Healthcare Analytics, Population Health Management, Healthcare Big Data

Tools & Strategies News

Artificial Intelligence a “Low Priority” for Healthcare Orgs in 2018

Artificial intelligence is taking a back seat to other big data basics due to unclear use cases and the need to create a more consumer-centered care environment.

Artificial intelligence in healthcare and big data

Source: Thinkstock

By Jennifer Bresnick

- Close to two-thirds of healthcare organizations are marking artificial intelligence deployments as low or very low priority in 2018, according to a new survey, indicating that AI may not be coming as quickly to the industry as some enthusiasts hope.

The Center for Connected Medicine poll, fielded to technology and clinical executives at more than 20 large health systems, shows that large health systems are more interested in bulking up their cybersecurity defenses, developing consumer-facing technologies, and leveraging predictive analytics for precision medicine than they are in AI innovation.

While machine learning and artificial intelligence may eventually become the driving technologies for many of these initiatives, organizations seem more likely to adopt readily available products that will help them meet their immediate needs than taking on the challenge of developing, piloting, or programming advanced analytics systems.

Just because AI isn’t on the table yet, however, doesn’t mean that organizations are ignoring the transformative impact of complex big data analytics.

Collecting and using patient-generated health data (PGHD) to use in consumer experience projects, population health initiatives, and chronic disease management programs is high on the list of goals for 2018.

READ MORE: How Soon Will Healthcare Connect Machine Learning with Consumers?

The ability to gather PGHD at scale requires high levels of consumer engagement – this, in turn, requires healthcare organizations to develop a number of new competencies.

In order to leverage PGHD, organizations must focus on deploying consumer-facing applications, optimizing EHRs to accept new data sources, understanding the nuances of the Internet of Things, developing patient-centered care teams that drive engagement, and ensuring the privacy and security of novel data sources.

Thirty-three percent of respondents to the poll said they will be working on integrating PGHD into the EHR in 2018.  Forty-six percent said they already accept structured PGHD data in their EHRs, while 8 percent have integrated unstructured patient-generated data sets.

The vast majority (88 percent) believe that patient portal data will provide the most value for their organizations in 2018, followed by 56 percent who see high value in home monitoring equipment. 

Sources of valuable patient-generated health data in 2018

Source: Center for Connected Medicine / Health Management Academy

READ MORE: Navigating the Hype of Healthcare Artificial Intelligence Companies

Mobile health apps (21 percent) and consumer-grade wearables (17 percent) produced significantly less enthusiasm, perhaps due to the fact that these data sets are generally unstructured and difficult to analyze using traditional tools.

“I see [most health data being generated by the patient from personal devices] as a trend, however I believe 3-5 years is quicker than the adoption will happen,” said a CIO participant. “The primary reason is security and liability related.”

Nevertheless, providers see significant potential for mobile health apps to improve the patient experience and create a consumer-centered environment.  One hundred percent of respondents to the poll are planning to promote health and wellness apps to their patients in 2018.

Physicians will be largely responsible for promoting or “prescribing” specific apps to patients, with 75 percent of participants planning to enlist their MDs for this task.  The same number will also use their patient portals to promote health and wellness tools.

“[Mobile health applications are] patient centric – by using these applications, patients feel better connected and they are receiving health care in the way they want to receive it,” said one CEO.

READ MORE: Top 10 Challenges of Big Data Analytics in Healthcare

Remote patient monitoring is also in high demand among patients.  Forty-two percent of organizations said they are implementing telehealth or home monitoring programs due to pressure from consumers, while 33 percent added that increasing patient access to care is a top driver for these initiatives.

Close to half of respondents see a clear case for using remote patient monitoring to lower costs, and a third believe it will help generate new revenue.

In contrast, the business case for artificial intelligence is much murkier, which may be why only 5 percent of respondents said AI solutions are among their top IT priorities.

“[It] boils down to what’s the return of investment and where does it get applied,” said a CFO. “AI in a more tried and true area, such as business functions (e.g., revenue cycle), may make more sense, and make it easier to justify the expense.”

Healthcare organizations are still working through the basics of big data analytics, added a CIO, and AI is simply not mature enough yet to slot easily into their data science roadmaps.

“I think that health care is still figuring out how to get value from data in general. We have use cases, but they are not ubiquitous,” the participant said.

“Until data is ubiquitous, it makes AI hard to prioritize. Additionally, it’s great to predict something, but if you don’t have a corresponding intervention it doesn’t do much. It’s interesting, but still an experiment.”

Despite the skepticism, more than half of organizations are using some sort of product described as leveraging “artificial intelligence,” the poll revealed.

Forty-six percent of AI users have clinical decision support (CDS) tools in place, followed by 33 percent using the strategy to support their population health initiatives.  A third have focused their AI implementations on readmissions, while 21 percent are targeting costs.

How artificial intelligence will be used in healthcare in 2018

Source: Center for Connected Medicine / Health Management Academy

Among those who are planning to implement AI in 2018, seventeen percent will also try to tackle revenue cycle management, cost-cutting, and financial health plan relations. 

Relatively few have precision medicine for oncology in their crosshairs: just 4 percent have adopted AI for cancer care, and only 8 percent are planning to do so within the next twelve months.  The cost and complexity of working with cancer-related datasets has made it difficult for some organizations to see progress.

“Cancer care has very complex formulas – very specific,” said a CIO. “We’ve worked on this for about three years – moving forward with a few more order sets and a few things around cancer care. With all the complexity, it has been very dicult to move forward.”

Other organizations have seen more success in the realm of precision medicine, however.  Thirty-five percent are currently using genomic testing and data analytics to personalize care, while an additional 22 percent are planning to start in 2018.

“We have a big investment from a research perspective in genomics,” said a CFO. “Researchers are more interested in the cutting-edge science, but it needs to be translated to how we deal with patient care. We have invested real money in it and it is a priority for us, but need to determine what the improvement [will be] in health outcomes and financial capability to keep investing.”

Building momentum to ensure data analytics remains a primary goal is challenging for most organizations responding to the survey. 

While more than 70 percent of providers believe that big data analytics can improve patient safety and care quality, reduce readmissions, and aid with CDS, sixty-seven percent said competing initiatives and scant resources were their top barriers to implementing predictive analytics tools. 

Thirty-eight percent pointed to cultural barriers within the organization as a reason why their big data plans may stall.

Organizations may be leery of overextending their big data competencies in the short term, but they appear to recognize the longer-term value of novel data sources and advanced analytics tools for improving quality and reducing costs.

“The change [in care delivery and patient experience] has been gradual,” noted a CIO. “However, I see the pace increasing in 2018.”

Focusing on cultural changes, shrewd allocation of resources, and forward-thinking planning for patient-centered care initiatives will help accelerate the adoption of sophisticated big data analytics tools in 2018 and beyond.


Join 25,000 of your peers

Register for free to get access to all our articles, webcasts, white papers and exclusive interviews.

Our privacy policy

no, thanks

Continue to site...