Microsoft's Healthcare 'Suite' and Google Tackles Administrative Burnout

The AI Clinic: Weekly Pulse #2 | Subscribe

Welcome to The AI Clinic, your go-to source for exploring the future of healthcare through the lens of AI. We’ve picked ourselves off the floor after another crushing Dallas Cowboys loss and are excited to discuss the week’s healthcare news.

From Microsoft's big AI product announcements to Google's latest research on how AI could relieve healthcare workers, we explore how technology is being applied to solve persistent industry problems. Here’s a quick overview of what’s inside:

In today’s newsletter:

• Microsoft’s New Healthcare AI Tools: Microsoft introduces a suite of AI solutions.

• AI Healthcare Startups on the Rise: Y Combinator’s Winter 2024 cohort reveals a significant focus on AI.

• UpToDate's New AI Medical Scribe: UpToDate gets AI-driven medical notes.

• Google’s Focus on Reducing Administrative Burden: A new study highlights the massive time and resource drain of administrative tasks in healthcare.

Microsoft Unveils New Healthcare AI tools

Microsoft has introduced several new AI tools tailored for the healthcare industry. These offerings include a service to help healthcare organizations build custom AI agents for tasks such as appointment scheduling, clinical trial matching, and patient triage, as well as models for medical imaging and healthcare data analysis.

Key Takeaways

1. AI Agent Service: Microsoft launched a customizable service enabling healthcare providers to create AI agents using templates for automating tasks like scheduling and triage, potentially easing administrative burdens.

2. Medical Imaging Models: The company introduced new models designed to assist with medical imaging tasks, such as analyzing scans and generating reports, which could support radiologists in their workflows.

3. Data Analysis Tools: A healthcare-specific data analysis platform within Microsoft Fabric allows organizations to manage and analyze health data, with a focus on improving reporting and patient outcomes.

4. Nursing Documentation Tool: In collaboration with Epic and several health systems, Microsoft is developing an AI-powered documentation tool for nurses to streamline data entry using ambient voice technology.

Why This Matters

Microsoft is positioning itself as a key player in healthcare, particularly outside of large electronic health record (EHR) providers. Their suite of Microsoft Office products is already HIPAA compliant and they’ve made significant acquisitions in the healthcare space over the past few years.

However, the broader question remains—who are these tools designed for? AI products often face challenges in practical implementation, especially if they aren’t integrated with existing tools. Other than their nursing documentation tool, it's unclear how Microsoft plans to ensure these new offerings seamlessly fit into the everyday workflows of healthcare practices. Full Article

A Quick Aside: The Incoming Class of AI Healthcare Startups

X user @deedydas recently shared an insightful graphic that’s worth a closer look. It visualizes the sectors represented in the Winter 2024 cohort of San Francisco's prominent startup incubator, Y Combinator.

What stands out immediately is that 90% of the 260 companies in this cohort are focused on AI, with approximately 10% directly targeting healthcare use cases. This concentration tells a clear story: the venture capital-backed startup community has embraced AI with evident enthusiasm, particularly in healthcare.

We've already seen multiple high-profile funding announcements for companies eager to capitalize on the opportunity to use AI in healthcare applications. This momentum in funding feels similar to the early days of smartphone-driven innovations—a wave that brought us culture-defining companies like Uber, Tinder, Instagram, Airbnb, and Venmo.

Tech optimists will tell you that this strong a surge of AI-driven startups could have a similarly transformative impact on healthcare. While the pessimists will tell you they’re still waiting for the culture-changing innovations from the crypto wave a few years ago.

UpToDate's New AI Medical Scribe

Abridge, an AI-powered medical note-taking startup, has partnered with Wolters Kluwer Health to integrate the UpToDate clinical decision support tool into its platform. This integration aims to enhance the quality and relevance of AI-generated clinical notes by embedding evidence-based insights directly into the documentation process.

Key Takeaways

• AI-Driven Note-Taking: Abridge uses generative AI to automatically convert patient-clinician conversations into structured clinical notes in real-time.

• Integration with UpToDate: Clinicians can now access evidence-based recommendations from UpToDate directly within the AI-generated notes, improving decision support at the point of care.

• Partnerships and Adoption: Abridge has formed significant partnerships, including with Kaiser Permanente, and is expanding its AI platform to major health systems like UChicago Medicine and Yale New Haven Health System.

Why This Matters

As more digital health products integrate AI functionalities, it’ll be important to distinguish between "bolt-on" AI tools — those that merely assist with surface-level tasks like summarizing or searching — and AI applications that meaningfully enhance clinical workflows. Currently, it’s difficult to know who among this wave of AI announcements is going to be a true difference maker, since so many companies simply seem eager to show participation in the "AI movement" (similar to companies slapping ‘blockchain’ on everything in 2019).

And while many companies will tout their embrace of AI integration, only those offering solutions that tangibly improve bedside care, beyond just clerical tasks, will have a lasting impact on healthcare delivery. Full Article

Google's Focus on Decreasing Healthcare Administrative Burden

A new study by Google Cloud and The Harris Poll, released this week, sheds light on on of healthcare's most common complaints - the administrative burden faced by U.S. healthcare workers. The study explores the opportunities for generative AI to alleviate these challenges. The report reveals that healthcare professionals are spending significant hours on paperwork, leading to reduced patient care, burnout, and increased human error. Both healthcare providers and payors express optimism about the potential of generative AI to streamline these tasks.

Key Takeaways

1. Administrative Burden is Significant: Healthcare workers spend an average of 28-36 hours per week on administrative tasks, such as documentation and pre-authorization, which impacts their ability to focus on patient care.

2. Widespread Burnout and Staffing Issues: Excessive administrative work is contributing to burnout and staffing shortages, with 82% of clinicians reporting that this workload reduces the time they can spend with patients.

3. Generative AI’s Promise: Both providers (89%) and payors (98%) are open to using generative AI to reduce administrative burdens. Key tasks where AI could assist include documentation, billing, communication, and claims management.

4. Public Acceptance: The general public is cautiously optimistic about AI in healthcare, especially when it directly improves patient care. 85% of respondents believe that using AI for administrative tasks is a good idea if it helps clinicians spend more time with patients.

Why This Matters

Google's study underscores a well-known frustration in healthcare: nobody is satisfied with the status quo. Patients, healthcare workers, and clinicians alike struggle with a system that hinders efficiency but offers few tools for meaningful change. The highly regulated nature of healthcare makes rapid innovation difficult—and a "move fast and break things" attitude can lead to serious harms.

Despite the very real ethical and safety concerns surrounding AI, this study highlights the urgent need for better solutions and the strong interest in what AI can offer. Generative AI presents a potential path forward to reduce the administrative overload that is draining resources and negatively impacting patient care. But it remains to be seen if any of the current suite of tools being brought to market will make a meaningful impact on this problem. Read The Study

Other things worth checking out

Deep learning AI model scans 'dark matter' of genomic data to find 70,000 never-before-seen RNA viruses

Study shows doctors and nurses disagree on value of AI generated portal messages

6 children's hospitals leading in AI