Startup Health Tech Companies are Facing Critical Junctures
In 2013, a patient was given 38.5 times the dose of his regular medicine due to a mistake in his electronic health record (EHR). The doctors, nurses and pharmacists involved carried out the order because that’s what the EHR directed.
Incidents like this have raised concerns over the rapid rise of technology in healthcare, creating trust issues among medical professionals and patients while drawing increased scrutiny from regulators.
Because the consequences of a technological failure at a hospital are much greater than in an office, medical professionals are hesitant to adopt new tech unless they’re absolutely sure it will work. Challenges around regulations and clinical relevance are also creating issues in health tech AI adoption, compliance and implementation.
A KPMG survey revealed that there are other factors slowing down AI implementation in healthcare, including costs and training. Despiteexcitement over the potential of AI, only 47% executives say their institutions offer any relevant AI training for employees and only 67% of employees are onboard with AI adoption.
It’s crucial for health tech innovators and leaders to understand the various challenges they face both at an industry and a startup level.
Challenges AI health techs face at the industry level
The healthcare industry – ever cautious, slow to change and riddled with regulations – continues to place roadblocks in the path of AI health tech startups, including:
- Risk aversion to new technologies: In the name of protecting consumers, healthcare institutions have blocked innovation.
- Lack of curated data sets: Given privacy rules and HIPAA restrictions, robust sets of data required for supervised learning are still difficult to come by in healthcare.
- Noisy data: Data sets from myriad sources, such as data related to drugs, can be inconsistent and lose integrity.
- Unclear ROI of implementing AI-based processes:Stakeholders aren’t clear on how to measure ROI and whether AI applications will live up to the hype.
- Failure to keep data secure due to digitisation: As medical data remains a valuable asset to hackers, medical institutions are more averse to data-dependent AI applications.
- AI solutions that lack context and understanding: AI solutions designed in a neutral environment rarely match the needs of real-world demands.
- A case study catch-22: Even though medical institutions insist on more case studies before they adopt an AI application, these are hard to come by without prior implementation.
- Knowledge gap: Most healthcare institutions lack the resources to train staff to implement AI in their daily practices.
- Complex shareholder relationships: There’s still a persistent sense among healthcare industry workers that AI tools will take over jobs.
- The age-old black box problem: While a successful AI application can arrive at the right decision, it can’t provide justification. Healthcare AI users are deterred from using the solution when they don’t know its inherent logic.
Challenges AI health techs face at the startup level
AI health tech leaders and innovators also face numerous challenges at the startup level:
- Vested interests in AI have led to too many redundant or not-true AI products. Without technical innovation, health techs will face steeper competition and lower their success rate. That is why Daphne Zohar, founder of PureTech, asks health venture entrepreneurs to first differentiate the concept, and ensure it’s really different from what other people are doing.
- Seamless integration of AI in healthcare with other apps. AI applications in healthcare need to integrate with increased computational power, digital interfaces, remote operations and telemedicine. AI solution providers also need to think about backend support and training users.
- Regulations in AI health tech are rising. These regulatory models will disrupt medtech and navigating them will require innovative strategies both from a technical and business perspective.
- AI health tech startups need strategic partnerships to survive. Some of the key players in the AI in healthcare market include IBM, NVIDIA, and GE. Strategic partnerships with mature entities will be crucial to build the profile of AI healthtech startups.
Zohar’s PureTech partnered with Enlight Biosciences and pharma companies for both strategic and financial profits. She says, “We have to identify opportunities that cross both areas of interest to provide both strategic value for our partners and financial value for us.”
AI in healthcare innovation will be a collaborative effort amongst several stakeholders, including product developers, investors, healthcare providers, patients and policymakers. Join leading industry experts, including Daphne Zohar, who was named one of the world's top young innovators by MIT's Technology Review magazine, to discover how these variables interplay in creating a successful venture in MIT's Leading Health Tech Ventures course.