Report: Healthcare Orgs Unprepared for Societal, Liability Issues of AI and IoT | Healthcare Informatics Magazine | Health IT | Information Technology Skip to content Skip to navigation

Report: Healthcare Orgs Unprepared for Societal, Liability Issues of AI and IoT

June 11, 2018
by
| Reprints

The healthcare industry is aggressively adopting intelligent technologies, such as the internet of things (IoT) and artificial intelligence (AI), but many health organizations need new capabilities to ensure that technology acts with responsibility and transparency as businesses evolve, according to a new Accenture report.

According to Accenture’s Digital Health Technology Vision 2018 report, more than three-fourths (77 percent) of the 100 health executives surveyed said they expect to invest in IoT and smart sensors this year — the highest among the 20 industries included in the broader Accenture Technology Vision research on which the health industry report was based. In addition, more than half (53 percent) of the health executives expect to invest in AI systems, with four-fifths (86 percent) of the executives saying that their organizations use data to drive automated decision-making at an unprecedented scale.

And, a majority of health executives (85 percent) surveyed agree that every human will be directly impacted on a daily basis by an AI-based decision within the next three years, and most (80 percent) agree that within the next two years, AI will work next to humans in their organization, as a coworker, collaborator and trusted advisor.

At the same time, healthcare executives using technology responsibly and in a transparent manner is critical. Ninety-two percent of health executives believe that ensuring the security of consumer data is important or very important to gain customers’ trust.

The study identified a range of issues related to the aggressive adoption of AI and the greater role it plays in healthcare decision-making, and the report also counsels the need for organizations to instill trust and transparency into the design of their technology systems.  

Eighty percent of surveyed health executives believe AI is advancing faster than their organization’s pace of adoption. More concerning, 81 percent of health executives agree that organizations are not prepared to face the societal and liability issues that will require them to explain their AI-based actions and decisions, should issues arise, according to the Accenture survey. As a result, 73 percent said they plan to develop internal ethical standards for AI to ensure that their systems act responsibly.

In addition, health organizations also face a new kind of vulnerability: inaccurate, manipulated and biased data that leads to corrupted insights and skewed results. More than five-sixths (86 percent) of health executives have not yet invested in capabilities to verify data sources across their most critical systems. In addition, one-fourth (24 percent) of the executives said that they have been the target of adversarial AI behaviors, like falsified location data or bot fraud.

The Accenture report also predicts key trends likely to disrupt business over the next three years, including virtual/augmented reality, blockchain and edge computing. Among the findings from healthcare executives about these technologies:

  • More than four in five (82 percent) of the executives said that extended reality—comprising virtual- and augmented-reality technologies—removes the hurdle of distance in access to people, information and experience, with nearly half (48 percent) of health providers and one-sixth (16 percent) of health payers planning to invest in these technologies in the next year.
  • Nine-tenths (91 percent) of health executives believe that blockchain and smart contracts are critical to enabling a frictionless business over the next three years, and approximately the same number (88 percent) believe that microservices will be crucial for scaling and integrating ecosystem partnerships. 
  • Four-fifths (82 percent) of health executives believe that “edge” architecture will speed the maturity of hyperconnected health environments, and slightly more (85 percent) believe that generating real-time insights from the volumes of data expected in the future will require computing “at the edge,” where data is generated. Yet the vast majority (86 percent) of health executives believe that they’ll need to balance cloud and edge computing to maximize technology infrastructure agility and enable intelligence everywhere throughout their organization.

“Intelligent technologies, such as AI, are enabling health organizations to evolve at speed, collaborate with other entities and create deeper, more meaningful relationships with patients across various care settings,” Kaveh Safavi, M.D., head of Accenture’s global health practice, said in a statement. “As this paradigm-shifting technology evolves—making business more dynamic than ever before—organizations will remain responsible for demonstrating data stewardship and designing systems with trust and transparency to bolster the societal benefits of these technologies.” 

 

2018 Seattle Health IT Summit

Renowned leaders in U.S. and North American healthcare gather throughout the year to present important information and share insights at the Healthcare Informatics Health IT Summits.

October 22 - 23, 2018 | Seattle


/news-item/it-applications/report-healthcare-orgs-unprepared-societal-liability-issues-ai-and-iot
/news-item/it-applications/healthcare-it-ma-activity-increases-23-percent-1h-2018

Healthcare IT M&A Activity Increases 23 Percent in 1H 2018

July 30, 2018
by Heather Landi
| Reprints

Mergers and acquisitions (M&A) activity in the healthcare IT industry rose during the first half this year, with volume increasing 23 percent, from 101 deals in the second half of 2017 to 124 deals so far in 2018, according to investment banking firm Berkery Noyes.

The firm’s first half 2018 M&A trend report examines deals in the healthcare and pharmaceutical information and technology industries.

The industry’s largest sponsored transaction in the first half of 2018 was Veritas Capital-backed Verscend’s announced acquisition of Cotiviti, a provider of payment accuracy and analytics driven solutions, for $4.82 billion. Veritas Capital also completed another deal in the top ten list during the half year period with the announced acquisition of the value-based care division from GE Healthcare for $1.05 billion, according to the report.

Seven of the industry’s top ten largest deals year-to-date occurred in the healthcare IT segment.

Overall, for the pharma and health IT industries, total M&A transaction volume in 2018 declined by three percent over the second half of 2017, from 239 to 232.

Total transaction value in the first half of 2018 rose by 54 percent over the second half of 2017, from $13.63 billion to $20.93 billion, according to the report.

The median revenue multiple, after rising from 1.9x in the first half of 2017 to 3.0x in the second half of 2017, declined to 2.3x in the first half of this year. During the last 30 months the median revenue multiple was 2.1x.

Berkery Noyes tracked 1,154 healthcare and pharma IT transactions between 2016 and the first half of 2018, of which 244 disclosed financial terms, and calculated the aggregate transaction value to be $72.62 billion. Based on known transaction values, the firm projects values of 910 undisclosed transactions to be $9.22 billion, totaling $81.83 billion worth of transactions tracked over the past two and a half years.

 

More From Healthcare Informatics

/article/it-applications/florida-health-it-summit-candid-look-promise-and-pitfalls-blockchain

At the Florida Health IT Summit, a Candid Look at the Promise and Pitfalls of Blockchain in Healthcare

July 25, 2018
by Mark Hagland
| Reprints
Dr. Ben Schanker shared his perspectives on the future of blockchain in U.S. healthcare

On Wednesday morning at the Florida Health IT Summit, sponsored by Healthcare Informatics, and being held at the Hilton St. Petersburg Bayfront in St. Petersburg, Benjamin D. Schanker, M.D., M.P.H., director of the American Board of Medical Quality and a practicing physician at the Zuckerberg San Francisco General Hospital and Trauma Center, as an employee of the Department of Public Health of San Francisco.

Dr. Schanker’s morning keynote address, “Blockchain and the Future of Digital Health and the Clinical Experience,” covered a range of questions and issues around that timely topic. He began by giving his audience a very broad overview of blockchain itself, explaining the concept of blockchain, some of its uses in other industries, and its current, still-emergent state in U.S. healthcare.

Moving into the subject of the use of data in healthcare, Schanker noted that “Data is overwhelming in healthcare. Deloitte published a study that found that it takes a physician 50 minutes per patient to adequately review a single patient medical record. I know as a clinician that I don’t have time for that. In healthcare,” he said, “we’re very good at putting data into [information] systems, but very bad at taking it out of them.” One of the initiatives he’s involved in at the Institute for Human Optimization/the Precision Medicine Research Group, he noted, is that “We’re looking at tying together personalized medicine and broad research around populations.”

Moving onto some specifics about Bitcoin, Schanker asked, “Is Bitcoin really the promised land of healthcare? Will it solve all our problems? I say, hold your horses.” After explaining the fundamental theories behind the use of blockchain—it is a distributed, decentralized ledger database that uses “blocks” of data, linked together in a chain,” and accessed by a peer-to-peer network of equal partners—he went on emphasize that the “immutable distributed database” undergirding any chain “engenders trust. When Alice and Bob want to share information,” he said, referring to two archetypal individuals involved in a blockchain community, “they can share that information with other parties who are on the chain. The data is immutable. It can’t be changed, because a number of different parties have validated it.” And, importantly, any new block of data added to the ledger must be validated by a consensus protocol, typically meaning 51 percent of the parties involved in that particular chain. “There is no single point of failure,” he emphasized. “If Alice falls off the face of the earth, there are still multiple copies in existence. In practice, this means having a backup of your database.”

All of those elements hold major implications for the adoption of blockchain in healthcare, Schanker noted. On the one hand, he conceded, “Blockchain is very difficult to implement, because it’s a democratic database. Oftentimes, there are hierarchies in organizations, and they are necessary to make things happen.” And that factor works against blockchain adoption in healthcare.


Benjamin Schanker, M.D., at the Health IT Summit

Also, there are many types of data in healthcare that are simply too big or complex to be used well in healthcare. For example, he said, “You would never want to put an MR”—magnetic resonance scan—“on a blockchain, because you’d have massive data that would have to be validated.” Instead, in many cases, clinical data in healthcare might be incorporated through the use of digital pointers, which can lead trusted parties to the digital locations of such data.

Still, some organizations are beginning to experiment with blockchain technology for some niched purposes. One project that he himself is involved in is “CareCoin, one I’m working on at UCSF,” Schanker noted. “It’s designed to incentivize patients to ‘act better.’ It is a tool to align incentives among doctors and patients, where both parties are incented to work together.” It provides rewards to both physicians and their patients for engaging in certain behaviors.

Elsewhere, Schanker noted, “In South Korea, there’s a project called MediLedger, in which they’re using blockchain for medication supply chain purposes, validating data from the pharmaceutical producer to the manufacturer, to the distributor. MediLedger shares information across that supply chain of medications among the parties.”

He also noted the development taking place of something called the Robomed Network, which is described on its website as “a revolutionary medical blockchain project connecting healthcare providers and patients via smart contracts and ensuring output-based approach in the relationship. Launched in Dubai and Russia, this is a great step towards value-based healthcare,” the Network’s website states.

He further referenced SolveCare, being sponsored by the government of Estonia, and Universal Health Coin, a U.S.-based initiative.

“We should focus less on who owns the data, and more on how it is used,” Schanker emphasized to his audience. He also noted that “We’ve got financial, social, professional, and spiritual incentives, in everything we do. We should focus on the non-financial incentives.” And, in that regard, he said, “We’ve got a lot of both ‘carrot’ and ‘stick’ incentives in healthcare. I don’t think one or the other is better,” he opined. “Both are tools that can be used as incentive mechanisms. When we think about personalized medicine and population health, those concepts seemed contrary to me. But I’m going to quote Adam Smith, he said, and then went on to quote a passage from The Wealth of Nations, the seminal 1776 book on macroeconomics by that author: “By pursuing his own interest he frequently promotes that of the society more effectually than when he really intends to promote it.”

Afterwards, Dr. Schanker spoke with Healthcare Informatics Editor-in-Chief Mark Hagland. Below are excerpts from that interview.

What should the CIOs, CMIOs, and other healthcare IT leaders of hospitals, medical groups, and health systems, be thinking about in all of this?

Version control is most important. You have an existing version of clinical data. You can keep that copy and add a specific piece to it that can always revert back if needed. So overnight, as databases are updated, you can add pieces to it that can be reverted back. As you’re iteratively experimenting with blockchain, you always have the option to revert back to a prior version.

Do you see any potential related to electronic health records [EHRs] in this?

Yes. Access control to patient records is a phenomenal blockchain use case.

In other words, one might wrap clinical data inside a blockchain? How would that work?

An example initially would be empowering patients to control who accesses their records and for them to see an audit trail of when clinicians access their records. That feedback loop of patients being able to see when clinicians access their records could be support the patient-clinician relationship and empower patients.

Since patient records are so large in terms of the data involved, might that involve more ‘portable’ pieces of data, such as the most recent notes and updates on a patient?

Yes, a basic patient synopsis that the patient can control and update.

That could be better than the PHR [personal health record] as it now exists?

Yes, that’s really the use case. EHRs in this day and age are meant for clinicians to communicate with each other. But we need to communicate better with patient and empower them to have control of their records, and to facilitate that process. The single most under-utilized resource in HC is the patient. The appeal of a PHR is lacking, and things like blockchain can excite people.

Will physicians really want to participate?

I think if the clinician is asked to do anything additional that involves taking any active steps, it will be a challenge to get them to participate. On the other hand, anything that’s going to streamline their workflow is a viable target. So if a patient is answering digital questions on an in intake, or in their PHR, and it’s auto-populating that PHR and it’s partially writing a note for them, physicians would be interested. It’s a matter of optimizing the clinical workflow.

So one might be potentially creating a separate piece, then, involving one element of the overall patient record?

Yes, that’s right. One area with great potential involves the patient intake process. To do the digital intake through a blockchain-based record system, has significant potential. I don’t know whether Epic or Cerner is trying to do it, but the current process of patient intake is one of the most inefficient processes that exists in healthcare. We do paper intake and then manually input the data to make it electronic. It’s a very wasteful and inefficient process, and there’s real potential for blockchain to be adopted in order to improve it.

 

 

 

 

 


Related Insights For: IT Applications

/news-item/it-applications/survey-more-effective-it-needed-improve-patient-safety

Survey: More Effective IT Needed to Improve Patient Safety

July 19, 2018
by Heather Landi
| Reprints

Nearly nine out of 10 respondents to a national survey of physicians, nurses and healthcare executives say their organizations are successfully improving the safety of patients, yet these medical professionals also identified ineffective information technology as a key barrier to achieving their patient safety goals, according to a recent online survey conducted by Health Catalyst.

Specifically, respondents said they need better health information technology to warn clinicians of impending patient harm, as well as more resources and greater organizational focus on the problem, the survey found. Health Catalyst, a Salt Lake City-based data analytics company, conducted an online survey of 462 medical, quality and pharmacy professionals in healthcare organizations of all sizes across the U.S. on patient safety issues.

There is compelling evidence that the healthcare industry needs to do more to address patient safety issues. Medical error is one of the leading causes of death in the U.S. While the morality statistics are alarming, non-lethal harm events are even more frequent, occurring at a rate 10 to 20 times higher than lethal events, according to a study in The Journal of Patient Safety.

Despite the evident room for improvement, confidence in current patient safety efforts is high, according to the survey. Seventy-nine percent of survey respondents rated their organizations’ success in improving patient safety either “somewhat good” or “very good.” Only 11 percent rated their patient safety efforts as “poor.” However, on the opposite end of the scale, just 9 percent gave their efforts an “excellent” grade. 

The survey results indicate that serious challenges prevent healthcare organizations from making a significant dent in preventable errors. Respondents identified several key obstacles that prevent them from achieving their patient safety goals:

  • “Ineffective information technology (data quality, patient matching, reporting)” and the related “lack of real-time warnings for possible harm events,” which requires technology – 30 percent
  • “Lack of resources” including staffing and budget – 27 percent
  • “Organization structure, culture or priorities” – 19 percent
  • “Lack of reimbursement for safety initiatives” – 10 percent
  • “Changes in patient population and practice setting” – 9 percent
  • “Other” – 6 percent

Organizations’ lack of effective information technology for patient safety is tied to a related finding from the survey – that healthcare organizations of all types are almost completely dependent on manual methods of tracking and reporting safety events. According to the survey, the four most common sources of data used for patient safety initiatives are voluntary reporting (selected by 82 percent of respondents), hospital-acquired infection surveys (67 percent), manual audits (58 percent), and retrospective coding (29 percent). Nearly one-third of respondents (28 percent) reported also using trigger tools as a data source for patient safety, which could mean either the manual process of chart review that relies on Institute for Healthcare Improvement (IHI) methodology, or home-grown reports that also follow the IHI methodology. 

According to a study published in the Journal of Patient Safety, these standard approaches to manual reporting of hospital safety events have been shown to find less than 5 percent of all-cause harm. Manual reporting is based on data that is at least 30 days old, and it requires extensive time and resources for data extraction, aggregation, and reporting, resulting in limited root-cause analyses, the Health Catalyst survey notes.

“As these survey results confirm, the current approach to using voluntary reporting to monitor patient safety gives health care organizations a false sense of tackling the ever-present danger of patient harm,” Stanley Pestotnik, Health Catalyst’s vice president of patient safety products, said in a statement. “Recent evidence continues to demonstrate that the majority of patient harm goes undetected and that medical injury is the third leading cause of death in the US—evidence that challenges voluntary reporting as an effective patient safety management strategy.”

Examining the factors that are most influential in driving organizations’ patient safety efforts, a majority of survey respondents (51 percent) named regulatory reporting as an influencing factor. Coming in second at 39 percent was “financial considerations” such as malpractice claims, value-based contracts and reduced reimbursement. Other choices included published accreditations and designations (34 percent); patient satisfaction scores (33 percent); data-driven organizational priorities (29 percent); performance against safety measures (27 percent); brand recognition, market competition (16 percent); and stakeholder interests (11 percent).

When asked to identify the areas where patient safety most needs improvement, survey takers rated four of the six choices within 3 points of each other. “Inpatient clinical” areas of focus such as length-of-stay, mortality and readmissions came out on top at 21.6 percent, barely ahead of “operations” (21.1 percent), an area that includes ED wait times and patient instructions at discharge. Two other areas most in need of improvement, according to survey takers, were “severity of illness” (19.5 percent), “outpatient/ambulatory clinical” (18.6 percent). 

Only “regulatory reporting,” including reporting of hospital-acquired conditions, seemed to require slightly less improvement than other areas, with 15.5 percent of respondents citing it. 

“The big picture takeaway from this survey is that although a small portion of respondents felt they have a good handle on their patient safety efforts, the largest portion of respondents still believe that they have room for improvement,” Valere Lemon, R.N., a senior subject matter expert for Health Catalyst, said in a statement. “Surveilling all-cause harm will aid healthcare organizations in bridging the gap from niche focused improvements to proactive harm identification and broader patient safety improvement interventions.”

   

 

See more on IT Applications