Serkan Kutan

The defining factor isn't the size of the company or startup, but the vision behind it. Take OpenAI, for instance. Until last year, unless you were deeply embedded in the ML world as a computer scientist/researcher, or enthusiast, you might not have heard of OpenAI. Meanwhile, a giant like Google was dominating the scene. However, a few years back, OpenAI was just another small startup. Now, it's rapidly evolving into a major player in big tech. Why and how? It comes down to having a long-term vision, a clear roadmap, patience, and solid execution (i.e., talents). Even in the context of healthcare AI, where (small) startups face many challenges (including the ones mentioned), the principles of vision and strategy apply. While merging with big tech might seem like an option for some, if a startup truly believes in its vision and has a well-defined path to achieve it, it should forge ahead independently. The landscape is evolving, and like OpenAI, today’s small startups with clear, visionary goals could be tomorrow's industry leaders. #starup #founders #openai #google #healthcare

3

There is a misconception emerging that AI agents will lead to immediate reductions in healthcare labor costs. While AI agents can automate end-to-end workflows such as booking appointments, answering basic patient questions, and performing form completion for prior authorizations, automating a single task is not equivalent to automating an entire job. Nurses are a prime example. Certain nursing tasks, such as care coordination and patient documentation, are highly automatable by AI agents; however, there is a tremendous amount of work nurses perform on the ground that cannot be fully automated with current technologies such as direct patient care, physical examinations, medication administration, wound care, patient support, and clinical assessments. These nuances make full role elimination less likely with current technologies. When articulating the ROI of an AI agent, we need to be both precise and accurate. Automating a task is more often not automating an entire profession. For roles that encompass many functions, such as nursing, AI agents can be invaluable tools for unburdening staff, increasing efficiency, and boosting throughput—benefits that are particularly valuable given current healthcare staffing and resource shortages. The most promising job candidates for full role elimination through AI remain positions with monolithic task structures, of which there are relatively few in healthcare today—scribes, medical coders, and data entry specialists for example. It is therefore unsurprising then that the most progress has been made in these categories.

124

18 Comments

One of the most underrated aspects of using an #interoperability tool like Metriport is not needing to know where the patient has received care previously 👇 While this may be obvious to providers who have leveraged Health Information Exchanges (HIEs) before, the reality is that so many providers today still need to ask their patient where they’ve received treatment previously (assuming the patient remembers them all), and then needs to call each facility and beg them to send over the records via #fax. This process can take days, if not weeks, and often still results in the provider having an incomplete, partial understanding of their patient's medical history 🤷‍♂️ Through #HIEs, a provider who leverages Metriport simply needs to enter the basic demographic info of their patient (name, DOB, gender, and address) to pull records from all the facilities where their patient has received care previously. Metriport handles this entire process in seconds, without any humans in the loop. This is by far one of the biggest “aha” moments for providers who use Metriport for the first time, after being reliant on manual retrieval methods previously. From initial skepticism, to providers who now can’t live without this, we see these tools as an inevitable part of evolution of modern #healthcare workflows.

24

2 Comments

Rock Health's latest report reveals a promising rebound for U.S. digital health startups, with $5.7B raised in H1 2024, potentially matching or exceeding pre-pandemic funding levels. The HLTH discussion highlighted the focus on practical AI applications like billing and coding, while noting that startups face competition from established players. As the market remains uncertain, strategic M&A is emerging as the most viable exit strategy. If you’re a small company or startup looking for guidance in the industry, feel free to reach out or connect to exchange ideas. https://lnkd.in/e_727ajd

9

1 Comment

One key reason why healthcare technologies like AI, telemedicine, and software haven't structurally lowered costs or broadened access for the entire system after decades of experimentation (excluding pure financial engineering of payer contracts) is the limited reimagining of care delivery itself, with asynchronous telemedicine as a notable exception. In the same way that digitizing physical paper failed to exploit the capabilities of a new computing medium for word processing, hosting physician-patient conversations over video does not transform the actual delivery of care. In telemedicine, many of the inputs (e.g., clinician time) and therefore cost structures of in-person care persist, which is why we often see fee-for-service telemedicine companies struggle to achieve economies of scale (i.e., growing revenues non-linearly while growing costs linearly). Until we deconstruct the components of a medical appointment – from intake and triage, to chart review, history gathering, physical examination, diagnosis, treatment, and follow-up – we will continue to be underwhelmed by the disconnect between the transformative potential of these technologies and real-world patient/system impact. I’m optimistic that the emergence of reliable and performant medical AI agents will urge us to perform this deconstruction and redesign the fundamental building blocks of a “medical appointment.”

215

39 Comments

Article in today’s WSJ highlights the skyrocketing cost of healthcare bills for payors and patients. 4L FWA Prevention powered by Integr8 AI can detect these schemes to incrementally ratchet up claims cost in mear real-time. Stop abusive billing schemes when thet start. Find, Fight and Prevent FWA Fast!

17

Exciting to see CodaMetrix named as one of Boston's top startups to watch in 2024! The administrative burden on physicians has grown tremendously, leaving less time for providers to dedicate to patient care. CodaMetrix helps solve that problem with their AI-powered solutions, allowing the providers to focus on their patients and giving the coders technology that helps them focus their expertise on complicated cases. #medicalcoding #AI #healthcare #transforminghealthcare

20

3 Comments

Amandeep Singh, Ph.D., Hell no! Well, all of us do hear startups trying to 'disrupt', a good innovation is where the target customers come wanting to have the solution in their lives and not excusing themselves by saying that they are 'burnt out'. A genuine solutions with a good GTM will ease the customer's life inavriably. What most of the healthcare solutions suffer from is an premature urge to take an 'bare' engineering solution to the market - a 'bare' solution bereft of product and design inputs. System design is not adding more 'seamless' patches to the existing workflows. It is about overhauling them. If it the overhaul is being done after there has been a significant change in the technology, patients expectations, newer diseases, the overhaul becomes all the more challenging. That said, such an overhaul, and the consequent design might be so daunting that no healthcare system would ever want to implement it. That's where you would need help of product managers. The multiplicity of stakeholders (typically the trio of user-operator-payer) is quite prevalent in the enterprise software world. Not a new challenge.

3

Just read Davis Liang's interview on Out-Of-Pocket and it got me fired up. We need more innovators tackling our outdated healthcare system. A few thoughts: First, a shoutout to Abridge for focusing on the nurse experience. Nurses are the backbone of our healthcare system, yet too often they get overlooked. Happy to see this changing. At Tensor, we're always asking: "How can we make healthcare workers' lives easier?" A study found that healthcare workers are experiencing stress and burnout at rates as high as 70%, which negatively impacts patient care. [1] It's really not just about tech - it's about supporting the humans at the heart of care. Davis mentioned the challenge of multilingual capabilities. Yup, I feel that. It's a huge issue, especially among diverse Medicaid populations. In fact, a systematic review of 14 studies involving over 300,000 participants found that language barriers in healthcare lead to a host of problems: · Miscommunication between medical professionals and patients · Reduced satisfaction for both healthcare providers and patients · Decreased quality of healthcare delivery · Compromised patient safety [2] Solving this could be a game-changer for health equity. Another critical aspect addressed was model drift - their approach to this challenge is spot on. For those unfamiliar, model drift is like your AI getting rusty. It's when your once-shiny ML model starts spitting out garbage because the world changed, and it didn't keep up. (This could mean the difference between life and death in healthcare.) As model drift shows, you can't just "set it and forget it" with AI in healthcare. Continuous improvement is key. Overall, I’m intrigued by their work on inpatient use cases. There's so much untapped potential there. It's inspiring to see companies like Abridge tackling these complex problems. While we might approach some things differently at Tensor, we're all working towards the same goal: better, more efficient healthcare for all. To me, the future of healthcare (with AI) will lead to: 1. Personalized care for all 2. Data-driven, increasingly rapid advances in standard of care 3. Prevention, prevention, prevention! 4. Better adherence to known screens to detect diseases before they become unmanageable It’s an exciting time! Hey Davis - if you're reading this, let's grab a coffee sometime. I'd love to geek out about healthcare AI with you!

11

3 Comments

In a recent report, only ~20% of healthcare executives claimed that they fully trust their data. At Fullsteam Health, we think that % is way too low. We're committed to reversing this trend with our automated data curation platform which was initially developed at Duke to address this very issue and allow for fast access to data that is needed and that has been cleaned, grouped, and continually monitored. https://lnkd.in/epWvPTDV

21

6 Comments

After a brief writing hiatus… Part III of “Who Pays for Healthcare AI” is live. The third installment defines five performance-based pricing models for healthcare AI and explores: • Four types of Return on Investment (ROI) in healthcare and life sciences • The importance of shielding at-risk pricing models from operational externalities • How better pricing model design can mitigate negative working capital risk under performance-based models • Taming the beast: attribution Check out Part III below, and subscribe to receive the final post of the series! https://lnkd.in/dPZSHMiy

122

7 Comments

Beebe Healthcare to Deploy Epic Enterprise EHR Across Network - HIT Consultant Beebe Healthcare Implements Epic EHR System #BeebeHealthcare #EpicEHR Beebe Healthcare, a leading healthcare provider, has announced the deployment of the Epic Enterprise EHR system across its network. This strategic decision aims to streamline operations, improve patient care, and enhance overall efficiency within the organization. #EHRdeployment Benefits of Epic EHR Implementation #EpicBenefits The implementation of the Epic EHR system will bring numerous benefits to Beebe Healthcare. These include improved data accuracy, enhanced communication among healthcare providers, increased patient engagement, and better clinical decision-making. #HealthITbenefits Enhancing Patient Care with Epic EHR #PatientCare By leveraging the capabilities of the Epic EHR system, ai.mediformatica.com #health #healthcare #epic #patients #this #suex #healthcaresystem #platform #strategic #carecoordination #careexperience #efficiency #digitalhealth #healthit #healthtech #healthcaretechnology @MediFormatica (https://buff.ly/4cJtDyR)

2

As the discourse shifts from models to compound AI systems / agents, we need better AI benchmarks to evaluate multi-modal and multi-step task performance, especially in healthcare and life sciences. When we wrote the first paper demonstrating ChatGPT's performance on the USMLE, we chose the US Medical Licensing Exam as a benchmark for accessibility, speed, and ease. This benchmark was never intended to represent AI model performance on real-world clinical tasks. Today, I still see so many research teams and startups using benchmarks (like the USMLE) that are ill-fitted for assessing the true clinical or scientific performance and utility of the models they are developing for real-world contexts. Benchmark development may be seen as a "less sexy" area of research, but it is of paramount importance. Years after the rise of the transformer, we still lack adequate benchmarks for so many single-step tasks in biomedicine. With compound AI systems (i.e., architectures that integrate multiple AI models to perform complex tasks) emerging, we need new benchmarks for agentic behaviors. I'd argue that developing an agent with novel capabilities without at least proposing a companion benchmark (if an industry standard does not yet exist) may hinder the adoption of said agent, especially for high-stakes workflows. Designing more benchmarks that capture/simulate real-world clinical and scientific workflows will help us mitigate the major discrepancies observed between in silico and in vivo performance and better support safe + effective deployment of AI in biomedicine. There are already brilliant people focused here, and we need more. DMs are open if you're researching or working in this area of multi-step/multi-modal benchmarking in healthcare and life sciences! #healthcare #ai #artificialintelligence #generativeai

100

18 Comments

Data governance is crucial to the implementation of AI but often doesn’t get the focus it deserves. Our most recent episode of Diagnosing Healthtech rectifies this. Listen in as Michael Sanky from Databricks joins Jeff Englander and Toryn Slater for a stimulating—and much needed—conversation. https://bit.ly/3UBYDZC #AI #governance #healthcare Healthcare Strategy Bullpen

1

CVS Health had its quarterly earning call yesterday. The stock was down over 16% after the call. ProntoNLP's #LLM based signal predicted this massive downturn in the stock as can be seen in the Forecast event in the below image. The Forecast was going down consistently over the last 5 quarters. In almost all of the #LLM topics CVS is at the bottom of the graph compared to the last 5 quarters. Utilizing ProntoNLP's signals and its #GenAI Pronto research could have prevented huge losses for portfolio managers (and generate big gain for those that were short on it). Contact ProntoNLP to start using our #quant signals and the Pronto Research platform for fundamental analysts. #stockpicking #portfoliomanagement #quantiativefinance

9

This analysis explores how Point of Care Testing (POCT) digital health technologies impact emergency care. It highlights how AI-powered remote monitoring can significantly improve operational efficiency in healthcare and enhance patient outcomes. Read more: [Nature](https://lnkd.in/gYb4hjXX)

1

I appreciate this poignant post from Marty Cagan on "transformation regrets," and nodding my head at all of these. I also want to give a 💯 to the bit on what happens when orgs looking to shift to Product Model put the majority of their investment primarily in boosting up Product Management as a function. I've personally seen this lead to a *decrease* in product team collaboration, which is critical for delivering higher value, higher quality, more competitive solutions, faster & more frequently. (This is a combo of the needs I've seen cause orgs to consider Product Model in the first place.) I'd love to see more transforming orgs prioritize the *system* of value delivery, and how the actors in that system interact efficiently, over the strength of any one actor. *This* is how you increase, for example, the % of OKR impacting solutions shipped, MoM. (Which is a good thing to optimize for, IMO :)

15

1 Comment

How to Implement Zero Trust Without Disrupting Clinical Workflows Continuous validation is central to the zero-trust framework, but related security controls shouldn’t burden clinicians. https://lnkd.in/g3YDa4Q8 #HealthcareIT

1

Big healthcare AI news yesterday as Tempus AI received the first-ever CPT code for an algorithm-only analysis of previously-sequenced, laboratory developed test transcriptomic data. TLDR: AI molecular phenotyping has a new business model. The PurISTSM test uses RNA profiling to identify molecular subtypes in patients with advanced pancreatic ductal adenocarcinoma, which may inform first-line therapy selection and management. While I'm skeptical that payer balance sheets are going to solve the "who pays for healthcare AI" problem, the reimbursement pipes for clinical AI are maturing. I am surprised to see more progress for usage-based vs. performance-based models given the many challenges with FFS economics. It will be interesting to see how payers assess the "medical necessity" and therefore reimbursability of clinical AI as these models come to market. https://t.co/eLe18SmihV

420

8 Comments