Agenda

  1. Welcome
  2. Participating Centers BPA Updates
  3. Preliminary Data/ BPA Insight- JHU Team
  4. Next Steps

JHU Updates

  • Deployment occurred on May 5th, and the data presented covers through August 5th.
  • 39% of the time, an acknowledgement button was selected.
  • 0.8% of the time, someone added the problem to the problem list.
  • 36% of the time, the advisory was dismissed.
  • 19% of the time, it was deferred.
  • 5% of people opened the smart set to add the device to the med list.
  • The most common acknowledgement reason was “patient declined or was not ready.”
  • Followed by “they’re already on an AID system. We’ll update the med list.”

Differences Between Adult and Pediatric Data

  • Pediatrics seems to be engaging more with the OPA than adults.
  • Risa stated that in pediatrics, over 50% of people are acknowledging or addressing the OPA.
  • Cancellation rate is at 30% and some deferral.
  • Risa: “patient declines or is not ready” was 42% of the responses in pediatrics.
  • 20% was the patient is not appropriate.
  • 17-18% was the patient was already on an AID, just wasn’t updated on the med list.
  • 11% were in a pre-contemplative phase.

Adult Data

  • 40% dismissed it, 30% had some acknowledgment, 22% deferred.
  • Very few are opening the smart set or adding the problem to the problem list.
  • The biggest acknowledgement reason was patient decline, followed by already on AID (will update med list).

Free Text Comments

  • Common themes on the peds side: Newly diagnosed, in a honeymoon phase, on an NDI, doing well, don’t need a pump yet, Patient was incarcerated and couldn’t get access, Subpar patient compliance, Not yet on CGMs, newly diagnosed, not quite ready, needs more education.
  • Other comments: Patient considering, not interested, discussed without detail, Safety concerns (vision/cognitive impairment), Some people wrote insurance when they probably meant cost.

Provider Feedback (Adult Side)

  • Alert fatigue is a big potential source.
  • Patients don’t have type 1 diabetes.
  • Patients are already on an AID.
  • Problem list reconciliation is a huge problem.
  • A provider reported the OPA fired inappropriately.
  • Training MAs to add devices to the med list at check-in since May.
  • Biggest feedback: right time in the workflow.
  • The main issues revolve around timing, workflow, and ensuring the right person (someone with type 1 diabetes not on an aid) is involved.

Participating Center Updates

Nationwide Children’s

  • The QI representative for ENDO will handle the data portion, while Leah focuses on slicer dicer sessions.
  • Leanne removed the cancel option to focus on actionable acknowledgement reasons.
  • Initial data showed many alerts per patient (8-10), which was scaled down.

Acknowledgement Reasons and Comments

  • There was a wide variety of reasons for patient decline, such as parents not feeling the patient is ready, patients playing football, or families traveling.
  • Some initial feedback indicates that the acknowledgement reason buckets don’t necessarily fit the institution’s specific workflow.
  • Patients attend a pump class with a diabetes nurse educator as a next step in the process.
  • There’s some misuse of acknowledgement reasons, with practitioners sometimes selecting “patient declined” for practitioner-related reasons.
  • The organization is using SNOMED concepts to capture diagnoses.

Note Templates and Smart Links

  • The organization uses Partners for Kids (PFK) to incorporate care coordination into patient-specific care, especially for long-term diagnoses.
  • A smart link is embedded into note templates to capture data and provide more specificity into the acknowledgement reason.
  • This smart link pulls information into the assessment and plan, offering more detailed descriptions that overlap with current acknowledgement reasons.
  • Smart data elements are associated with each acknowledgement reason to capture information, and an extension is used to pull in the comment.

Alert Configuration and Feedback

  • The appointment trigger was removed from the note template to prevent the alert from firing during pre-charting, based on feedback.
  • The alert now only fires for office visit encounters, not video visits.
  • Instead of a 90-day lockout, a 72-day lockout was implemented to ensure the alert would fire at the next encounter if acknowledged in a way that it should.
  • There is a consideration to suppress the alert in pre-charting, but input from other groups is desired.
  • It may be necessary to reassess the decision in 8-12 months to see if patients want the alert removed entirely.

Grady Update

  • Grady went live the first week of August.
  • The build looks like Johns Hopkins’ build, except Grady doesn’t have the order set.
  • Grady has the option to add the problem list.
  • All details match what JHU is doing.
  • Grady is a safety net hospital with historically high no-show rates.
  • The hospital is trying to improve continuity of providers.
  • There is a dedicated clinic for Type 1 diabetes in adults that is expanding.

 

Feedback on the New System

  • Initial feedback from providers was that it was “too soon” and they needed more time to go through the chart.
  • There was a need for better education on what the buttons meant for accept, dismiss, and defer.
  • A faculty member suggested education around the “defer” button, specifically if it could pop back up after a certain time.
  • The IT team said there aren’t many trigger options outside of “open encounter.”
  • A suggestion was made for the storyboard to have the deferred item pop back up between 5 and 15 minutes later.

Discussion on “Defer” Functionality

  • Some groups want the deferred item to pop back up because they are forgetting about it.
  • One suggestion was for it to pop back up 15 minutes later or right before closing the chart.
  • The current time frame for defer is set at 60 minutes.
  • There is no “close encounter” trigger yet, but there is talk of it.
  • A “sign note” or “mark note” trigger was suggested but not accepted.

Technical Aspects and Alternative Solutions

  • The minimum time for an item to pop back up is unknown.
  • One group turned off pre-charging due to negative feedback.
  • The OPA trigger is “open chart,” which shows up when the MA rooms the patient, but the provider hasn’t spoken to the patient yet.
  • The storyboard and practice advisories tab are used to remind providers.
  • Other OPAs have alerts that come up and won’t let you sign the encounter if certain requirements aren’t met.
  • OPA customizations are limited in flexibility compared to others.
  • There are no override reasons.
  • Links and graphics cannot be added.
  • Currently, a user action is required to trigger the alert again.

Validation and Re-triggering

  • The speaker asks about a validation process where placing a diagnosis code re-triggers the alert if it was snoozed or dismissed.
  • 30 minutes is a reasonable amount of time to lock out patient charts because patients are seen by multiple people (nutritionists, diabetes nurse educators, practitioners).
  • The goal is to capture data on patients using technology (pumps) that hasn’t been documented.

Aid Use Reporting: T1DX-QI

  • Dhruvi asks about a consensus on a year to use as a cutoff date for when pump models became AID compatible.
  • The question is: If there isn’t discrete data to confirm AID use, can an assumption be made that if a patient has an insulin pump or CGM, they are likely using an AID?
  • Omnipod 5 was commercially released in May 2022, and Control IQ has been available since 2018.
  • Issue: Centers report general pump use but not specific pump variables or AID use.
  • Dr. Mathioudakis asks what percentage of patients with type 1 diabetes using pumps are in manual mode.
  • Concern: If the logic for defining an AID changes, it will affect the baseline for the project.
  • Suggestion: Combine pump and AID metrics in the T1D exchange portal.
  • Consensus: 2022 is a fair cutoff year.

Ut Southwestern Updates

  • Deployed the intervention on 7/11/2025.
  • The AID fires in pre-charting and office visits.
  • Enhancement: A smart link to update AID information.
  • Educational efforts included division meetings and emails with educational material.
  • Targeted individuals who missed meetings or had low engagement.
  • Patient decline, not ready: More education is needed.
  • Several people got started on pump again.
  • Process: No button to start pump.

Alerts Engagement

  • 181 BPAs fired for these 42 encounters.
  • 71% engagement.
  • One month data: 246 completed encounters with 187 acknowledgments.
  • No shows were being pulled up during the pre-charting and were being counted.
  • Removing no shows resulted in about 76% engagement this past month.
  • Some fellow visits were being counted twice.
  • Some people say discussing pumps is annoying.
  • “Once deferred, same thing, I forget about it. Can it pop up again after a certain time, 15 minutes or at encounter close?”
  • There should be a button to sign up for pumps as a part of the workflow.

SUNY Upstate Update

  • Went live yesterday.
  • Training and overview meeting with providers on July 3rd.
  • BPA includes dismissed reasons, opening the order set, and adding problem.
  • Correct criteria in building the BPA:
  • Patient is type 1 diabetes.
  • No AID system listed on their medication list.
  • No diagnosis related to an insulin pump.
  • Patient must be older than 18 years.

BPA Details

  • An order set lists pumps and CGMs used, with a referral to diabetes education for pump classes.
  • Dismissed reasons have been implemented, with an option to add more detail or a comment.
  • A BPA locks out a dismissed reason for 90 days, retriggering if the patient isn’t on an AID system after that period.
  • A defer button allows pre-charting providers to postpone the BPA, which will retrigger when the encounter is reopened.

Potential Modifications

  • Turning off pre-charting.
  • Implementing a snooze function with a 30-minute lockout.
  • Adding a hyperlink for documentation within the smart form.
  • Implementing a validation system at the time of entering a visit diagnosis.

Next Steps

  • Next meeting: September 18th.
  • The meeting will include feedback from SUNY upstate, Seattle, and Hassenfeld.
  • Hopkins will share the impact of any modifications made.