance with setting up the app and understanding all of its components and features. We learned that patients could best use a CIP if they had a preexisting level of comfort with smartphone apps and could read English.

As previously mentioned, there was some delay and confusion with obtaining two different prescriptions needed for 1) a CIP and 2) insulin cartridges. Therefore, patients needed reminders about obtaining both prescriptions and assistance coordinating with their pharmacies and DME company. Transition from traditional insulin pens to a CIP was also delayed for some patients who had a large supply of insulin pens at home that they wanted to finish before starting a CIP.

CIP data could be obtained various ways, and during this project, data-sharing was implemented through a CIP data portal, which was helpful in obtaining patients’ data remotely. However, we found that setting up patients’ CIP data accounts and linking them to their providers was somewhat cumbersome and felt that this task was best coordinated in a face-to-face visit.

The authors thank the Leona M. and Harry B. Helmsley Charitable Trust for funding the T1DX-QI. The authors also acknowledge the contributions of people with diabetes, families, diabetes care teams, and collaborators within the T1DX-QI who continually seek to improve care and outcomes for people living with diabetes.

This study was funded by an education grant from the Eli Lilly & Company Global Medical Affairs Division, Indianapolis, IN. The T1DX-QI is funded by the Leona M. and Harry B. Helmsley Charitable Trust.

No potential conflicts of interest relevant to this article were reported.

V.L.H. and N.S. were the clinical providers working with patients with CIPs, provided CIP education to other health care team members, created the supplementary figures, contributed to the discussion, and reviewed and edited the manuscript. H.S. coordinated access for patients to receive CIPs, contributed to the discussion, and reviewed and edited the manuscript. S.A.T. provided guidance for the discussion, gathered study data, and wrote and edited the manuscript. V.L.H. is the guarantor of this work and, as such, had full access to all the data reported and takes responsibility for the integrity of the data and the accuracy of the data analysis.