• Healthcare Huddle
  • Posts
  • Unlocking the Potential of Generative AI in After Visit Patient Summaries

Unlocking the Potential of Generative AI in After Visit Patient Summaries

Huddle+ | Inefficiency Insights #36

What’s the Problem?

The after-visit summary often fails to effectively communicate critical medical information to patients due to information overload, complex terminology, and time constraints on physicians, leading to confusion and decreased patient engagement in their own care plans.

Problem Description

The after-visit summary is a nuisance and challenge for both the patient and the physician. I'll share my perspective from both perspectives.

Patient Perspective

As the patient, I’m sitting there listening to the physician discuss the workup and management for my chief complaint. I’ll keep it relatable: let’s say it’s abdominal pain.

The physician says, “We’ll do this and that. If this is the case, then we’ll start that medication. But if that’s the case, then we’ll start this medication.”

I sit there nodding in agreement. I’m not writing anything down, just listening—but there’s no way I remember any of this. Plus, the medications sound like a toddler made them up.

Physician Perspective

Unfortunately, as a physician, I find myself doing the same thing as my physician example above (which is why these Inefficiency Insights are a great time to reflect on my own idiosyncrasies in practice!). I ask the patient questions, discuss the next steps with them, and then send them on their way.

At the end of each visit, an “After Visit Summary” is printed out. It’s very templated. It includes:

  • Instructions

  • Vitals from the visit and issues discussed

  • Follow-ups

  • Changes to the Medication List

These are all automated. For instructions, it will pull in new medications and tell you, the patient, to pick them up. Vitals are uploaded from the visit. Issues discussed are added based on the diagnostic codes for the visit. It goes on and on. The physician can take time to type in instructions to summarize everything that was discussed during the visit. But when the physician only has so little time, this seldom gets done.

The Challenges

  1. Information Overload: Patients are often overwhelmed by the amount of information given during a visit. Without a clear, concise summary, important details are easily forgotten.

  2. Complex Medical Terminology: The medical jargon used can be confusing for patients, making it difficult to understand their own health care plan.

  3. Time Constraints: Physicians are pressed for time. Crafting a personalized after-visit summary is often deprioritized in favor of other pressing tasks.

  4. Automation Limitations: While the automated aspects of the after-visit summary (e.g., vitals, follow-ups) are helpful, they lack the nuanced, personalized instructions that patients need.

  5. Patient Engagement: Without a comprehensive and understandable summary, patients are less likely to engage with their care plan, potentially leading to non-compliance and poorer health outcomes.

If done right, the after-visit summary can be invaluable for patients. It’s a vital communication tool that reinforces the patient’s understanding of their health care plan. The challenge lies in getting it done right, ensuring it is comprehensive and comprehensible within the time constraints of a typical physician’s schedule.

Insights

Root Cause Analysis: 5 Whys

The 5 Whys process in root cause analysis involves repeatedly asking "Why?" five times to drill down into the root cause of a problem by exploring the cause-and-effect relationships underlying the issue.

The problem: The after-visit summary often fails to effectively communicate critical medical information to patients.

  1. Why?: the information is often overwhelming and confusing for patients.

  2. Why?: it contains complex medical terminology and lacks a clear, concise summary of the visit.

  3. Why?: the automated templates used are not personalized to the patient’s level of understanding.

  4. Why?: physicians have limited time to customize these summaries during patient visits.

  5. Why (root cause)?: they are burdened with numerous other tasks and time constraints during each patient visit.

Impact Analysis

Impact analysis is the assessment of the potential consequences and effects that changes in one part of a system may have on other parts of the system or the whole.

  • Patient: this often results in confusion and misunderstanding of their care plans, leading to non-compliance with prescribed treatments and poorer health outcomes. They may feel overwhelmed by the complex medical terminology and lack of clear instructions, ultimately eroding their trust and satisfaction with their healthcare experience.

  • Clinician or Provider: the need to explain and re-explain care instructions in subsequent visits or follow-up calls takes valuable time away from other patients, exacerbating their already heavy workloads. This inefficiency contributes to burnout and diminishes the quality of care they can provide.

  • System: poor patient compliance and understanding can lead to increased readmission rates, higher healthcare costs, and more frequent emergency visits. These inefficiencies strain resources and reduce the overall effectiveness of healthcare delivery, highlighting the critical need for improvements in the communication tools used in patient care.

Become a Huddle+ member for full access to my weekly Inefficiency Insights newsletter 👇️ 

Potential Solutions

  • Implementation of Ambient Generative AI: Utilize ambient generative AI to listen to conversations between the doctor and patient, summarizing the discussion and care plan in real time. This summary would be immediately available to the patient before they leave the office, ensuring accurate and personalized instructions that enhance patient understanding and compliance. Or, physicians can (in a HIPAA-compliant way) copy and paste their progress notes into Doximity GPT and basically make a patient-friendly version of it.

  • Patient-Friendly Technology Integration: Develop and integrate patient-friendly digital tools, such as mobile apps or web portals, that allow patients to access and interact with their after-visit summaries. These tools can offer simplified explanations, visual aids, and interactive elements to help patients better understand their care plans and manage their health proactively. These would be an upgrade from the standard After Visit Summary. Right now, we’re able to interact with PDFs with ChatGPT. Perhaps patients can interact with their After Visit Summaries using ChatGPT to answer any questions they may have.

  • Incentivized Wrap-Up Time: Offer additional compensation for physicians who spend an extra three minutes at the end of each visit to review and personalize the after-visit summary with the patient. This incentive encourages physicians to ensure that patients fully understand their care plans, improving patient satisfaction and health outcomes. This is wishful thinking.

POLL

Would you allow generative AI to summarize your visit with the doctor?

Login or Subscribe to participate in polls.

LAST WEEK’S INEFFICIENCY INSIGHTS

You have access to recent Inefficiency Insights here.

HUDDLE #TRENDS

Monthly deep dives into pivotal healthcare trends, crafted as enduring resources for research.

Check out the most recent Huddle #Trends report below 👇️ 

You have access to recent Huddle #Trends reports here.

P.S. Need Huddle+ for your team? Send me an email ([email protected]) for more info about Huddle+ corporate subscriptions and discounts.

Please be advised that the content expressed within 'Inefficiency Insights', including all views and opinions, are solely those of the author and do not represent or reflect the views, policies, or practices of the institution with which the author is affiliated. The inefficiencies and related matters discussed in this newsletter are synthesized from a variety of sources, including but not limited to academic journal articles, generalized patient experiences, and the author's cumulative experiences in the healthcare sector over several years, and should not be construed as specific to or indicative of the practices at the author's place of employment.

Join the conversation

or to participate.