On the third and final day of HAS 21 Virtual, Paul Horstmeier welcomed attendees to the Hotel Catalyst Dubai. This incredible virtual environment was modeled after the Burj Al Arab and included a massive Dale Chihuly glass sculpture and an aquarium. Attendees also got to peek behind the scenes and meet some of the crew who produced this year’s event—no small undertaking!
Horstmeier shared the rest of the summit’s agenda, with a lineup once again filled with industry leaders, including Amy Compton-Phillips, who oversaw the care of the first known COVID-19 patient in the United States. The breakout sessions not only featured a robust set of topics, but also a special edition of Jeopardy focused on healthcare improvement and financial success, which Tom Burton hosted.
Additionally, the analyst team shared insights into the great attendee engagement they’ve seen throughout the summit. This includes 232 Braindates, with top topics such as All Things PowerBI, Data Science and Agile, and Beyond Burnout. Braindate opportunities were scheduled to continue through Friday, September 23.
Lastly, it wouldn’t be HAS 21 Virtual without #SocksofHAS. The competition this year was intense, with dogs, cats, babies, and some favorite snacks all making appearances. All entries were strong, and three deserving winners took the top prizes.
Amy Compton-Phillips, MD, President, Clinical Care at Providence St. Joseph Health
In January of 2020, Providence admitted the first-known U.S. COVID-19 patient. The health system soon realized they would need to carefully plan, leverage data and technology, and quickly scale solutions across the organization to respond. Over the next 18 months, Providence led a robust and multifaceted response. During her session, Amy Compton-Phillips shared several of Providence’s data-driven initiatives and explained how her organization uses data to accelerate the development of tools, infrastructure, and thinking.
As the pandemic unfolded, Providence worked continuously to stay ahead of potential issues. The organizations developed a chatbot to proactively answer questions from people concerned they might have contracted the virus before testing was widely available and resources were strained. Providence deployed a remote monitoring solution to allow patients who didn’t need hospitalization to recover safely at home, making hospital beds and critical care providers available for those who needed them most. Next, the Providence created a system to leverage its data, track virus progression, and use natural language processing to monitor reported symptoms to predict how many patients might eventually be hospitalized.
The above programs and other innovative initiatives enabled Providence to maximize its staff and facilities, deploy resources where they were needed most, ensure equitable delivery of the vaccine, and conduct real-time research by tapping into patient-reported outcomes.
Dan Unger, MBA, SVP and General Manager, Financial Transformation Business
There is no “if” when it comes to COVID-19’s impact on the healthcare industry, only “how,” according to Dan Unger. Unger discussed how the pandemic fueled unprecedented financial headwinds. New competitors—including vertically integrated insurers, innovative care delivery providers, and large corporations (e.g., Amazon and Walmart)—and physician-hiring battles are just a few of the major hurdles traditional health systems face, Unger said.
The answer to navigating these growing challenges, according to Unger, is for CFOs and finance leaders to play a bigger role in enabling and empowering physicians to deliver better care. He cited five key competencies financial leaders can apply to drive this transformation:
With decades of deep data (e.g., clinical, cost, claims, and social determinants of health), Unger said traditional healthcare organizations still have a strong advantage over their competitors—but only if they use that data to make more strategic decisions and play to win.
Sadiqa Mahmood, DDS, MPH, SVP and General Manager, Life Sciences Business, Health Catalyst
Sadiqa Mahmood, DDS, MPH, asserted that life sciences and pharma need a role model to execute on the promise of clinical research. This role model, she said, will enable research to overcome areas of ignorance and develop cross-industry collaboration and partnerships. Dr. Mahmood explained that the historic approach to clinical research hasn’t leveraged adequate data, yielding a lengthy and challenge-ridden process, leaving clinical research ripe for transformation.
Beginning in March 2020, COVID-19 forever changed how the industry thinks about clinical research innovation and assess vaccine risk. With support from the U.S. government’s COVID-19 vaccine initiative, Operation Warp Speed, researchers pursued innovative mRNA vaccines and published early clinical trial data by summer 2020. Remarkably, by August 21, the FDA granted emergency use authorization (EUA) to the first of these vaccines—just months into the clinical research.
Until that August 2020 EUA milestone, said Dr. Mahmood, the fastest a vaccine development process had taken four years (the mumps vaccine in 1976), with 8.5 years being the average. She explained that because researchers had access to existing mRNA technology, this achievement has changed the clinical research paradigm. It wouldn’t have happened without innovative partnerships across clinical and technology R&D sectors.
The innovative partnerships, or learning networks, Dr. Mahmood described are data centric, leverage quality information, and have terminals that enable information sharing. She outlined the following six considerations for use of data to support clinical research:
Dr. Mahmood added a call to borrow ideas from other industries that use more data than healthcare. She offered the example of Formula One car racing, explaining the vast partnerships that support each car and driver. Cars are equipped with hundreds of sensors, generating multiple terabytes of data that 10s to 100s of on- and off-sight employees analyze. Additionally, many different team members with different skillsets and strengths are on the track, around the car, addressing needs the data communicates. The large and dispersed team comes together to utilize data to create a winning partnership. Dr. Mahmood attests that clinical research can achieve this same level of partnership for success.
Rana El Kaliouby, PhD, Co-Founder and CEO, Affectiva, Pioneer and Inventor of Emotion and Human Perception AI
As artificial, or augmented, intelligence (AI) becomes mainstream, users and innovators have largely focused on its logical and transactional aspects. In other words, we know AI has a high IQ but tend to give it less credit for emotional intelligence.
Rana El Kaliouby, PhD, seeks to change the way we perceive the emotional side of technology, positing that AI can identify human emotion. For example, a computer can distinguish between a smile and a smirk, thereby accessing the primary way humans communicate, which is nonverbally through facial expression.
However, coding human expression manually is too labor and time intensive, Dr. El Kaliouby said. Instead, she explained that developers use computer vision, machine and deep learning, and video data to train algorithms to recognize facial patterns and map them to emotional and cognitive states. This involves looking for features common between the same expression (e.g., a smile or brow furrow) on different faces. This way, when a machine sees a new face, it can tap into deep neural networks for emotion recognition to assign a probability score to facial expression.
Computers at Dr. El Kaliouby’s company, Affectiva, can now detect over 35 expressions, including levels of attention, cognitive overload, wakefulness, heart rate, and more. In healthcare, she explained emotion AI will transform mental health, providing an objective measure of the patient’s well-being, versus the traditional mental state survey. She offered other examples for emotion AI in healthcare, including Parkinson’s disease detection, rebuilding smiles for stroke patients, and adding an “empathy score” to measure clinician-patient interaction. Additionally, Dr. El Kaliouby described robot companions for the terminally ill that could understand a patient’s well-being and call in a human nurse or physician when they deviated from baseline, a virtual therapist that understands emotion, and an assistive device to help people with autism understand emotions and social cues.
While Dr. El Kaliouby offered a lot of promise for emotion AI, she also touched on a top issue in AI overall—mitigating bias with diversity, equity, and inclusion. She said innovators must be intentional when designing algorithms to not accidently add bias, leveraging population data that’s balanced with gender, dress, skin tone, etc.
Phillip Rowell, MJ, Vice President of Clinical and Business Intelligence, Carle Health
There is a distinction between executive governance and data governance, but health systems often conflate the two. According to Phillip Rowell, executive governance is a common agreement between executives about how to leverage data within the system. Data governance focuses more on defining key metrics and how to distribute and apply data to identify improvement opportunities. A lack of executive governance will impede successful data governance, which is critical to maximizing a system’s precious data assets.
Rowell discussed Carle Health’s three-step journey to achieve executive governance and the role it played in laying the groundwork for data governance:
Although executive and data governance are internal initiatives, Rowell emphasized their far-reaching effects of improving patient satisfaction, enhancing care delivery, and increasing revenue opportunities.
Jack Beal, JD, Vice President, Performance Improvement Deputy General Counsel, The University of Kansas Health System
Maksym Ovsiyenko, MBA, MHA, Director of Applied Analytics and Business Intelligence, The University of Kansas Health System
As a large academic health system, The University of Kansas Health System was already using a data-informed problem-solving process. But, although the organization widely distributed data, this wasn’t always leading to action—or to a meaningful ROI. Leaders determined that the lack of action was primarily a people problem, not a data problem. For example, clinical departments had varying levels of support for improvement work, and many felt disconnected from system leadership. Additionally, there was an “us-versus-them” mentality, which misaligned incentives further complicated.
So, what did The University of Kansas Health System do differently? It focused on better connecting the right people to the process. Critical success factors included identifying and involving key stakeholders (especially clinicians) early and often; clearly defining success; providing frequent communication; using continual PDCA (plan-do-check-act) improvement cycles; and aligning resources and incentives to promote change.
Jack Beal, JD, and Maksym Ovsiyenko, MBA, MHA, shared several examples of success after making these adjustments. One standout was advancing value-based performance (VBP) at their organization—moving away from compensation based on volume to compensation based on value. By engaging key stakeholders early in multiple pilots—and measuring and communicating shared value—these efforts have already led to $600,000 in shared savings and continued work toward integrating VBP metrics into physician contracts.
Finally, Beal and Ovsiyenko provided details of a more deliberate four-step Strategic Performance Improvement System, which in its final stages of building and testing. The goal is to continue to leverage and deploy data in a way that connects the right people in the right ways to compel the right actions.
Chester Ho, MD, Chief Medical Officer, Health Alliance
April Vogelsang, RN, MS, Senior Vice President and Chief Clinical Integration Officer, Health Alliance Medical Plans and Carle Health
Though vertically integrated, Carle Health (a healthcare provider) and Health Alliance (a health insurer) still functioned separately when it came to patient care. April Vogelsang, RN, MS and Chester Ho, MD, saw the separation as a missed opportunity. They believed the organizations could integrate more to improve care for the 50,000 patients they shared while cutting waste. To do so, Carle Health and Health Alliance created and implemented a new population health care model with interdisciplinary teams to help patients manage their care.
At the beginning, Vogelsang and Dr. Ho struggled to get patients to sign up. Why? As one newly diagnosed cancer patient told them, “I already have to talk about cancer every day. I don't want to talk about it even more on the phone to some stranger from an insurance company.” With such answers, Vogelsang realized they would need to meet patients where they already were—face to face at their provider's office.
Using data from their analytics partner and clustering algorithms to visualize patterns of care, Vogelsang and Dr. Ho selected five locations to start. They embedded care management staff in those clinics physically and/or virtually. The teams included mostly existing staff with only three new hires needed to support the model.
After the first year, the patient engagement effort saw a 3-to-1 return on their investment, a 40 percent reduction in emergency department admissions, and a 30 percent reduction in readmissions (all approximations). Among many other learnings, Vogelsang and Dr. Ho concluded that working with a data analytics partner was a key part of the model's success.
Holly Fetter, MS, Healthcare Data Scientist, Albany Med
Holly Fetter, MS, discussed the most intense and rewarding project she worked on at Albany Med—the development and launch of their denial management tool, ReClaim. Fetter began the session with a review of the typical claim and denial process, explained how decentralized management, limited data access, siloed workflows, and hospital/practice ownership can impede effective denials management.
For a successful first version of the denial management tool, Fetter and the team had to understand the process. They sat with subject matter experts, asked a lot of questions, and became the experts themselves. Then, with each new version of the tool, the team had new challenges to overcome, which the biggest challenge being incorporating new source data from various systems.
The team’s persistence paid off. So far, ReClaim has helped Albany Med achieve the following:
Fetter ended the session with four recommendations: identify key players, define strategic objectives upfront, utilize integrated data and analytics, and become an expert.
Sadiqa Mahmood, DDS, MPH, General Manager & Senior Vice President, Life Sciences Business,
Health Catalyst
Qin Ye, MD, MS, Principal, ZS Associates
Imagine a world where providers and pharma come together to improve the quality of care for Alzheimer’s, diabetes, or even cancer. Sadiqa Mahmood, DDS, MPH, and Qin Ye, MD, MS, discussed the importance of the provider-pharma partnership. Though the need for collaboration already existed, the COVID-19 pandemic made it more important. The COVID-19 vaccination is an example of the ideal partnership—providers contributed data while pharma and biotechnology contributed expertise and technology.
Although providers may face common challenges (economic pressure, regulatory changes, patient outcomes, etc.), the suggested collaboration framework leverages their collective strengths to diminish the challenges. The collaboration framework has three key parts:
To reinforce the importance of these relationships, Dr. Mahmood and Dr. Ye shared results from provider and pharma partnership interest surveys and additional real-world examples around diabetes, metabolic disease, and cancer survivorship.
Mahmood and Ye ended the session with takeaways that could benefit any provider interested in collaboration:
David B. Pryor, MD, Former Executive Vice President and CMO, Ascension Health, Former President & Chief Executive Officer, Ascension Clinical Holdings
Brent C. James, MD, MStat, Clinical Professor, Clinical Excellence Research Center (CERC), Department of Medicine, Stanford University School of Medicine
A board-certified cardiologist and over 18 years of executive experience at Ascension, Allina Health, and Tufts Medical Center makes David Pryor, MD, an expert at driving systemwide, large-scale change. As Dr. Pryor sat down with Brent James MD, MStat, to discuss his approach to achieving change, he said that not only does better care save lives, it’s also good business.
In one of Ascension’s initiative centered on reducing waste, clinically led efforts saved the system over $1 billion in one year. So, how can an executive—whose position is seven layers removed from where the change needs to happen—inspire change in the clinical setting? Dr. Pryor’s answer includes four concepts:
Dr. Pryor dove deeper into concepts three and four. When setting organizational goals, he advised limiting the goals to one transformational idea, then reducing that idea into an elevator speech before marketing it broadly to the system. To effectively work towards that goal, Dr. Pryor suggested dividing it into smaller segments that are easy to implement within existing workflows and emphasized the importance of culture in fostering systemwide change. He recalled an experience with a mentor early in his career who told him that a strategy usually doesn’t fail because it’s wrong, but because of the organization’s culture. Lastly, Dr. Pryor closed by reminding leaders that it’s more important to be effective than right because effective leaders empower team members to drive change at their individual level.
Taylor Larsen - Sr. Director of Data Quality & Operations, Data Business Unit, Health Catalyst
Healthcare organizations increasingly rely on data to inform strategic decisions. This growing dependence makes data quality across the organization more critical than ever. A recent Gartner survey found that organizations estimate the cost of poor data quality at $12.8 million per year. Poor data quality results in lost revenue, bad business decisions, duplicate efforts, and employee turnover. Those costs are amplified in healthcare due to the human costs involved, from patient safety issues to provider burnout.
In this session, Larsen posited that attendees were likely there for one of two reasons: they’ve invested in data quality but are not seeing the results they want, or they know that their data quality investment is inadequate, but don’t how to get the resources they need. Either way, Larsen said without optimization, it’s going to be difficult to improve analytic time to value and prepare for the future. He then highlighted three main data quality optimizations that healthcare organizations can focus on to drive better results and secure more resources:
1. Increasing data quality visibility across tow time-related dimensions (calendar and processing time).
2. Reducing the pain of getting data quality in place.
3. Making data quality important to the broader organization and scaling its expansion.
Larsen then walked attendees through actionable ways to make these optimizations and highlighted successful case studies for each. Making these optimizations will help quantify the value of data quality to help drive adoption for future success.
<h3>Session 31 – AI-Enabled Population Health = Healthy Patients and a Positive ROI
Rhiannon Harms, Executive Director of Strategic Analytics, UnityPoint Health
Kristin McKay, Executive Director of Care Management, UnityPoint Health
UnityPoint Health makes the following promise to their community: “Care that is easier and more personal.” Kristin McVay and Rhiannon Harms delivered a trove of information on how they use artificial intelligence to fulfill this promise and improve population health.
McVay discussed how the UnityPoint Health team determines good candidates for their care management program, including high risk of utilization, risk stratification, probability of admission and readmission, and clinical judgment.
A close partnership between the care management and analytics teams enabled the analytics team to build the tools required to best serve their patients and providers and support the following:
The team tracks their results through the Outcomes Analyzer tool, which showed an impressive reduction in utilization as the care management program has grown:
Thomas D. Burton, MBA, Cofounder and Strategic Advisor, Health Catalyst
Channeling his inner Alex Trebek, Tom Burton led three lucky healthcare contestants—a chief information officer, chief finance officer, and chief analytics officer (or perhaps their impersonators)—through an educational, Jeopardy-style game. The contestants’ objectives were to prove their knowledge of the key transformations necessary to sustain organizations in an environment of decreased hospital volumes and lower revenue.
Using categories that reflect the five necessary types of transformation, Tom presented their core principles—in the form of questions, naturally—to help audience members gauge their organization’s proximity to the ideal state:
These transformations—all which education can enable—allow organizations to adopt quality as a business strategy, accelerating the shift to value-based care and enabling providers to do the right thing for the patient while also improving the bottom line. Acknowledging that it’s not a fast or simple journey, host Burton explained that it’s nevertheless worth it. Organizations that successfully transform can double or triple their operating margins, despite lower revenues.
Fans of puns won’t be surprised to learn that contestant “Cat A. Lyst” won the game.
Nancy D. Lin, Vice President, Real World Insights & Evidence, Health Catalyst
Carla Rodriguez-Watson, PhD, MPH, Director of Research, Reagan-Udall Foundation for the FDA
The rapidly evolving use cases in COVID-19 therapeutics, diagnostics, and vaccines demand evidence. To that end, Carla Rodriguez-Watson, PhD, MPH, and Nancy D. Lin described how a national collaborative community of stakeholders convened to share insights, compare results, and answer solve key challenges. The collaborative uses real-word data (RWD) and real-world evidence (RWE) to answer questions about COVID-19 diagnostics and therapeutics.
Despite the massive investment and digitization, healthcare hasn’t achieved sufficient data for surveillance, supply chain, and regulatory decision-making—a reality the pandemic has acutely laid bare. In response, the Reagan-Udall Foundation for the FDA and Friends of Cancer Research developed the COVID-19 Evidence Accelerator to convene a RWD healthcare ecosystem. This community of data and analytics partners is ready to urgently address questions about COVID-19 in therapeutic, diagnostic, and vaccine workstreams.
Evidence accelerator goals include running several analyses in parallel to quickly field multiple questions, addressing critical questions related to COVID-19 through data already collected by participating data partners, align on a common analytic plan, deploy against a common analytic plan, and report results for presentation. Using parallel analysis methods, potential questions the accelerator can answer span from patterns of general outcomes for people with COVID-19 to real-world performance of serology tests to real-world safety of therapies used in COVID-19. In use cases, notable challenges have included a lack of standard terminology and data curation, which the collaborative will address with standardization, mapping, recruitment of local partner expertise, and transparency.
Captain Paul Horstmeier welcomed attendees back for the final gathering at the close of the multicity HAS 21 Virtual adventures. Horstmeier invited participants to take a brief survey to make next year’s summit even better and added a reminder to earn CME credit. Horstmeier then announced the winners of various event competitions, and highest rated sessions, and the 2021 Annual Flywheel Award winner—ChristianaCare for using data and AI to reduce sepsis mortality rates and saving millions in reduced costs.
Dan Burton took the stage to close the 2021 summit, offering a heartfelt thanks to everyone involved and reminding attendees about the value of taking a multi-domain analytics approach (even when it comes to basketball). Dan likened the process of transforming care to the small-town basketball team competing for the state championship in the movie Hoosiers. Just as the players felt overwhelmed walking into a huge stadium, so do healthcare organizations when facing an unpredictable future. However, the coach reminded the players of the foundational principles, to not get distracted by the big stadium and flashy lights, and so did Burton. Some things will always remain the same no matter what we face, he said, such as our desire to make healthcare better through data and analytics.
Would you like to learn more about this topic? Here are some articles we suggest:
