AI in Healthcare – news picked by GLI /06
Welcome to AI in Healthcare for October 2023. We’ve created this very series to provide you with news, updates and viewpoints handpicked by the Graylight Imaging (GLI) team. In the current edition, we’ll delve into healthcare reimbursement policies and real-world AI in healthcare applications.
The reimbursement can truly stimulate innovation
In our latest post about the radiology AI landscape in number, we wrote about the significance of reimbursement policies, which have a substantial impact on the adoption of AI in healthcare by directly affecting the financial incentives for healthcare providers. I’m confident that we all concur that thoughtfully crafted reimbursement policies can stimulate innovation and the creation of AI solutions tailored to tackle distinct healthcare challenges. This confidence is commonly shared by probably all medical societies, but lately has been brought up by the Society of Cardiovascular Computed Tomography (SCCT).
Since its establishment, SCCT has directed its endeavors toward promoting an increased number of clinical coronary CT angiography (CCTA) studies with the aim of gathering essential data to bolster the broader utilization of cardiac CT within clinical practice. In 2021 CCTA received Multiple Class 1, Level A recommendations in 2021 New Chest Pain Guideline [1]. SCCT’s focus has recently shifted to advocacy and reimbursement policies: the society hired a lobby firm in Washington, D.C. and asked its members to write letters to legislators on various CMS funding issues [2]. We keep our fingers crossed for the efforts that they’re undertaking, as the analysis of medical images is the pivotal area of AI in healthcare applications.
AI in healthcare reimbursement strategies in European countries
Digital and AI-enhanced solutions in healthcare require reimbursement, but what are the strategies that particular countries implement in order to proceed with them? A recent publication by Robin van Kessel et al. [3] investigates the reimbursement landscape in 8 European countries (Belgium, France, Germany, Italy, the Netherlands, Poland, Sweden, and the United Kingdom) and Israel.
We highly recommend reading it – here we’d like to share our 3 takeaways:
- The pricing of solutions is primarily established through dialogues between national committees and the manufacturers. Nevertheless, the specific qualifications required for participation in these committees currently remain unclear. There is a need for greater transparency in this area and possibly the establishment of international guidelines to enable patients to benefit from these solutions to the same extent even when staying outside their own country.
- A fee-for-service approach is the most applied. However, it also opens the door to supplier-induced demand, which could lead to an increase in healthcare spending.
- Among the analyzed countries only the landscape in Poland remains very limited. Polish reimbursement strategy primarily encompasses digital consultation tools facilitating interactions between healthcare professionals and patients, as well as consultations among healthcare professionals, which seems to be a post-pandemic legacy with no further steps towards the adoption of digital healthcare.
AI in healthcare and AI-based cyberattacks
Is the mentioned example of Polish reimbursement strategy isolated or truly it was a covid-19 pandemic that transformed our approach towards innovations? There is no discussion about that when it comes to cloud computing adoption (check our post Cloud computing in healthcare – a promising avenue for the industry’s advancement?). Until 2020, cloud computing appeared to be merely an appealing option in healthcare – today it is listed by tech giants as the main direction. In a recent post by Michael Giannopoulos, Healthcare CISO & CTO, Dell Technologies another interesting point is made [5]: AI is not only a promising option. The emergence of AI-driven cyberattacks and deepfakes presents an elevated risk that healthcare institutions and their technology collaborators must be ready to address. In 2019 we all read about the malicious software developed by an Israeli research team that was able to transform CT and MRI images and fake patient progression or regression [6].
A video of the simulated attack can be viewed on Cyber Security Labs at Ben Gurion University YouTube channel:
If that sounds scary (and it should) we must address vulnerabilities and take advantage of the state-of-the-art solution in an informed and secure manner. With cloud computing, it is not that hard, which we prove in the below real-world example.
AI for colonoscopy video analysis – real-world and secure application of AI in drug development
Inflammatory Bowel Disease (IBD) is a group of chronic, inflammatory disorders that primarily affect the gastrointestinal (GI) tract. According to research [7], the number of people living with inflammatory bowel disease (IBD) is increasing – currently approximately 1.6 million Americans have IBD. These figures indicate the need for additional research to find a cure. ‘AI for colonoscopy video analysis’ was the title of a presentation by Benjamín Gutierrez Becker, R&D Imaging Data Analysis Expert at Roche during this year’s edition of BioTechX Congress Europe in Basel. The attendees were able to familiarize themselves with this AI-enhanced research supported by our Graylight Imaging’s experts. We are honored to be a part of such an important project and to contribute our knowledge to the fight against a disease that affects tens of thousands of patients annually.
References
[3] van Kessel R, Srivastava D, Kyriopoulos I, Monti G, Novillo-Ortiz D, Milman R, Zhang-Czabanowski W, Nasi G, Stern A, Wharton G, Mossialos E, Digital Health Reimbursement Strategies of 8 European Countries and Israel: Scoping Review and Policy Mapping. JMIR Mhealth Uhealth 2023;11:e49003, https://mhealth.jmir.org/2023/1/e49003 DOI: 10.2196/49003
[5] https://www.bbc.com/news/technology-47812475
[6] https://www.crohnscolitisfoundation.org/sites/default/files/2019-02/Updated%20IBD%20Factbook.pdf