Skip to content Skip to footer
Search

Loading Results

Adoption of artificial intelligence in healthcare

AI in healthcare is just getting started

Investment in artificial intelligence is growing in every industry, but particularly so in healthcare. AI funding in healthcare is at historical highs, confirming that the healthcare industry is emerging as one of the frontrunners in AI research. Various investigations have revealed that the potential benefits for the industry are huge – according to research we concluded in 2017, AI use for the prevention, diagnosis and treatment of just three prevalent conditions (childhood obesity, dementia and breast cancer), could yield cost savings of over EUR 170 billion over the next ten years, while improving access and availability of timely care. The potential is huge, yet, adoption of AI applications have been rather slow to take off. 

Download the report

(PDF, 1.0mb)

Challenges indicated by healthcare stakeholders in the Netherlands

The data challenge

Availability of data is fundamental to the development and use of AI applications. The interoperability of data is therefore one of the most critical challenges for AI in healthcare. Some developments, like the move towards developing a European Electronic Health Record exchange format for EU countries are a step in the right direction, though a lot more needs to be done. Other data related issues, such as who is the owner of patient data systems in hospitals, are equally important. To lead the way to an AI-based future, the owners of such systems should be well versed with the potential of data, and not just its traditional uses, to allow for more innovation.

The mindset gap

AI based tools have long remained black boxes, not just for patients but also for doctors and other care professionals. The proportion of medical professionals who have the skills needed to fully understand the tools remains small, hindering the adoption of AI in healthcare. Apart from that, having the mindset to prioritize innovation is also vital. Bringing medical professionals to the table to prioritize new technologies’ adoption is often difficult, purely by virtue of their priorities being set in providing care on a day to day basis.

Coordination failures

Considering the black box issue with AI, which is unlikely to go away anytime soon, co-creation might well be the only way to ensure trust and transparency. Market coordination is a prerequisite to ensure investments are directed effectively, yet, at the moment AI research in the EU is rather fragmented, with no definite ecosystem in play. Apart from co-creation and AI development, it is equally important to have a coordinated dissemination strategy from the government, that spells out how adoption of AI tools will move from large hospitals to smaller ones.

Payment and incentive models

The existing payment systems, where healthcare providers are largely paid through fee-for-service systems, do not align well with the aims of AI. These payment systems incentivize maximizing patient volumes, whereas AI is aimed at creating more value by improving health outcomes. In other words, the extra value that AI generates in healthcare needs to be financially rewarded more directly, by gradually moving towards value-based healthcare.

Set priorities for implementation

AI development for healthcare applications is not slowing down. In view of the technical developments, it will be important to set priorities to allow for simultaneous adoption across healthcare services. We emphasize focusing on local challenges while building adoption strategies for the EU.

1. Determine a national coordinated strategy for AI in healthcare

Though there have been various efforts to develop a national strategy for AI, there needs to be higher coordination among stakeholders in the country to ensure widespread adoption. Correspondingly, at the macro level, collaboration between European and multi-national policies and programs would also be vital for the creation of a strategy that allows for seamless adoption of AI for healthcare, that are accessible throughout the region.

View more

2. Prioritize applications based on need and ease of adoption

When it comes to ease of adoption, not all AI-based applications are created equal. To allow for widespread access and use, it will be important to focus on applications that are low on investments and suited for wider use. Decision makers must understand that it is not just the technology and its feasibility that is important. Equally vital is to understand the business side of things, so that the added value is realized for the patients and the providers. PwC’s BXT model defines the different points of view that must be considered before taking such investment decisions.

View more

3. Address data issues urgently

AI development for healthcare, as well as effective use is highly dependent on availability of data. Focusing on EHRs is a good strategy, as is clearing issues and concerns related to data privacy. Innovation should be focused on answering data challenges – including data management and data governance, but also developing devices and systems for protecting individual health information.

View more

Contact us

Jan Willem Velthuijsen

Jan Willem Velthuijsen

Chief economist PwC, PwC Netherlands

Tel: +31 (0)62 248 32 93

Chantalle Wullems-Beister

Chantalle Wullems-Beister

Partner, Risk Assurance , PwC Netherlands

Tel: +31 (0)61 094 44 83

Willeke Bakker

Willeke Bakker

Partner, PwC Netherlands

Tel: +31 (0)61 089 31 82

Sander Visser

Sander Visser

Partner, PwC Netherlands

Tel: +31 (0)62 279 31 20

Follow us