Linda Wilson

home

authors

linda-wilson

To help its members prevent or manage diabetes, Blue Cross and Blue Shield of Minnesota opted for a virtual solution.  

The Minnesota health insurer has partnered with Omada Health, based in San Francisco, to offer technology-enabled disease education and management programs to its members with diabetes or prediabetes. Omada Health has separate programs for each disease that are based on nationally recognized training and education programs.

The Omada programs feature live coaches who help patients improve their biometric data, such as weight or A1C levels. Using data analytics to parse data collected through connected devices, including scales and glucose monitors, Omada Health derives a personalized program for each participant. Using a digital app on their smartphones, members connect with their coaches through texting or virtual visits. 

The intimacy of a personal coach and the convenience of virtual care appeal to Blue Cross members, explains Jen Benner, clinical product manager at the Minnesota Blues plan, whose parent company has invested in Omada. The pairing of a coach and a virtual care is especially important for individuals with diabetes, she adds “because they have so many decisions to make every day.”

What Will Be the Most Costly Medical Conditions In 10 Years?

Omada Health is not alone in selling technology-enabled programs to help peopleprevent or manage diabetes. These programs are moving quickly from being novelty items for early adopters into mainstream use. The move from the fringe is something to encourage, says Michael Sturmer, senior vice president of health services for Livongo, a publicly traded disease management company in Mountain View, California. 

“What we are really focused on with our health plan partners and our PBMs is really changing their mindset from digital health as this add-on or buy-up to digital health as part of the core benefits that a member expects to get when they sign up for their health plan,” says Sturmer.

Virtual, tech-enabled programs for diabetes are catching on for any number of reasons. First, there is prevalence of diabetes and the associated expense. The American Diabetes Association says about 30 million Americans have diabetes and estimated that in 2017, an American diagnosed with the disease incurred, on average, $16,750 per year in medical expenses. Second, the programs seem to be producing promising results. Independent studies are few and far between, but the vendors have released their own study results showing weight loss and improved A1C levels. 

Virta Health, a San Francisco company that says its goal is not just to manage type 2 diabetes but reverse it, uses technology to help participants limit carbohydrate intake. The company released results that it claims show that its technology reversed type 2 diabetes in 60% of its patients. This means the patients’ blood glucose readings are in a normal range, without diabetic-specific medications. For some of its earliest patients, those results have been sustained for four years, according to CEO and co-founder Sami Inkinen.

Blue Cross and Blue Shield of Minnesota began using Omada’s diabetes-prevention program in 2015. The program for individuals with full-blown type 2 diabetes was adopted in 2019.

The prevention program includes intensive coaching for weight loss, goal setting, and an online peer-support group. It is designed to last 16 weeks, but members can remain engaged in the program for as long as they wish, say Benner, the Minnesota Blues clinical products manager. Participants also get a wireless digital scale that sends weight readings to Omada’s database. 

Benner says the cellular feature of the scale is important because not all consumers have internet access at home. Since 2015, 37,000 members of Blue Cross and Blue Shield of Minnesota have enrolled in the virtual program and lost a combined 200,000 pounds, an average of 5.4 pounds per person. 

Intermountain Healthcare, which also has invested in Omada Health, has seen similar results. The Salt Lake City health system began offering Omada’s diabetes prevention program to its employees in 2019. From January 1 through November 30, 2019, 1,533 Intermountain employees lost a total of 9,800 pounds, an average of 6.4 pounds per person. 

“A 5% weight loss is associated with a nearly 60% reduced risk in developing type 2 diabetes. A 10% weight loss from baseline - 20 pounds for a 200-pound person - is associated with an 85% reduction in the likelihood of developing type 2 diabetes,” says Elizabeth Joy, MD, senior medical director for wellness and nutrition at Intermountain. However, she was referencing research findings for in-person diabetes prevention programs, not tech-enabled plans like Omada’s.

“Type 2 diabetes is largely preventable for the vast majority of people by making lifestyle changes that result in relatively minimal weight loss,” Joy continues. “That is a really important point to make, because we get saturated with TV shows like ‘The Biggest Loser’ that basically say the only weight loss that is meaningful is a 50- to 100-pound weight loss. Clinically meaningful weight loss is 3% to 5% from baseline.”

The potential market for tech-enabled diabetes management could be gigantic because of the number of people with diabetes and relative lack of alternative offerings. In a 2014 study, for example, the CDC found that only 6.8% of commercially insured, newly diagnosed patients received formal diabetes education. 

Blue Cross and Blue Shield of Kansas City offers its members access to a diabetes management program from Livongo. The insurer, known as Blue KC, launched a pilot program in 2016 and made the Livongo program available to all of its members during the fourth quarter of 2019. As of January 2020, Blue KC had enrolled 1,237 members in Livongo’s diabetes management program, according to Tim O’Brien, vice president and executive director of care delivery at Blue KC.

Blue KC plans to integrate patient data generated by Livongo’s platform into the electronic health records at Spira Care, which is Blue KC’s combined health insurance and primary care plan. Spira’s seven care centers provide members with access to primary care and behavioral health providers, health coaches, generic prescriptions, laboratory services, and digital X-rays. 

Livongo says its customers, which include self-insured employers as well as health plans, yield savings. For example, Livongo compared health claims data for 3,664 participants in its diabetes management program with data from 12,065 people with diabetes who were not enrolled in Livongo’s program. Livongo found that participants in its diabetes management program netted a mean reduction of A1C ranging from 7.8% to 6.9% and a 5.8% decrease in the cost of healthcare claims. 

Some of the tech-enabled programs offer value-based pricing. Virta bases its charges on clinical outcomes. “If we hit certain thresholds, we get paid, and if not, we don’t get paid,” says Inkinen. The metrics Virta uses are improvements in blood sugar control, reductions in medications, and weight loss. The company bills customers monthly for patient cases meeting specified thresholds. 

“To me, that is how healthcare should be,” says Inkinen. “We shouldn’t be here to do more transactions. We should be here to make the patients better.”  

Linda Wilson is an experienced writer and editor specializing in the healthcare industry.   

Articles

Several Blues plans have contracted with vendors that supply tech-enabled diabetes disease management and prevention programs. Sometimes the tech is combined with coaching - by a real, live human being.

21st Century-Style Diabetes Disease Management: Coaches and Tech-Enabled

The top healthcare technologies for 2020 aren’t new, but they are on the cusp of moving into broader use.

In a sign of the growing maturity of some of these technologies-such as machine learning, voice recognition, and natural language processing-vendors are combining them to build solutions that solve real-world problems for providers and insurers.

That’s because the focus of digitally enabled healthcare “is less about the technologies to be used and more about doing things differently,” says John Bosco, chief technology officer at Northwell Health, a 23-hospital health system based in New Hyde Park, New York.

Here’s the list of technologies behind the transformation of business and clinical processes in 2020:

Voice-to-text solutions are emerging to help physicians and other providers cut down on the amount of time they spend documenting patient care.

“The ability to use voice to automate documentation for clinicians and reduce the administrative burden and use of manual scribes is something that has been in research and development but is now being more widely deployed across the United States,” says Brian Kalis, managing director of digital health and innovation at Accenture, a consulting, technology and outsourcing firm based in based in Dublin, Ireland.

The players in the market include established voice-recognition vendors, large technology companies, and startups.

3 Technologies Providing Personalized Healthcare

For example, Suki combines voice recognition, natural language processing, and machine learning to create a documentation solution that runs on laptops, and Android and iOS devices.  The Redwood City, California-based vendor claims that its software decreases the amount of time physicians spend on a clinical note by 76%, or from 13 minutes to three minutes.

Suki also claims that it’s integrated with the following electronic health records vendors: athenahealth, Watertown, Massachusetts; Cerner, Kansas City; eClinicalWorks, Westborough, Massachusetts; and Epic, Verona, Wisconsin, according to KLAS, a research and analysis company based in Orem, Utah, and focused on health information technology.

Suki rival Notable Health-based in San Mateo, California-has developed a workflow solution that combines patient intake and clinical documentation. Patients complete tasks such as registering and completing medical histories using their smartphones, while providers use digitally enabled watches to dictate visit summaries.

Chicago-based CommonSpirit Health, a 142-hospital Catholic health system, announced plans in 2019 to pilot Notable Health’s platform at some primary-care centers in California, with a goal of implementing the solution broadly across the organization.

Two other startup vendors of voice-to-text clinical documentation software are Robin Healthcare, Austin, Texas, and Saykara, Seattle, KLAS said in a 2019 blog post.

Legacy voice-recognition vendors also are developing intelligent voice-enabled documentation solutions. For example, Nuance Communications-based in Burlington, Massachusetts-and Microsoft-based in Redmond, Washington-are partnering to create a product to compete with the offerings from the startups. They hope to begin pilot testing their solution in 2020.

Meanwhile, Vanderbilt University Medical Center, an academic medical center based in Nashville, is combining voice recognition and natural language processing to create a solution that helps clinicians look up information in the electronic medical record.

Called Vanderbilt Enhanced Voice Assistant, the application “does what your phone does now. The idea is to ask a question of the health record,” explains Kevin Johnson, MD, Cornelius Vanderbilt professor and chair of the department of biomedical informatics and professor of pediatrics.

For example, providers could ask about their schedule of patients for the day or information about specific patients, such as their weight or medications. Johnson says providers could even ask a question such as, “Is there anything that I need to do today that I left in my notes from the last visit?”

“What we are trying to enable you to do is to have that conversation without having to stare at a computer screen,” he says.

The academic medical center, based in Nashville, Tennessee, plans to pilot the application in pediatric endocrinology in 2020 and then expand its use to other departments.

Mountain View, California-based Google also is investing in solutions that ease providers’ workflow burden, in part, through the use of natural language processing.

In a blog post in November 2019, Google Health CEO David Feinberg, MD, said Ascension-a 150-hospital Catholic health system headquartered in St. Louis-is piloting a tool that uses Google’s finely honed search capabilities to help providers extract useful information from patients’ medical records. The idea is “to make health records more useful, more accessible and more searchable by pulling them into a single, easy-to-use interface for doctors,” Feinberg wrote.

Google is not alone. Accenture’s Kalis says that health systems and insurers also are using natural language processing to improve their interactions with consumers.

For example, he says, WellCare Health Plans, headquartered in Tampa, Florida, has been using a product from Sensentia that uses natural language processing to help call center employees search through insurance documentation using plain English to answer enrollees’ questions about health insurance benefits. Based in San Francisco, Sensentia uses the same technology for a product that allows enrollees to look up benefits information online by themselves.

On the health system side of the industry, one example is an online symptom checker at Milwaukee-based Froedert & the Medical College of Wisconsin health network, a five-hospital network and academic medical center, developed using natural language processing and machine learning. Called Buoy Health, the interactive tool asks questions to help guide consumers to the appropriate medical care.   

Machine learning will continue to be adopted in healthcare in 2020, focusing on “back office” types of applications, Johnson says. He describes these “as technologies that are designed to improve the overall throughput of healthcare either by predicting patients who have specific risks, like readmission risks or length of stay risks, or by potentially doing things like decreasing denial days or improving overall compliance with billing.”

Vanderbilt is currently testing two risk-prediction tools-one for sepsis and another for suicide-by running them in the background. What this means is that the tools assign risk scores to patients and then Vanderbilt’s staff compares the tool’s predictions to what actually happens to patients.

If the results are promising for either tool, the next step would be a randomized controlled trial involving a clinical intervention, Johnson says. Without an intervention that mitigates the risk effectively, a tool isn’t useful, he says. “There are multiple steps to do high-quality work in precision analytics.”

“At Vanderbilt we believe very strongly in the learning health system, which allows data to be used for knowledge generation, knowledge to be used to change action, and action to be used to be used to generate more data. That is the cycle,” Johnson says.

Gregg Pessin, analyst and senior director for healthcare delivery at Gartner, a research firm based in Stamford, Connecticut, agrees. He says machine learning won’t be widely deployed until the healthcare industry addresses two issues. First, the algorithms require access to large, verifiable sets of training data. Second, the AI engines need time to mature because their accuracy improves as they continue to learn by ingesting new data.

Digitally connected devices-known collectively as the internet of things, or IoT-is a promising source of the large-scale data sets needed to drive machine learning algorithms.

Many of these medical-grade digital devices-such as blood pressure cuffs and glucometers-are already recording patient data and transmitting the information to clinicians.

Pessin says the next step involves collecting and analyzing data from consumer devices such as digital watches and fitness trackers.

“In 2020, the Internet of Things itself will pass the tipping point for healthcare, and we will see that it becomes ubiquitous. There won't be a hospital, at least in the United States or the Western World, that isn't fully invested in IoT by the end of the year,” Pessin says.

In addition to incorporating data from consumer devices, Northwell Health’s Bosco says health systems are beginning to investigate how to move from intermittent data transmission from IoT devices to continuous patient monitoring.

Kalis says the movement toward continuous patient monitoring has been on the horizon for a number of years, but it is gaining some traction now because of the lower price points of the devices and sensors and the emergence of 5G on wireless networks.

Nonetheless, Bosco says there are three issues the industry needs to resolve before health systems adopt continuous patient monitoring at scale. He says that healthcare organizations need to implement the following:

Technologies that can ingest and store vast amounts of real-time data.

 AI tools to analyze and interpret the information.

 Clinical interventions aimed at improving patient care based on the information generated by the AI algorithms.

Bosco says continuous patient monitoring “holds a lot of promise for driving down costs and improving the quality of the care that we are giving,” particularly in-home care. “In general, the home is the most convenient, most safe and least costly venue of care,“ Bosco says.

Robotic process automation-software that carries out high-volume, error-prone repetitive tasks-has been deployed in manufacturing, financial services and other industries. Now, some of those vendors are adapting their solutions for the healthcare arena.

One promising area involves automating repetitive tasks in the revenue cycle such as claims management, explains Doug Tolley, vice president of business development at KLAS.

Some examples: Automation Anywhere, is a vendor of robotic automation tools, located in San Jose, California; UiPath, New York, a vendor or robotic automation tools; and Blue Prism, a vendor of robotic automation tools, located in Warrington, United Kingdom.

Olive, headquartered in Columbus, Ohio, is slightly different because it was founded in 2012 specifically to develop solutions for the healthcare arena. KLAS, which interviewed five of  

Olive’s customers for a recent report, says the vendor’s solution automates such tasks as verifying insurance eligibility checks, pre-authorization requests, and patient registration.

“I think it'll be far broader than just the rev cycle when it's really deployed into healthcare, but it started with some of the simpler areas of rev cycle,” Tolley says. “However, we haven't been able to yet, at KLAS, validate very many use cases outside of automating some of things in rev cycle,” he says. “We're looking for those test cases.”

With robotic process automation as well as the other top technologies for 2020, the real promise lies in their potential to help humans transform workflows, improving quality and reducing costs in healthcare. As Pessin explains, “We've been doing healthcare the same way for decades now, and we're on the precipice of a new way of delivering care. We're not in the new era, but we're just kind of walking through the doorway.”

Linda Wilson is an experienced writer and editor specializing in the healthcare industry.

The top technologies for 2020 aren’t new, but they are on the cusp of moving into broader use in healthcare.

Top 5 Healthcare Technologies

After analyzing data about risk of death and emergency department usage for its cancer patients, Northwest Medical Specialties opened a walk-in clinic.

A nurse practitioner staffs the clinic, which is open from 9 a.m. to 5 p.m. on weekdays. The goal is to treat the acute symptoms that cancer patients experience-such as pain, dizziness, and diarrhea-and avoid trips to the emergency department.

The practice-with six locations in the Tacoma, Washington area-plans to hire a second advanced practice provider to staff the acute-care service at a second location.

Before launching the clinic, the practice conducted a six-month pilot, demonstrating sufficient patient demand and satisfaction with the service to justify the cost.

“This was an example of a practice transformation-of how this practice took information and made a dramatic change in the clinic,” says Sibel Blau, MD, medical director, hematology and oncology divisions at the 10-physican practice. The overall goal, she says, is to position the practice for success under value-based reimbursement models.

Northwest Medical Specialties is not alone. As value-based payment arrangements become more common, community cancer providers are installing data-analysis tools to help them figure out how to meet cost and quality targets.

Can Data Analytics Lower Hospital-Acquired Condition Incidence?

Payers and providers are focusing on improving the quality and efficiency of cancer care because of the disease’s prevalence and treatment costs. The American Cancer Society projects that there will be more than 1.7 million new cancer cases and 606,880 cancer deaths in the United States in 2019.

The costs to treat those patients keep rising. The National Institutes of Health estimates that the United States will spend at least $158 billion on cancer care in 2020, a 27% increase over expenditures in 2010.

Because a significant proportion of those patients are Medicare beneficiaries, CMS developed a five-year pilot project, the Oncology Care Model (OCM), which began on July 1, 2016. Nearly 200 physician groups and 17 commercial payers are participating in the program, according to CMS.

CMS reimburses providers based on their performance during six-month episodes of care for chemotherapy treatment using either a one-sided or a two-sided risk model.

West Palm Beach, Florida-based Integra Connect, which develops information technology solutions for healthcare providers, has tracked the performance of its customers involved in the OCM program, which totals about 16 practices and 1,000 oncologists. “What we see is a steady improvement,” says Charles Saunders, MD, CEO, Integra Connect. With each successive performance period, more of the practices have kept their actual costs of care under the target cost set by CMS, he says.

Nonetheless, Saunders says some of his clients did not earn a bonus payment during any of four performance periods, ending with the first half of 2018. Those practices would have decided by December 2019 whether to accept two-sided risk, or drop out of the program.

In cancer care, the largest buckets of costs are chemotherapy and other medications; facility services, such as hospitals and skilled nursing facilities; and care at the end of life, says Saunders, who regularly writes and speaks about value-based cancer treatment.

CMS isn’t the only payer applying value-based care models to cancer treatment. In April, Humana announced its value-based program for oncology in which it will reward oncology practices for exceeding quality and cost benchmarks over a one-year period.

Like Northwest Medical Specialties, Texas Oncology-a large cancer treatment and research provider in Texas specializing in hematology, pediatric and radiation oncology, which also participates in CMS’ OCM-has worked to decrease rates of hospital admissions and emergency department visits.

For example, the oncology practice analyzed data on telephone calls and response times for patients reporting acute medical issues. What Texas Oncology learned was that when patients called, they usually left a voicemail but often for the wrong department. The inefficient approach made it difficult for staff to return calls quickly.

Call back times are “a huge barrier,” says Lalan Wilfong, MD, executive vice president of value-based care and quality programs, Texas Oncology. If patients don’t receive a timely return call, they go the emergency department for help, he says.

To solve the problem, Texas Oncology, with more than 460 physicians and 210 locations, is in the midst of implementing an electronic triage platform from Navigating Care.

This is how the new process works: When a patient calls, a locally based operator working with an electronic dashboard determines which department should respond to the call. If it is a nursing issue, the operator asks targeted questions to triage the patient. “Somebody who calls in with a fever of 102 [degrees] is a higher risk than somebody who calls in because his or her toe hurts,” Wilfong says.

When dedicated triage nurses return calls, they also have access to the pre-screened patient information on an electronic dashboard, Wilfong says. Because the tool is electronic, nursing managers monitor the dashboards, and assign additional staff to return calls on unusually busy days, he adds.

The objectives of the electronic triage program are straight-forward: A “live person” answers the calls. When the call involves an acute medical issue, triage nurses call the patient back within 30 minutes, Wilfong says.

In addition to improving acute symptom management, oncology practices nationally are also using data analytics to improve their care coordination programs.

Northwest Medical Specialties uses a machine learning­–enabled tool from Jvion to risk stratify patients. The tool produces lists of patients at risk of clinical deterioration on seven vectors, including 30-day mortality and avoidable hospital admissions. The tool also details the clinical and non-clinical factors driving the risk and recommended actions to mitigate the risk.

By feeding data from its electronic health record into the tool, Northwest Medical Specialties gets a daily list of patients who score high on at least five of the seven vectors. A care coordinator works with those patients to schedule the appropriate services, such as evaluation and treatment for pain, distress, or depression; completing advance directives; or referral to hospice care.

By risk stratifying patients and acting on the information, Northwest Medical Specialties has increased referrals for palliative and hospice care by 81%, depression screenings by 68%, and case management evaluations by 80%.

In addition to heading off adverse events before they occur, it’s also important to quickly find out when patients visit an emergency department or are admitted to a hospital or skilled nursing facility.

To gain reliable access to that type of information, Regional Cancer Care Associates recently implemented software from PatientPing, which electronically notifies the 34-location cancer provider when one of its patients receives care outside of its facilities in New Jersey, Maryland, and Connecticut.

Using the software, staff at Regional Cancer Care Associates can also drill down to find out “the working diagnosis” for each of its patients, says Lani Alison, vice president, clinical affairs, Regional Cancer Care Associates.

The tool also allows Alison to look for opportunities to improve performance by analyzing aggregated data, such as how often a patient has been admitted to a hospital or how many of a doctor’s patients have been hospitalized.

Before implementing PatientPing, gaining access to such timely data was haphazard, manually driven, and didn’t lend itself to data analytics. As Alison says, “It was clunky.”

Linda Wilson is an experienced writer and editor specializing in the healthcare industry.

Risk-stratification and electronic triage programs are improving patient care and decreasing costs for health systems.

Data Analytics Improving Care Coordination Programs  

From the adoption of value-based, consumer-centric care to the arrival of new industry players, such as retailers, change seems constant in the healthcare industry. That’s why it’s more important than ever for organizations to deploy technology wisely to support their business vision and strategy.

The information technology (IT) operation “needs to move from a cost center to more of a value center,” says Subra Sripada, managing director in the healthcare practice at Navigant Consulting. “Simply said, the IT tactical plan needs to align closely with the strategic plan,” he says.

While this is not a new concept, achieving it isn’t easy. So Sripada, former chief information officer and chief transformation officer at Beaumont Health in Southeast Michigan, and other experts share eight recommendations about how to align the work of information technology with the organization’s strategy.  

1. Create a structured governance process

At FHN-a rural health system based in Freeport, Illinois-an information technology governance group composed of C-suite executives meets throughout the year to make sure the IT department’s portfolio of projects supports the organization’s strategy to become a consumer-centric health network. “It really needs to be owned by the IT governance council because the IT department needs the support of senior executives to really get what they need: the resources to get things done,” explains Mark Gridley, president and CEO of FHN, which includes a 100-bed hospital.

Carol Chouinard, vice president of provider technology at Optum Advisory Services, says regular governance meetings provide a venue for C-suite executives to learn about the work of the IT department throughout the year. “Too many CIOs go through this decision-making process, and then they walk away. Then there are these surprises and frustrations the next time they get to the table,” he says.

2. Adopt project management for the entire enterprise

While many healthcare organizations use a project management process within the IT department to vet, manage and track projects, fewer of them have adopted a process that encompasses the entire organization.

But that’s what executives at Albany Medical Center in New York decided to do earlier this year. In March, the academic medical center launched an enterprise program management office. The office coordinates and tracks all initiatives from departments across the organization, such as IT, facilities and clinical service lines, as well as cross-functional projects.  

Projects are completed on a day-to-day perspective at the department level, but the project details from IT and other departments also are fed into the enterprise program management process. “It is completely bi-directional,” says George Hickman, executive vice president, CIO, and chief analytics officer at Albany Medical Center.

Mayo Clinic, Google to Form Healthcare Partnership

“We can see all the interrelationships.” Hickman continues. “Frankly, it makes the prioritization process transparent and fundamentally clear to all of senior management.” Because many projects compete for the same people and financial resources, centralized project management helps executives balance competing priorities, he notes. 

A team from information technology and finance manages the enterprise program management office. The team reports to a leadership group that oversees Albany Medical Center’s overall strategic plan-an effort the organization calls “Pillars 22,” Hickman says. 

During the annual budget process, CIOs should help other members of the C-suite decide how to deploy technology to fix business problems. They should also provide their peers with insight about the benefits and risks of each option.

As Gridley explains, “For myself as the CEO, I may have a vision for where I think things need to be, but my IT team is the one that needs to help frame the vision.”  

For example, he says, FHN’s executives wanted to provide telehealth services as an alternative to in-office visits for patients. After evaluating the risks and benefits of developing a telehealth program in-house, FHN decided to contract with MDLive, which manages the virtual technology and remote providers but brands the service as an FHN entity. The new service has been live for several months.

“The mature CIOs will always anticipate and be able to manage through changes in the project portfolio. We are in an industry where things change,” Chouinard explains. “The most common is budget pressure and budget priority changes. The second is around M&A activity. The third is leadership changes. New folks come in and they have new priorities.”

For example, he says he worked with a large health system that pulled the plug on a Cerner electronic medical record deployment in one local market because a large, important medical group insisted that the organization’s local hospitals use Epic’s EMR.

Most IT departments need to improve how they manage budgeting, forecasting and spending revisions for the projects in their portfolios, Chouinard says. Even more important, they need to think about IT spending in a strategic way.  He explains the strategic thought process as, “What is the right level of affordability? How much can we do with the money we have?”

Sripada encourages his clients to hold town hall meetings to report back to their staff about the progress the IT department is making to support the organization’s strategic goals. “I’ve always felt that gives a sense of purpose to the IT team,” Sripada says.

Hickman says members of the data analytics team are assigned to Albany Medical Center’s process-improvement projects, so they become familiar with the organization’s corporate strategy through their day-to-day work.

Beyond that, Hickman says he includes an agenda item on Albany Medical Center’s strategy at IT department meetings. He also encourages information-sharing at meetings. “I try to create venues-whether I am there or not-where we go around the table and talk about what we are working on, and the approaches we are taking,” he says.

7. Align the IT project portfolio to the corporate culture.

Organizations have different appetites for the amount of risk they are willing to take on and the rate of change they are comfortable managing. CIOs should fully understand those cultural dynamics before embarking on large initiatives or changing the mix of existing projects, Sripada advises.

“Just because there are technologies available to solve a business problem doesn’t mean the organization is ready to adopt it and benefit from it,” he says.

During his last four years at Beaumont Health, Sripada oversaw the day-to-day activities involved in merging three health systems into one-a process that subjected employees to a rapid pace of change. In hindsight, “we are glad we did it the way we did it,” he says, while also saying that it was a challenging process. “There were a lot of balls that were in the air.”

The IT team needs to keep abreast of new technologies that could help the corporate organization reach its strategic goals, such as third-party apps for consumers. The department also needs to plan for changes to the underlying infrastructure, such as the electronic medical record, to support future IT investments.  

“It is so easy to get lost in the 400 requests that we get, and we have to manage on a regular basis,” Chouinard says. “There is a risk that we will lose sight of what the bigger picture is.”

is an experienced writer and editor specializing in the healthcare industry.

Ensuring your IT plan is aligned with the organization’s strategy can be critical to success. Here are eight things to keep in mind.

8 Ways to Optimize Your IT Operations

To take advantage of emerging software tools that incorporate artificial intelligence, healthcare organizations first need to overcome a variety of challenges.

Some leading-edge organizations are beginning to do just that, focusing on machine learning, a subset of artificial intelligence (AI) that encompasses statistical methods in which computer systems recognize patterns or correlations in data by ingesting large sets of training data. They improve their performance, or “learn,” over time as they incorporate new data; revising their approach as needed without human programmers updating the rules.

In the healthcare industry, most machine learning applications are in the research stage. “There is not a ton of clinical use,” according to Brian Edwards, independent validation consultant for AI vendors.

One area with a lot of research activity is 

, where the industry is investigating how to use 

machine learning to detect signs of disease from digital images

. “Wherever you have crisp clean data is where you should start. Images are the highest quality data that you have in a health system in terms of reliability,” Edwards says.

Machine learning has been applied to other areas, such as assessing patients’ risk of a hospital readmission, exacerbation of a chronic medical condition, or coming down with sepsis during a hospital stay.

“I think planning is where it starts,” says Bob Fuller, managing partner for healthcare at 

, an information technology consulting firm focused on data analytics. Fuller says healthcare organizations should assess their overall business strategy and how AI could be deployed to solve specific problems, such as hospital readmissions or claims fraud. The next step is to allocate financial resources to transition the information technology infrastructure, so it becomes AI-ready, Fuller says.

This includes a large and diverse set of reliable data to train the machine learning models-whether they are developed internally or purchased from AI software vendors.

Six Healthcare Technologies Coming in the Next 10 Years

“In every situation, it is necessary to train the AI,” Edwards says. “That knowledge is really institution specific. The heterogeneity of data makes it very difficult, if not impossible, to take something from one organization-or a generic packaged product-and implement it widely. It really is implemented one-at-a-time in an almost ala carte type of way, where you need to customize it,” Edwards says.

Another barrier to implementing machine learning in healthcare organizations is access to high-quality data. Healthcare organizations need to have rigorous processes in place to ensure they have clean and well-defined data, Fuller says. While this has always been true, it becomes even more important as the volume and types of data that healthcare organizations capture continues to grow, he adds.

-an integrated delivery network with 13 hospital campuses and a nearly 600,000-member health plan-has tapped into its vast store of diverse datasets to develop machine learning applications internally.

It has data on 2 million patients in its electronic health-records system. It also has a stable patient population in Pennsylvania and New Jersey, which allows it to build longitudinal datasets, spanning 20 years.

In its digital imaging system, it has two petabytes of data, which it accumulated over 19 years. Most of this data is from radiology, but some is from other medical disciplines, such as cardiology.

“The ability to have that data and to use if for machine learning is one of our strengths,” says Aalpen Patel, MD, chairman of radiology at Geisinger.

Some applications are in daily clinical use already, including a machine learning model that detects intracranial hemorrhage, or bleeding in the brain, from CT scans. Geisinger uses the model in daily operations to read CT scans of the brain taken in the outpatient setting. If the algorithm detects bleeding, the case is automatically reprioritized as a STAT case in radiologists’ work queue.

While all CT brain scans from hospital units or emergency departments are considered STAT and read within 30 minutes, similar scans taken in the outpatient setting are read within 12 hours. By sending those cases from the outpatient setting through the machine learning model, Geisinger reduces diagnosis times significantly for critical cases coming from the outpatient setting.

“We don’t take on problems that we don’t think are relevant,” says Brandon Fornwalt, MD, PhD, chairman of the department of imaging science and innovation at Geisinger. The integrated delivery network’s goal is to move from the research phase to implementation in clinical workflows quickly, he says.

In addition to large datasets, ample processing speed is necessary for machine learning. Geisinger solved this problem for its first machine learning application in imaging by purchasing a GPU (graphics processing unit), which accelerates the processing of computational workloads. Since then, it has upgraded its architecture to include multiple GPUs because its need for processing power has grown.

Fuller says accessing cloud options-such as from Amazon Web Services, Microsoft Azure, or Google Cloud-might be more cost effective for many healthcare organizations than building the required infrastructure internally because the cloud allows you scale computing resources up and down to match current needs.

Real-World Applications of Artificial Intelligence in Healthcare

Having a plan to evaluate machine learning products on the market today also is important. Salt Lake City-based Intermountain Healthcare, which has 23 hospitals and more than 170 clinics focuses its efforts on purchasing commercially available software, has investigated many more products than it has purchased, according to Lonny Northrup, senior health informaticist at Intermountain.

Northrup says Intermountain typically won’t move forward with a tool unless the vendor can point to clinical improvements and cost reductions that occurred at another health system.

After analyzing the results the platform achieved elsewhere, Intermountain assesses how well a product works with Intermountain’s patient data. To do this, the health system feeds the product a set of training data-such as on colon surgery-that the health system has already studied, allowing it to verify the insights that the tool derives.

The next step is a pilot test. For example, Intermountain is working with a vendor of patient engagement software on a clinical trial using the product to encourage patients with complex cases of congestive heart failure to follow their medication regimens. Using machine learning methods, the software platform personalizes the recommendations it makes about how to prod patients to behave in ways that improve their health. As part of the project, Intermountain provides 24/7 availability of clinical personnel to respond to these patients’ needs, Northrup says.

No longer a pipe dream, AI is already in use. Here are two real-world examples of machine learning in healthcare.