Rajiv Shrestha

Allof Health is now private-beta testing!🥳 If I had a dollar for every time someone asked how interoperability could transform health data management, I’d be sitting on a mountain of cash. Three years ago, I got exclusive insights into the potential of decentralized technologies in healthcare through PharmaLedger Project and the European Blockchain Convention. Around the same time, frustrated by the challenges of managing patient data within and across health institutions in Nigeria, I was driven to find a solution that would enable seamless communication between health institutions using patient data and ensure individuals truly own, share, and benefit from their health data. Fast forward to today. After years of dedication—and invaluable support from Co-creation HUB (CcHUB), Futurize and Africa Blockchain Institute —we’re finally cracking this complex puzzle. (P.S.: We’ve had my fair share of challenges along the way, but it’s all worth it as we are now live on TESTNET and set for a private beta test: yes, we're building Allof Health with your needs in mind.) Now, we’re thrilled to open limited slots for forward-thinking health institutions—hospitals, clinics, and community pharmacies—who are ready to lead healthcare’s digital transformation to join our community. Whether you’re looking to boost visibility, create a digital identity, or leverage our health data-sharing platform, this is your opportunity to build the future of healthcare. What’s In It For You? For Patients: • Control, share, and manage health data securely. • Easily book appointments and locate healthcare providers. For Providers: • Access and update patient records without local data storage needs. • Efficiently manage medication inventories and prescriptions. 🔗 Join Our Demo List: https://lnkd.in/dBQqjxE9 We also believe education is the key to driving healthcare digitization forward. That’s why we’re offering workshops on health data management for interested institutions. These sessions cover best practices in data collection, sharing, and security. 🔗 Register Your Institution for a Workshop: https://lnkd.in/dkem6RZ7 Special thanks to my teammates Alinafe Kaliwo Toluwani Michael Alese Lekan Lawal Esther Ogodo Nanna Ponfa who made this possible. Let’s build the future, together. Etheros HealthData Foundation Nigeria Health Watch Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH Africa CDC HealthTech Hub Africa Private Sector Health Alliance of Nigeria The Pan African Health Informatics Association (HELINA) David Chen Rhea Singhla Smart Africa Digital Health Africa Digital Transformations for Health Lab Jude Feranmi Adejuwon InnovateHealth Africa

58

21 Comments

In this week’s The Yuan, I share my thoughts on advances in #QSM (Quantified Self Movement) and how newer technologies (like #wearables and #LLMs) can unlock the power held in #biomarkers to help us live healthier for longer. Medicine 3.0 is here! #mitoplus #mito #humanedge #docm #longevity #healthspan #healthcare #everdayhealth

6

This graphic is a great visual by #RezaZahiri, PhD! It illustrates how strong the #MedTech industry is financially, a proxy to the essential connection with #innovation in #healthcare. Most if not all of these companies have made substantial commitments to decrease #GHG emission’s and achieve more sustainable products. They are exploring #circularity by design, extending the life of #MRI, changing packaging materials to decrease waste, enhancing biodegradation and taking on the burden and opportunity of changing #singleuse to #reusable instruments. With healthcare struggling to make ends meet, paying for the disposal of #waste is expensive! Yesterday it was my privilege to talk with over 350 #surgeons at the American College of Surgeons Leadership meeting in DC about #sustainability. They were surprised to realize that sustainability is closely tied to #equity and #humanslavery. This week surgeons, especially the next generation of leaders, will “see” the opportunity for change more clearly. The evidence is there in every OR. One call to #action is for hospital leaders to communicate to their buying groups that sustainability is an important component to consider when deciding which companies to purchase from. #partnership #acs #interventionalists

15

Does the heterogeneity of early phase oncology studies preclude it from the need for patient reported outcomes (PROs)? The Oncology Dosing Toolkit by the FDA (as part of project OPTIMUS) does not seem to think so! But wat does "As Applicable" mean exactly?

10

2 Comments

**Top 5 Myths About Ethical AI Use in Healthcare** 📊 1. Data governance is all that you need to ensure unbiased AI: ⚠ This oversimplifies the complex challenge of bias in AI systems. While data governance is crucial, addressing AI bias requires a multifaceted approach that includes continuous monitoring, sociotechnical assessment, and clinical feedback loops. 🎭 2. Once deployed, AI systems will maintain consistent performance: ⚠ This myth is debunked by many model performance evaluation needs, including the concept of "model collapse," defined as "the degradation of AI performance over time, potentially leading to inaccurate diagnostic tools." 👨‍👨‍👧‍👦 3. Including patients in prioritizing AI use cases is too difficult: ⚠ This myth underestimates the value and feasibility of patient involvement. Responsible AI governance should include patient perspectives. Engaging patients provides crucial insights into real-world needs, potential risks, and ethical concerns, leading to more patient-centered and effective AI applications in healthcare. ⚕ 4. AI-driven healthcare is more efficient and accurate: ⚠ To date, LLM and MLLM research has identified risks like "poor fit for clinical use" and "misaligned diagnostic outputs," indicating that AI systems may not always perform optimally in real healthcare settings. While there are existing GenAI applications already in use, more research is needed covering a range of health conditions, specializations, and use cases for wide-scale adoption to be successful. 🤔 5. AI can replace human decision-making in healthcare: ⚠ As stated above, there are many remaining challenges to adopting AI in healthcare settings. These include warnings about "excessive automation" in patient care and "over-reliance" on AI in critical care decision-making, without maintaining significant human-in-the-loop oversight and approvals. 🚨 These myths underscore the importance of a GOVERNANCE-FIRST approach to AI in Healthcare! By understanding and addressing these misconceptions, healthcare leaders can work towards more ethical, effective, and patient-centered AI implementation. 💡 ⚡ #AIinHealthcare #EthicalAI #HealthTech #PatientCentric #GovernanceFirstAI #SociotechnicalAI #ResponsibleAI #AIMyths

10

7 Comments

As generative AI becomes an increasingly integral part of healthcare, the FDA's Digital Health Advisory Committee has been emphasizing the need for a regulatory framework that ensures both innovation and patient safety. A centralized data repository to monitor real-world performance and track risks could make postmarket evaluations scalable and efficient. Notably, the panel has stressed the importance of guarding against biases inherent in AI by proposing certification programs to enforce accountability. By emphasizing transparency and rigor, these efforts aim to bridge the gap between cutting-edge AI applications and equitable, effective care. But one major question still remains: How can we ensure AI consistently elevates healthcare closer to the gold standard without compromising trust? #AI #GenerativeAI #DigitalHealth #PatientSafety

2

Words of wisdom from Joshua Tamayo-Sarver, MD, PhD, FACEP, FAMIA regarding the "Perilous Pilot." #Healthcare and #LifeSciences #startups, heed his helpful guidance on #governance when building your #AI solution. Microsoft for Startups #startupsuccess #innovation #AIinHealthcare #HealthcareOnLinkedIn

7

1 Comment

Stanford just published a paper: Large Language Models in Healthcare: Are We There Yet? 👇👇👇 My summary: Actionable Insights for Healthcare Executives on Large Language Models 1. Prioritize Real-World Data Testing: Current Gap: Most evaluations of LLMs use curated datasets, not real patient care data. Action: Implement rigorous testing of LLMs using real-world patient data to ensure reliability and safety in clinical applications. 2. Expand Evaluation Scope: Current Focus: Majority of studies focus on medical knowledge and diagnostics. Action: Broaden LLM evaluation to include administrative and operational tasks, such as billing and clinical note writing, to maximize efficiency and reduce physician burnout. 3. Adopt Diverse Evaluation Metrics: Current Issue: Evaluations often prioritize accuracy but neglect other critical dimensions. Action: Implement a holistic evaluation approach that includes fairness, bias, robustness, and deployment efficiency to ensure comprehensive performance assessment. 4. Implement Scalable Evaluation Frameworks: Current Challenge: Manual evaluations are time-consuming and limit progress. Action: Utilize AI agents and automated evaluation mechanisms to scale the assessment process, ensuring rapid and thorough evaluation of LLMs. By focusing on these areas, healthcare executives can ensure that large language models are safely and effectively integrated into clinical practice, enhancing both physician workload management and patient care outcomes. #AI #GenAI #healthcare

11

1 Comment

This was enlightening and the last bullet caught my attention! Per a study from Tufts Center for the Study of Drug Development it takes 7 years for an innovation to go from from piloting to adoption in drug development and 20 years for the industry to achieve adoption by 2/3 of companies in that industry. ↳ Your product won't scale until you have convinced key opinion leaders (KOLs) and scientific societies with robust scientific evidence. It typically takes up to 3 years from the first scientific paper until it really picks up (= there is a wide adoption). What evidence do you need for trials with decentralised elements. This sponsor experience report on implementing decentralized elements from TransCelerate BioPharma Inc. gives a good overview of where we are https://lnkd.in/gENjpneb

44

13 Comments

Attention All ErgoScience FCE Providers: The New and Improved Web-Based Software Is Now Available From May 27 through June 1, 2024, we’re offering our current FCE Providers an exclusive 15% discount on FCE refresher training (optional) and our new web-based software, PWPE Cloud™. Renew your software license and take advantage of the many benefits including these upgrades: -Enhanced Security: Data security is paramount. Our transition to cloud-based software ensures the protection of your valuable patient information. -Increased Efficiency: Your time is valuable, so we’ve streamlined the FCE process to boost your productivity. -Improved Customization: With PWPE Cloud™ you’ll have even greater flexibility for job-specific tests. -New Report Format: The new software generates a more attention-grabbing report, helping your clinic stand out from the crowd. Schedule a call with us today to discuss how our new software can help you excel even more: https://bit.ly/3UKPcHi

4

Endless possibilities vs definitive end points. Identify the end point before looking at the endless possibilities. Wearing my practicing clinician hat, I feel the above is the biggest problem when it comes to #health #ai solutions. AI and #genai have given rise to endless possibilities (just check out the mindboggling video in Andreas Horn’s post below). When it comes to the application of these technologies in #healthcare however, it is imperative for us to realise that in the Health Industries, every single one of us is working towards a definitive end point. Before embarking on building or buying health technology solutions in this space, time needs to be spent on identifying solutions of true value and impact for there are often less than you might first imagine. Identify the end point before looking at the endless possibilities, not the other way around, at least when it comes to healthcare. HiruX Pte. Ltd. Rafael J. Grossmann, MD, MSHS, FACS Professor Shafi Ahmed

14

5 Comments

AI is only as good as its input and training data, as well as the questions we ask of it. When we build new systems, it’s easy for existing limitations to elicit a suboptimal solution for a diverse population. #aihealthcare #aihealthequity #healthequity #ai #datadisparity #healthdisparities

7

2 Comments

In case you missed it, last week the Coalition for Health AI (CHAI) released a draft consensus framework for responsible health AI. Developed with input from over 100 contributors representing a diverse network of healthcare stakeholders, this guide proposes actionable evaluation criteria throughout the AI lifecycle—from identifying use cases to deployment and monitoring. Key examples outlined in the guide include: • Predictive EHR Risk Use Case (Pediatric Asthma Exacerbation) • Imaging Diagnostic Use Case (Mammography) • Generative AI Use Case (EHR Query and Extraction) • Claims-Based Outpatient Use Case (Care Management) • Clinical Ops & Administration Use Case (Prior Authorization with Medical Coding) • Genomics Use Case (Precision Oncology with Genomic Markers) The draft framework is open for public review and comment for the next sixty days. Please use this form to submit feedback: https://lnkd.in/eYughFYC Kudos to the CHAI team led by CEO & President, Dr. Brian Anderson, MD, on this milestone. Eric Horvitz Jennifer Goldsack John Halamka, M.D., M.S. Michael Pencina Micky Tripathi Nigam Shah Suchi Saria Troy Tazbaz #healthcare #ai #artificialintelligence #generativeai https://lnkd.in/e7VRv6fD

125

1 Comment

Breaking complexity to more manageable steps can allow #AI to be deployed in a controlled, responsible way: “#Systemengineering decomposes complex tasks into modular steps, enabling tasks like #clinicaltrial matching, clinical note generation, and automated coding through interconnected #LLM modules and human-in-the-loop validation.”

6

1 Comment

There are two related studies: (1) Study of LLMs and physician case management decision making: https://lnkd.in/gQHu8Gpm (2) Study LLMs and diagnostic reasoning:  https://lnkd.in/gpBETjqS On the management study (1), performance of the LLM alone was found to be about the same as the performance of the LLM + physicians + conventional resources. Both were better performing than physicians + conventional resources (no LLM access) on the set of medical challenges considered. In the diagnostic study (2), the LLM alone did better than the LLM + physician + conventional. And the LLM + human did about the same as human + conventional resources (without the LLM). That is, the availability of GPT-4 to physicians did not significantly bolster their diagnostic reasoning. The results in both papers indicate that the LLMs (in the context of the scenarios evaluated) are not being harnessed to ideally complement human intellect, per setup, training, context. There's much to do and explore on better training of physicians to use LLMs and, more fundamentally, on methods for advancing human-AI collaboration, in pursuit of dream of leveraging AI to augment human cognition. Stanford Institute for Human-Centered Artificial Intelligence (HAI) Stanford Center for Artificial Intelligence in Medicine and Imaging (AIMI)

116

9 Comments

"Subtle but transformative" so aptly describes Apple's work. My latest for Forbes: how Apple is just getting started in healthcare. I specifically examine its progress across a multitude of services, ranging from the Health App to the #AppleWatch, and now, its ground-breaking work in spatial computing with #VisionPro Last month, Apple released widely celebrated report on how Vision Pro has the power to disrupt healthcare across numerous functions I also discuss two papers that look at the efficacy of the ECG function on the Apple Watch: -The first is the ground-breaking Apple Heart Study which was published by the NEJM Group, and was spearheaded by Marco Perez, Kenneth Mahaffey, John S. Rumsfeld, MD PhD, et al. -The second is a recent paper by Shari Pepplinkhuizen, Willeke van der Stuijt, et al. in the Cardiovascular Digital Health Journal via the Heart Rhythm Society. As always, keen to hear your thoughts & feedback! CC: Lily Peng, Kuber B., Anura Patil, Sumbul A., Kevin Lynch, Eugene Kim, and many more! #apple #AI #genAI #hardware #economy #tech #inflation #healthtech #innovation #access #compute https://lnkd.in/gXeayzvv

138

19 Comments

“These approaches could be the enabler of robust digital twins of individual patients (i.e., representations of up-to-date individual patient data in digital form, serving as a record of patient health and enabling personalized predictive analytics) with LLMs used to rapidly create stable individual patient KGs (knowledge graphs) as stable robust data structures, which could be used to augment and verify data interpreted by LLMs from newly conducted consultations.”

9

1 Comment