Philips uses its intelligence for outcomes-based incomes

Philips prioritises artificial intelligence-based technologies as it seeks to grow its Connected Care and Health Informatics unit

When Philips posted its quarterly results on Monday its Connected Care and Health Informatics unit was a bit of a let-down. Its sales were flat, largely owing to a consent decree issued by the FDA in October that banned the company from selling its external defibrillators in the US. But the unit is working hard to improve, partly by prioritising artificial intelligence-based technologies linked with patient monitoring.

“With the help of AI we’re not just observing and analysing better but also predicting,” says Carla Kriwet, chief executive of Philips’s CCHI unit. “We can take data together which wasn’t linked in the past, and can tell you which patient is going to get sickest and need intervention.” The potential of these technologies is enormous, but the difficulties of implementing them in often cash-strapped healthcare systems are substantial.

One of the issues is getting different technologies, often from different manufacturers, to work together. When setting up a new piece of software like an image analysis system there is no guarantee that it will work with the existing infrastructure – the software that comes with the ultrasound or MRI machine, perhaps.

“Interoperability is one of the hurdles for analytics and AI,” says Ms Kriwet. “We’ve committed ourselves to open systems, which was different from in the past. Now when you talk to the younger generation it’s all open source. You cooperate with competitors, start-ups, universities, customers. Interoperability becomes a key thing.”

Philips recently acquired Forcare, a Dutch group whose technology is designed to support interoperability. Ms Kriwet describes this as “a bit like Google for a patient” – the user types in the patient’s name and all the data on the patient, including clinical images and lab results, are in one place. This is just what healthcare staff want, she says.

“I’ve never heard somebody say, ‘give me more data.’ They say, ‘give me my time back.’ If we can help them to get the data and cut the bureaucracy there’s a benefit for them.”

Put your money where your mouth is

If integration into existing tech infrastructure is tricky, making money out of this type of sophisticated software is less so. For once, payers seem to be relatively easy to get on side.

“AI is helped by reimbursement that is focusing on value,” says Ms Kriwet. These technologies can benefit from an industry-wide push towards paying for improvement in patient outcomes, she says, since the data they collect can be used to prove their own clinical value and effect on productivity.

Philips has taken this to the logical conclusion, and believes that it can track when the use of its technologies saves a patient’s life. This might be a bit much for payers, but they will surely accept, for example, date-stamped images that prove more patients can be imaged in a single day using Philips software than was possible before.

The group is also at the forefront of the industry-wide shift towards subscription-based payment models. Speaking on the first-quarter results conference call, Philips’s chief executive, Frans van Houten, said the CCHI business in particular would “feel this more profoundly because it … lends itself to these subscription models”.

Ms Kriwet says Philips offers a range of payment models. Customers can pay the old-fashioned way, up front for a device, or they can opt for a service-based deal where they pay per usage of the product. But the company is seeing increasing interest in what Ms Kriwet calls the outcome model, where Philips only gets paid if productivity, clinical outcomes and sometimes even patient and caregiver satisfaction standards are met.

“If we say we can reduce the time for radiography picture analysis by 20%, if we make 20% we get X. If we make 30% we get X more – if it’s less than 5% we get nothing,” she says. “More and more hospitals say, ‘I don’t care how you do it, all I care about is the outcome. Put your money where your mouth is.’”

Back-end loaded

The downside to subscription-based payment, of course, is that you do not get the cash up front. On the call, Mr van Houten said the time window of order recognition was about 12 to 18 months. He added that for the CCHI unit this financial year would be back-end loaded.

The other main reason for that is the consent decree the FDA slapped on Philips’s automated external defibrillators last autumn. The agency said the products were made in violation of current good manufacturing practice requirements, and prohibited Philips making or selling some of the devices made at two US factories until it complied with regulations.

In its first-quarter report the company recorded €20m ($24m) of charges related to the consent decree, bringing the total it has paid to €40m. The company believes that it will have rectified the issues and got the shipping hold lifted by the end of this year. Ms Kriwet calls this a learning opportunity.

Consent decree or not, Philips cannot be faulted in its efforts to grow the CCHI unit. The company is buying in new software and healthcare IT technologies fairly aggressively, and Ms Kriwet says more acquisitions are definitely part of the unit’s future strategy, along with partnerships.

As well as buying Forcare in December, Philips has completed seven tuck-in deals in the areas of informatics and other software and apps in the past two years.

Still, it has not announced any new purchases so far this year. Perhaps it now intends to settle down and await the fruits of its deal-making. If the technologies it has bought are as appealing to payers as Philips hopes, financial 2018 could end rather better than it started. 

Philips’s recent healthcare informatics and IT acquisitions  
Date Target Technology
Dec 2017 Vitalhealth Population health management technology
Dec 2017 Forcare Open standards-based interoperability software 
Nov 2017 Analytical Informatics Web-based software to enhance clinical and operational performance of imaging departments
Jul 2017 Tomtec Imaging Systems Intelligent image analysis software, especially for diagnostic ultrasound
Jul 2017 Health & Parenting Healthcare and family-related mobile apps for expectant and new parents
Jun 2017 Electrical Geodesics Non-invasive technologies used to monitor and interpret brain activity
Jul 2016 Wellcentive Population health management technology
Jun 2016 PathXL Digital pathology image analysis, workflow software and educational tools
Note: financial terms undisclosed except for Electrical Geodesics, acquired for £29m cash. Source: EvaluateMedTech and company website.

To contact the writer of this story email Elizabeth Cairns in London at or follow @LizVantage on Twitter

Share This Article