Hospitals invest in AI – How can they assess the return on investment?

Will artificial intelligence (IA) be made or will break the finances of the hospital? The excitement around the promise of AI has attracted billions of dollars in investment, but it remains difficult for hospitals to predict the value they can expect to see these technologies in the future.

A recent McKinsey survey revealed that, even if about half of the leaders of health systems provide for a return on investment (King) of the AI, only 17% were currently able to measure a positive return. With so many hospitals operating on negative or thin margins, they do not have the flexibility to play their resources on tools which can or may not support long -term growth.

Historically, hospitals have experienced innovation passively rather than driving it actively – seen in cases such as the implementation of the compulsory DSE and the digital entry doors focused on consumers. To prevent the history from repeating themselves, health leaders must seize the opportunities that AI can offer, in particular operational efficiency and profitability.

Hospitals’ challenge

New technologies have historically had a mixed assessment of their promise – evoking a healthy skepticism of hospital leaders. The traditional business model of health care technology has improved the results of software companies, but health systems – which are counting on these technologies for everything, from radiological imagery to invoicing patient complaints – have not always seen the same advantages. While they continue to operate on a competitive market with median margins below one percent, productive and responsible investments in AI are increasingly vital for their economic survival and their ability to serve patients. AI must work for health systems, but decision -makers need confidence that they make the right choices.

The combination of tight margins and a struggle for hospitals even well reinforced to stay afloat can force leaders with hasty AI investments. Anxious, some enter the first solution that seems to stimulate efficiency, often without sufficient planning of the return on investment. Others wait for a magic box to appear and solve their problems. The two can threaten future solvency and the capacity of hospitals to provide care, and no longer a strategy.

To operate AI for their hospital, managers need a plan – a clear approach to measure value, selection of tools and scaling what works.

Start with the advantages

Thinking of the king should start by appreciating all the potential advantages. Many people think exclusively of AI Automation and cost savings, but this is only part of the image: AI can also help teams do better work – improve precision, expand the scope and allow more intelligent decisions.

Take the example of the integrity of clinical documentation (CDI). Even the most experienced CDI teams can only be systematically capable of the most common 40% of missed codes – such as sepsis and respiratory failure. But what about the long tail – the 60% others? AI can help discover less frequent but high impact diagnostic codes, considerably improving capture and income without replacing the human team. Conclusion: IA is not just about doing the same job for less. It is also about doing a better job – smarter, faster and more complete.

And even if you have great ambitions, restrict your reach to start with a small victory. Hospitals should start using AI with a relatively small number of people – with a department, for example – then develop once at ease with how to select and operationalize AI. When AI strategy managers start small, then demonstrate a positive return on investment, they will earn political capital that can be exploited for more investments.

Align success measures with analysis teams

Many hospitals do not have the internal analysis capacities to measure the return on investment by themselves, they therefore count on AI sellers to do so. As a former medical director of transformation in a large hospital responsible for supervising applied AI and in my current role leading to a clinical company in AI, I saw this challenge on the hospital side and table sellers. Before a hospital selects a supplier, he must align himself with his analysis teams and suppliers about how to allocate the value and what success looks like.

This is a crucial step: you do not want poorly designed measures to work against your IA investment. If a supplier reports that efficiency has improved by 80% by only measuring the cost, but 100 people do a task that could be performed by 20, you may think that you have not really improved efficiency. Define how the value will be attributed in advance and be as specific as possible. Make sure your supplier is on board – and kept responsible.

Help the sellers have help you

Sellers want to understand North Star’s metrics for which they optimize, but they may need help. This is a common problem for hospitals. To solve it, hospitals and sellers should cross the steps they would take to resolve their challenges together. If a supplier cannot clearly explain how its product will provide value in your context, it is a red flag. And if they can, give them the data and the context they need to succeed. King is a shared responsibility.

Build the right team for the success of AI

The success of AI does not only depend on technology. It depends on the people who choose, the piloting and the champion. A working group on AI of a few highly capable leaders, supported by a solid internal analysis team, can help a hospital system to make smart bets by working closely with technological partners and internal stakeholders to assess and validate AI tools.

You don’t need a huge committee. Just a few strong people, curious and analytical to the analytical spirit can make a big difference.

Learn by experience to plan the future

The adoption of AI in hospitals is on an exponential curve as confidence in its performance increases. The capacities which once seemed futuristic – such as the understanding of the man of clinical documentation – are now commercially viable with powerful models of large -scale language. If hospitals are not starting to learn from AI now, they may not only late. They may also miss the upward AI that AI can provide.

AI already proves its value in the back office of health care. Take the management of the income cycle: AI can now perform second level exams of each patient table before billing – an application that increases efficiency while generating a return on investment of 5: 1. It is not a future potential. It’s a real performance today.

AI does not need to be a black box – and hospitals do not need to invest according to the blind faith. With the right structure, the questions and measures in place, health leaders can reduce overhaul and make decisions that really stimulate value.

In a financial landscape where each investment counts, AI cannot simply be promising – it must be productive. And for prospective hospitals, this is already the case.

Image: Warchi, Getty Images

This message appears through the Medcity influencers program. Anyone can publish their point of view on business and innovation in health care on Medcity News through Medcity influencers. Click here to find out how.