Ewout Kramer

This study examines the impact of AI-based Clinical Decision Support Systems (CDSS) on patient-physician trust, focusing on competence, integrity, and benevolence in high-risk and low-risk clinical scenarios. 1️⃣ Cross-sectional vignette study with 398 patients from the University Medical Center Utrecht. 2️⃣ Patients were divided into high-risk and low-risk scenarios, with control and intervention groups for each. 3️⃣ Trust was measured using a questionnaire based on competence, integrity, and benevolence on a 7-point Likert scale. 4️⃣ High-risk scenario: Slight decrease in perceived competence and integrity when CDSS was used, particularly among women. 5️⃣ Low-risk scenario: Trust remained high overall, with increased trust in technology positively influencing perceptions of benevolence and integrity. 6️⃣ While patient-physician trust was generally high, the use of AI applications showed a potential negative impact, especially among women in high-risk situations. Trust in technology plays a crucial role in the acceptance of AI-supported medical decisions. 7️⃣ In low-risk scenarios, patients with higher trust in technology perceived greater physician benevolence and integrity when CDSS was used. 8️⃣ Effective communication about AI's role and patient involvement in the development and use of CDSS are crucial for maintaining trust. ✍🏻 Anna Zondag, Raoul Rozestraten, Stephan Grimmelikhuijsen, Karin Jongsma, Wouter van Solinge, Michiel Bots, Robin Vernooij, Saskia Haitjema. The Effect of Artificial Intelligence on Patient-Physician Trust: Cross-Sectional Vignette Study. J Med Internet Res 2024;26:e50853. DOI: 10.2196/50853 📄 Planning to submit your research to a peer-reviewed, open-access journal? Use the discount code AMJB200 at checkout to save $200 USD on the article processing charge at any JMIR Publications journal. Start your submission here: https://lnkd.in/deuAEN2r. It's important to me that you know this isn't a sponsored post. I'm simply committed to keeping you informed about opportunities to disseminate your research effectively.

134

7 commentaren

𝐑𝐞𝐬𝐩𝐨𝐧𝐬𝐢𝐛𝐥𝐞 𝐀𝐈 𝐢𝐧 𝐇𝐞𝐚𝐥𝐭𝐡𝐜𝐚𝐫𝐞⚕️ #𝐀𝐫𝐭𝐢𝐟𝐢𝐜𝐢𝐚𝐥 𝐈𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐜𝐞 (AI) has the potential to revolutionize healthcare. It can help doctors in many ways like diagnosing diseases early, personalizing treatments, and improving patient care. But there are also risks. Biased algorithms and privacy issues can cause harm. #𝐓𝐡𝐫𝐞𝐞 𝐊𝐞𝐲 𝐏𝐫𝐢𝐧𝐜𝐢𝐩𝐥𝐞𝐬 👉 𝐁𝐞 𝐀𝐜𝐜𝐨𝐮𝐧𝐭𝐚𝐛𝐥𝐞: Imagine AI as a team member. We need to be responsible for our actions. Clear rules and decision-making processes are essential. When AI suggests a diagnosis or treatment, we should know why it made that choice. 👉𝐓𝐡𝐢𝐧𝐤 𝐑𝐞𝐬𝐩𝐨𝐧𝐬𝐢𝐛𝐥𝐲: AI is more than just coding; it means knowing what's going on. We must consider biases in data and how our algorithms affect people. Is everyone getting equal access to AI benefits? 👉𝐁𝐞 𝐓𝐫𝐚𝐧𝐬𝐩𝐚𝐫𝐞𝐧𝐭: Imagine AI as a helpful friend. We want it to explain its decisions. Transparent AI models show us why they recommend certain things. This builds real trust. #𝐑𝐞𝐬𝐩𝐨𝐧𝐬𝐢𝐛𝐥𝐞 𝐀𝐈 aims for fairness. It shouldn’t favor one group over another. By sharing knowledge, we create a world where AI helps without causing harm. Let’s build a healthier future with AI we can trust! #ResponsibleAI #AI #AIInHealthcare

18

2 commentaren

Speaking on panel at HLTH Europe in #Amsterdam. Topic: Cross border use of #HealthData. There’s a whole new world opening up thanks to the #EHDS European Health Data Space legislation stressing the possibilities to advance science to save lives. Data should be handled safely and individual integrity must remain protected BUT as the panel agreed: It is unethical NOT to help science save lives just because we fail to join hands on Health Data. Action is needed by most EU countries to comply with EHDS. Our lives may depend on it.

35

4 commentaren

🎏 OASYS NOW is bringing its Medical AI to Japan! 🇯🇵 Sara Okhuijsen and I, together with other pioneers of AI in Dutch Healthcare 🇳🇱, are selected by RVO (Netherlands Enterprise Agency ) for a Trade Mission to Japan's most prestigious healthcare and research institutes. 🔬 From OASYS NOW, we will showcase our AI-powered Clinical Trial Patient Identification & Recruitment platform, which speeds up the process of patient pre-screening from months to near real-time. 🔍Our platform leverages state-of-the-art AI and privacy techniques that can process deep patient records (structured and unstructured) to provide a high-precision matching between patients and clinical studies. Essentially, freeing medical personnel from thousands of hours of boring, repetitive, manual tasks behind computers so that they can focus on what they enjoy the most: interacting with their patients. 🩺 Imagine a world where healthcare professionals can dedicate their time to what matters most: caring for patients. That's the future OASYS NOW is building. Join us on this journey! ☎️ If you know anyone in Japan (or here in the Netherlands) who should know about what OASYS NOW offers, share this post with them and tell them to contact me or Sara Okhuijsen to share more details with them! #AIinHealthcare #ClinicalTrials #AIforGood #PatientCare #OASYSNOW

52

11 commentaren

This paper by Steven Schalekamp, Kicky van Leeuwen and colleagues highlights the most likely avenue where AI software could do the bulk of the work, rather than a human, in reporting/reading the entire chest imaging study (CXR in this case). Current prominent examples of AI in thoracic radiology are most commonly only identifying selected abnormal findings (nodules, PE, pneumothorax etc) rather than sorting out entire studies. The software (LUNIT CXR) can identify normal CXRs reliably - AUC was 0.92. We may be now approaching the point where AI accurately reads and reports an entire study (CXR in this case) when normal and the principal, if not only role, of the staff radiologist is to 'sign off'. Whether this augments, displaces, or replaces human readers, is to be seen (I know what I think ...). This all may seem far-fetched but then I am reminded that essentially all CBCs (complete blood counts) are repoted in an automated fashion, and not the normals! https://lnkd.in/gx2KWXp4

13

#Google has introduced a new AI tool called #𝐌𝐞𝐝-𝐆𝐞𝐦𝐢𝐧𝐢. #Med-Gemini is specifically designed by #Google for medical applications and aims to assist healthcare professionals in their work. 𝐇𝐞𝐫𝐞 𝐚𝐫𝐞 𝐬𝐨𝐦𝐞 𝐤𝐞𝐲 𝐢𝐧𝐬𝐢𝐠𝐡𝐭𝐬:- 👉 Its long-context processing capabilities can analyze extensive health records and research papers, providing valuable insights into patient conditions, required care, and more. 👉 Med-Gemini is based on Gemini 1.0 and Gemini 1.5 LLM, and it offers enhanced text, image, and video outputs. Google claims that Med-Gemini provides both accuracy and reliability in clinical reasoning tasks.  👉 In benchmark testing, Med-Gemini outperformed previous models and competitors, scoring a rate of 91.1% on MedQA (USMLE). 👉 Currently, Med-Gemini is only available for beta testing. But once it’s fully launched, it promises patient care and medical research. #AI #Healthcare #MedGemini This is a major step forward in the field of Medical AI and certainly going to change the Healthcare industry. https://lnkd.in/gzV-w9MV

25

6 commentaren

Very nice example of a useful application of AI for health: our paper on using AI for extracting lifestyle (smoking, alcohol use, substance abuse etc) from clinical notes is live. Have a read, looking forward to feedback!

3

Generative AI in healthcare information flow. Important safeguard: no full access to the patient file. Big reduction in time needed from the specialist and increased quality for the patient! If you want to experiment with GenAI in Health, make sure to learn from Elisabeth-TweeSteden Ziekenhuis (ETZ) and UMCG . Thanks Margot de Waal-Verheijen and Jessica Workum for taking the lead as innovators 🙏🏻 #health #innovation #genai #experiment #himss #nza Nederlandse Zorgautoriteit (NZa)

38

5 commentaren

How to create AI-ready culture in healthcare organizations 👉 A fantastic session and practical tips by Ran Balicer, Cris Ross, and Stéphanie Allassonniere at #HIMSS24EUROPE👇 ✅ Healthcare faces the #FOMO (fear of missing out) regarding #AI. There is big AI (genAI, etc.) and small AI that healthcare organizations should focus on (voice-to-text systems, AI algorithms, control towers). Mayo Clinic already has 200 algorithms in use. ✅ When implementing #AI algorithms, always start with “why” and “what next?”. Any new algorithm implies the necessity to develop procedures that follow (translation outcomes into actions). ✅ Predictive analytics can be helpful but are not a one-stop solution. Their application must be embedded into clinical workflows to make the results reliable and actionable (from data to action approach). ✅ Model building is so 2010. Now, it’s about implementation. According to Cris Ross, three modalities are critical: 👉 embedding AI into EHR, avoiding alert fatigue, moving to suggestion-oriented models 👉 one generalized framework for all AI algorithms (unified platform) 👉 chat-based interface, doctors want to speak to EHR, not click ✅ When adopting AI, mind the usefulness and user-friendliness for doctors. Doctors are tired of alerts, messages, and new applications; do not increase the burden of tech innovations on doctors ✅ Generative AI will help to extract information from data (talking to EHR), but when you want to make predictive models, the present genAI version won’t help; you have to reach for deep learning solutions ✅ Multi-modal AI opens completely new opportunities in healthcare. Even if they are not 100% reliable, they can be applied as supportive tools (f.e. for triage) ✅ Start with the question: Can we systematically deploy a new AI innovation? You have to talk to clinicians (do you need it?), IT staff (can we integrate it?), AI office (can we embed it into data flow?) ✅ Before implementing AI, build capacity: train IT teams because otherwise, they will stop every AI project due to lack of expertise; train healthcare professionals because otherwise, they will reject every new AI tool due to lack of understanding of how they can support their work ✅ AI can narrow knowledge gaps in healthcare: doctors need to be experts in many domains, but it’s impossible to have all the medical, up-to-date knowledge only specialists have (memory capacity is limited). AI can bridge this gap and widen doctor’s capabilities ✅ Most of the hospitals, investing 2-3% in IT, will depend on AI models provided by vendors. Think about long-term integration even if they ensure you “the solution is easy to implement, doctors can use it on their smartphones) ✅ AI is excellent, but what about regulations? Marco Marsella from the European Commission argues that the #EU #AIAct won’t hinder the benefits of AI in healthcare, just like “medical devices regulation didn’t stop the development of medical devices.”

59

1 commentaar

The next 5 years of AI in healthcare https://lnkd.in/eH37yW9g #AI #ArtificialIntelligence #healthcare #data By Laura Dyrda via Becker’s Hospital Review

3

The Dutch government is allocating €𝟏𝟎𝟖 𝐦𝐢𝐥𝐥𝐢𝐨𝐧 over the next few years to transform the healthcare and welfare sector. 💡Smart healthcare technology eases the burden on caregivers and streamlines work for healthcare professionals. The glucose monitor featured in the video below is a great example! As a provider, you can receive up to: 🟢 €25,000 in subsidies during the initial phase 🟢 €750,000 if you’re evaluating and scaling an existing idea For more information on the Technology Support and Care Incentive Scheme (STOZ) follow this link: https://shorturl.at/fhuzU. The first application period runs from May 14 to December 20, 2024. Are you a provider looking at developing or improving your own solution, and would you like to take advantage of the subsidies? Our white label platform can significantly reduce your time to market. Get in contact with me for more information. #WOZO #TAZ #TEC #DevelcoProducts #WDTM

14

Challenge your understanding of healthcare innovation! Embracing the power of digital innovation, the upcoming webinar with Ansys and Ghent University will dive into Computational modeling and simulation (CM&S), sharing insights for enhanced liver cancer treatment. We'll explore the transformative potential of 3D technology in personalized medicine, enabling clinicians to tailor treatment strategies with unparalleled precision. Whether you're a healthcare professional, researcher, or industry enthusiast, this webinar brings you to the forefront of healthcare innovation. Secure your spot today!

1

1 commentaar

"For AI tools to succeed in the clinical setting, only 20% is about the algorithm; 80% is about process redesign". I watched a great panel discussion from the Lean Summit 2024 (link below in comments). One key point was that although many healthcare AI models have been developed, very few have demonstrated clinically meaningful benefits. This gap is known as the "AI chasm". Taking sepsis as an example - thousands of papers have been published on sepsis AI models, but only 2 have actually shown real-world improvements in patient outcomes. Why is this so? Even if the AI model is making accurate predictions (no bias, etc), this may not translate on the ground to any concrete actions / changes to care delivery. The problem is that many AI projects start with building the model first and then figuring out how to deploy it in a clinical setting. A better way would be to start with the workflow, and *then* build the AI model that is required to redesign the workflow. Or to use Lean language, don't jump to solutions (in this case, AI tools); the technology should serve your people and processes, not the other way round. Christopher Longhurst also shared this journal editorial which commented on UCSD Health's work and which made the same point - more than just the algorithm, a whole ecosystem is required for AI to perform well in the clinical setting. My own takeaway is that healthcare AI is not all about data scientists and clinicians. While they are obviously essential in this ecosystem, people with skills in process improvement, change management and human factors can play an important role too. Lean Enterprise Institute https://lnkd.in/gFWCZY3i

57

3 commentaren

🚀 𝗜𝗻𝘁𝗲𝗿𝗼𝗽𝗲𝗿𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝗨𝗻𝗹𝗲𝗮𝘀𝗵𝗲𝗱: 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆'𝘀 𝗥𝗼𝗹𝗲 𝗶𝗻 𝗛𝗲𝗮𝗹𝘁𝗵𝗰𝗮𝗿𝗲 𝗧𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻 🚀 I'm excited to share the first part of our new blog series, "𝗜𝗻𝘁𝗲𝗿𝗼𝗽𝗲𝗿𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝗨𝗻𝗹𝗲𝗮𝘀𝗵𝗲𝗱: 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆'𝘀 𝗥𝗼𝗹𝗲 𝗶𝗻 𝗛𝗲𝗮𝗹𝘁𝗵𝗰𝗮𝗿𝗲 𝗧𝗿𝗮𝗻𝘀𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻." 🌐 In this series, we dive deep into the world of healthcare data interoperability, exploring the challenges of fragmented data systems and the transformative potential of achieving true interoperability. In 𝗣𝗮𝗿𝘁 𝟭: 𝗧𝗵𝗲 𝗕𝗿𝗼𝗸𝗲𝗻 𝗘𝗰𝗼𝘀𝘆𝘀𝘁𝗲𝗺 - 𝗪𝗵𝘆 𝗜𝗻𝘁𝗲𝗿𝗼𝗽𝗲𝗿𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝗠𝗮𝘁𝘁𝗲𝗿𝘀, we discuss the critical issues arising from disconnected data and the profound impact on patient care, clinical workflows, and medical research. From real-world examples to the opportunities that interoperability can unlock, this installment sets the stage for a comprehensive exploration of healthcare's digital future. 🔍 Read the full blog post here: https://lnkd.in/ewKtv_AX Join us on this journey towards a more efficient, effective, and patient-centric healthcare system. Stay tuned for the next installment, where we will delve into the lessons learned from past attempts by tech giants like Google and Microsoft. #Healthcare #Interoperability #DigitalHealth #HealthIT #PatientCare #MedicalInnovation #HealthcareTransformation #AdaptiveProduct

4

3 commentaren

It's time to talk AI. 🤖 🎙 In a recent interview with Forbes, our Co-founder, Professor Prof. Daan Dohmen, PhD explained how we shouldn't be using AI to replace humans and why, in healthcare especially, data privacy is an ever-increasing concern for new technological advancements. What are the other mistakes leaders should avoid with AI? Read more in the Forbes article: https://lnkd.in/gJewSa2k #digitalhealth #digitalhealthcare #digitalhealthinnovation #digitalhealth2024 #virtualcare #virtualwards #remotepatientmonitoring #nhs #nhsengland

24

Take time to study the #guidance coming out of the #EU member state #HTA Coordination Group #HTACG and you may realise how important this #PSI #EFSPI discussion is for the successful #implementation of the #HTAR. Reduction of unnecessary #duplication and raising quality #standards that #EUnetHTA worked for was enabled with the #regulation. Now it's up to the busy HTA doers and the applied scientists to do exactly this in practice! Multidisciplinary topic for #HTAi #ISPOR to foster! https://lnkd.in/dKjB6_qb

14

2 commentaren

If there’s one conclusion we can easily draw from AI in healthcare — AI's potential goes beyond individual treatment. It can analyze complex biology to fast-track drug discovery for diseases like cancer. AI-powered tools can also analyze medical images with even greater accuracy, helping doctors catch illnesses earlier. Plus, AI can free up doctors' time by automating tasks and generating reports, allowing them to focus on what matters most – their patients. Of course, with great power comes responsibility. Ethical considerations, biases in AI algorithms, and data privacy are important issues to address. But the potential benefits are undeniable. AI has the power to usher in a new era of personalized, efficient, and preventative healthcare. Here at smartData Enterprises Inc., we're committed to incorporating these advancements into our software development to be part of this exciting future! #smartData #healthcare #healthtech #AI

156

4 commentaren

How to improve EMRs in HMIS by using Clinical Process Improvements tools The new buzz word in the hospital industry is using EHRs (Electronic Health Records) and AI (Artificial Intelligence). The typical HMIS (Hospital Management Information System) is used for digitising to improve a) General Administration and b) Clinical Services, Clinical Quality & Patient Safety In India, the majority of hospitals use HMIS which is mainly used for routine transaction and administrative applications. Most of these HMIS applications lack clinical applications. The flavour or buzzword of the hospital industry and individual Clinicians is EHR & AI. A Clinical module EHR covering one or more specialties typically covers initial clinical assessment of the patient covering complete clinical history and clinical findings and based on these order sets for diagnostic services, medications, interventions, second opinion from other clinical specialities (if required) are used and a typical patient CARE PLAN is made. The indoor patients will require charting of daily medical parameters, medications, adverse events and day to day medical progress in the patient's medical condition till he/she is cured & discharged. Thus, all aspects related to clinical care, clinical quality and patient safety by various hospital teams, specialities, departments are captured digitally. In India, due to nationwide lax definitions and applications of CLINICAL PROTOCOLS by various clinical specialties has resulted in big Physician practice variations. With a few exceptions, most of the hospitals' medical board/administration are also not empowered to implement clinical protocols in their institutions. A typical HMIS vendor, in the absence of detailed clinical protocols pushes low end EMR solutions. The right way is to involve clinical leadership in drawing & documenting clinical workflows (as in use) processes using SOPs (Standard Operating procedures) approach. A good EHR definitely helps in improving clinical efficiencies by application of clinical protocols, tracking & coordinating patient health & safety and tracking Clinical Quality Indicators. These in turn leads to a positive cycle of CQI (Continuous Quality Improvements) of the clinical services, whether in individual clinician practice or in an indoor setting. Artificial Intelligence and speciality wise Clinical Decision Support Software (already commercially available) applications in medicine comes into picture and help healthcare organisations only when there is a good foundation of digital medical / clinical data is available The Clinicians and Clinical Leadership in the healthcare / hosptal industry needs to become proactive by shedding their fears in anything which requires digitisation of their medical records. Dr Brita Hansen, MD, Chief Medical Officer, LogicStream in her article 'Three Pillars of Clinical Process Improvement & Control' in Healthcare IT TODAY has beautifully described her approach

5

1 commentaar