Thursday Recap: 2016 Healthcare Analytics Summit Finale
The final day of the 2016 Healthcare Analytics summit began with a reminder about an easy-to-use tool that will help health systems improve outcomes: Health Catalyst’s Outcomes Improvement Readiness Assessment. The assessment gives health systems an in-depth review of their competencies in five main categories:
- Adaptive leadership and culture
- Best Practices
- Financial Alignment
HAS attendees will get the rare opportunity to not only take this assessment, but also have their results interpreted for them by the assessment creators (Health Catalyst staff), including what’s working and what’s not in their health systems.
Keynote: Can Clinicians Take the Lead in Health Care Reform?
Don Berwick, MD, Former Administrator, CMS; Founding Chief Executive Officer, Institute for Healthcare Improvement
“What does American healthcare need today?” asked Dr. Don Berwick, former CMS Administrator and founding CEO of IHI. The answer according to seventy-nine percent of the audience: “A fundamental redesign.”
American healthcare accounts for 18 percent of the GDP with massive waste, from overuse to fraud and abuse. It’s fraught with conflict among stakeholders, workforce demoralization, and the pressure to produce. Healthcare consumers live in the tension between the beautify of healing relationships and an industry obsessed with production—a treadmill of processes shaped by anonymous forces over which clinicians have little control.
Healthcare is focused on incentives, when it should prioritize improvement and innovation. In an incentive-based, carrots-and-sticks healthcare era, Dr. Berwick imagines a new era, and outlined nine steps to get there:
#1: Stop excessive management.
#2: Abandon complex incentives.
#3: Decrease focus on finance.
#4: Avoid professional prerogatives.
#5: Recommit to improvement science.
#6: Embrace transparency.
#7: Protect civility.
#9: Reject greed.
Dr. Berwick’s primary message was that we have gone through two eras in healthcare, and that it’s time to enter a third.
Era 1 was defined by trust, prerogative, inquiry, mentorship, and research. He translated it as the era of stethoscopes and white coats.
Era 2 was (is) defined by accountability, scrutiny, measuring, incentive, and doubt. He translated this as the era of FFS and pay for performance.
Before he defined Era 3, he gave us a new way to think about each era’s qualities, in terms of quality control, quality improvement, ad quality planning. Quality control is keeping things in order, like fixing the flat on the car. Quality improvement is making the car better, keeping it maintained. Quality planning is innovation, like abandoning the car and creating an airplane instead.
Era 1 spent little time on control and innovation, and lots on improvement. Era 2 spends a lot on control, and little on the other two. Era 3 needs to balance all three.
This new era will focus less on money and more on patients and their families. After all, Dr. Berwick argues, medicine is about relationships. The quest, he says, is clear: “It’s not power, reward, or profit; it is a search for meaning in the value of the person in need. In healthcare, we need to ask, “Who is this person and what can I do to help?” And it’s not just the frontline responsible for the transition—its entire communities. It will take many villages working together to improve American Healthcare.
Keynote: Leveraging Data and Strong Partnerships to Thrive in the Land Between Volume and Value
Craig E. Strauss, MD, MPH, Cardiologist, Medical Director, Minneapolis Heart Institute for Healthcare Delivery Innovation
Healthcare organizations that leverage data and forge strong, strategic partnerships will thrive in the land between volume and value—that somewhat uncomfortable place American healthcare will be stuck in for a while. Health systems need strategies that work in this hybrid world. In addition to creating a shared vision, aligned interests, supportive infrastructure, platforms to allow ongoing investment, and, trust, there are three keys to thriving in this transitional state:
- Physician leadership.
- Strong care team collaboration.
- EDW with access to data (and an analytics team that knows how to leverage that data).
According to Dr. Craig Strauss, Cardiologist and Medical Director at the Minneapolis Heart Institute for Healthcare Delivery Innovation, “strong committed leadership is essential for success, innovation is required to navigate challenges, and collaboration is essential to achieve the Triple Aim.” Health systems can learn from the Minneapolis Heart Institute. Its Healthcare Delivery Information Center leverages data using retrospective, real-time, and predictive clinical intelligence tools and is in relentless pursuit of quality improvement.
Dr. Strauss included a reminder about the fee-for-service world in which the industry still operates: “In the fee-for-service world, there’s still an opportunity to improve care for under-treated patient populations; interventions for patients who will benefit from them; interventions that will reduce the cycle of readmissions; interventions that will, ultimately, mitigate healthcare costs.”
Other benefits from their improvement efforts included:
- The ability to understand the true pricing of their services, which helped with bundled payment programs.
- The ability to understand the difference in outcomes among diabetes patients compared to patients without diabetes; this allowed them to drive out variation in that particular line of care.
- Reducing complication rates in very specific patient cohorts, such as CABG patients with a BMI over 40
Keynote: PHM Is Here to Stay (And We Need Better Data to Get It Right)
David F. Torchiana, MD, President and Chief Executive Officer, Partners Healthcare
Currently, the US pays significantly for healthcare, yet we’re not seeing this return in life expectancy. When compared with other countries of similar economic standing, the US spends more on healthcare but less on the social fabric of society. In fact, when healthcare outcomes are aggregated, spending on social needs correlates with healthcare outcomes: when you underinvest in social programs, you end of with very costly healthcare.
Dr. Torchiana explained the steps needed to move us from fee for service (FFS, the historic payment model) to a system that pays for value. Under FFS the insurer has incentive, so there’s less uptake of disease management programs. Furthermore, FFS doesn’t cover coordination of care, leaving room for less than desirable outcomes.
He views U.S. healthcare as a never-ending narrative with many chapters:
Chapter 1: The High Cost of Healthcare
The U.S. spends significantly more on healthcare than social services and more than one third of expenditures are wasted in healthcare.
Chapter 2: Healthcare Quality
A 1999 IOM report stated that 98,000 Americans die each year as a result of preventable errors in hospitals – the equivalent of a 747 crashing every other day.
Chapter 3: Coronary Artery Disease Mortality
Coronary artery disease mortality has been on a steady decline since 1968; an example of the value of medical advances being greater than the costs. Healthcare expenditures can have real value in terms of preserving life and quality of life.
Chapter 4: Staying Healthy
The fundamental question is, “What if healthcare was more effective at keeping people healthy and diminishing disease?”
Chapter 5: Underinvestment in Social Programs
U.S. healthcare outcomes are less than ideal because it underinvests in social programs.
Chapter 6: Quality Is Improving, but it Won’t Ever Be Perfect
Quality of care is improving, but as long as opinionated, non-compliant humans are involved, it won’t ever be perfect.
Breakout 22: Outcomes Improvement Governance: The Quest to Achieve More with Less
Susan Easton, Senior Vice President, Health Catalyst; and Tom Burton, Co-Founder and Executive Vice President, Health Catalyst
Easton and Burton launched the world debut of the card game, Governance Quest, to teach core outcomes improvement governance principles. Attendees played the card game with carrying sets of rules that mimicked real-life challenges they may face when trying to implement an improve initiative. The principles included: Engage the Right Stakeholders, Create Shared Understanding, Use Alignment and Consistent Methodology, and Focus by Practicing Disciplined Decision Making to Prioritize Projects and Resources. Attendees also received a handbook entitled “Implementing Governance for Outcomes Improvement,” that covers the 11 steps needed to create successful, sustainable outcomes improvement governance structures.
Breakout 23: Integrating Detailed Patient Level Costs With Outcomes and Quality Metrics
Charlton Park, MBA, MHSM, Chief Analytics Officer, University of Utah Health Care
The University of Utah Health Care has embraced a culture of transparency to achieve accurate and actionable cost accounting and outcomes data, which are essential to creating value. In fact, Dr. Robert S. Kaplan and Dr. Michael E. Porter stated in a past New York Times article, “Understanding [patient level] costs could be the single most powerful lever to transform the value of healthcare.”
The first step to capturing and understanding patient level costs is to build the platform, which requires strong leadership support. Next, a team must be established with clear goals and must incorporate ongoing feedback from providers and executives. The University of Utah Health Care found that the team members were strongly motivated by understanding the importance of their efforts as the organization transitioned to a value-based care environment.
Leveraging a “sequester team” model, the team members met offsite three times per week with the specific objectives of developing a prototype within three months — complete with outcomes and quality measurements — and implementing an organizational rollout within six months.
The second step was to create value. This was achieved through a commitment to transparency, accurate data, actionable analytics and engagement.
As healthcare moves from fee-for-service to capitation bundles and value-based care, the identification of patient level costs, transparency in sharing that data and implementation of the actionable insights to create performance measures and standardized, best practice care will enable healthcare systems to not only live – but also succeed — in this new world of care.
Breakout 24: Deploying Predictive Analytics: A Practitioner’s Guide
Eric Just, Senior VP, Product Development, Health Catalyst
Guiding principle of Eric Just’s technical session was that predictive analytics is about using pattern recognition to predict future events. But predicting something isn’t good enough; we must also have the data to act and intervene. Machine learning plays a key role in predictive analytics because it explores the study and construction of algorithms that can learn from and make predictions on data.
There’s been a huge insurgence of predictive analytics since 2010 because limitations on index models are driving more specific models, data availability is better, analytics is more pervasive, and because we now have better machine learning tools.
The audience learned that to fully leverage the analytics environment requires a data warehouse to gain access to lots of pre-defined and raw data. It’s important to clean up and aggregate data for machine learning, to conduct feature engineering in order to feed machine learning algorithms, and to do this engineering in the data warehouse environment.
Other learning highlights were how to standardize tools and methods using production quality code and how to deploy with a strategy for intervention. Eric showed how predictive analytics appears in a workflow by demonstrating a CLABSI case study where nurses could see how many patients on the floor were at risk for CLABSI. The dashboard showed the specific number of high-risk patients and their risk scores, with individual patients listed at the bottom ranked by risk. This demonstration drew the most applause.
Breakout 25: Healthcare Analytics—Are You Just Buying a Car or Actually Planning to Go Somewhere?
Taylor Davis, Vice President, Analysis and Strategy, KLAS
Taylor Davis, Vice President of Analysis and Strategy at KLAS, provided an overview of the healthcare business intelligence (BI) market, based on KLAS’s surveys of hundreds of organizations about the BI solutions they are purchasing. Unfortunately, according to Davis, relatively few of the organizations report that they are achieving their goals. Davis shared KLAS survey data and his own insights to help explain what may be holding these organizations back from arriving at their desired “destination.”
Davis used the metaphor of buying a care to describe four categories of BI solutions available on the market and how their “owners” are responding to them:
- Buy for the engine: These are the cross-industry BI toolset vendors like IBM, Microsoft, and SAP who’ve provided BI solutions for multiple industries over many years. They claim to have big, powerful engines that a lot of large enterprises have used to drive success. Buyers rank these solutions lowest among the four when asked if their BI tools have been successful, according to KLAS data. They also rank last when buyers are asked to quantify their optimism in achieving success in the future using their BI tools.
- Buy for the GPS: These are the healthcare-only BI solutions like Health Catalyst that claim to understand healthcare better than the other vendors because they grew up in healthcare and it’s their only focus. They can be viewed as having the best GPS to help a provider organization reach their destination. Buyers rate this solution type No. 1 among the four categories both in current success and future optimism, according to KLAS. “they’re the only ones holding people’s hands and getting them where they want to go,” said Davis.
- Buy for the Tires: These are the enterprise EMR vendors who have built their own BI solutions. These foundational (tire) solutions claim to have the best access to BI data since it already resides in their systems. Solutions in this category rank No. 3 in both categories of current success and future optimism, because “(buyers say) they don’t have the functionality we need to reach our destination,” said Davis.
- Buy a Self-Driving Car: These are visual data discovery tools. Many organizations who select these solutions say “if we throw this self-driving tool out to a lot of different users, some of them will stumble onto the destination,” according to Davis. These “self-driving” solutions rank a very close second to the healthcare-only vendors in both current success and future optimism.
Breakout 26: New Competencies for Succeeding in Risk-Based Arrangements
Bobbi Brown, Senior Vice President, Financial Engagement, Health Catalyst; Dan Unger, MBA, VP Product Development, Financial Decision Support, Health Catalyst; and Lynn M. Guillette, CPA, MBA, Vice President, Payment Innovation, Dartmouth-Hitchcock Health
Over the next three years, value-based care and at-risk contracts will increase. Health systems will succeed in risk-based arrangements by developing seven key competencies:
- Realize your true risk adjustment factor (you don’t have much control over reimbursement in a capitated world, but you DO have the data and the capabilities to understand, track, and impact this metric).
- Improve self-pay collections.
- Prioritize high-value quality measures.
- Avoid adverse, unnecessary events.
- Actively manage high-risk patients (five percent of your population accounts for 40 percent of costs).
- Advanced cost reduction.
- Utilization reduction via public health-ish initiatives (health systems fail to get full financial credit for the risk they’re taking due to lack of timely insights into patient access and documentation).
Dartmouth-Hitchcock, a five-hospital system, embraces these competencies. It’s effectively managing the payment system transition. It knows the current fee-for-service payment system is unsustainable and that APMs offer an avenue to transition to pay-for value. It participated in the pioneer ACO program for three years with mixed financial results and positive quality performance. It withdrew due to concerns with benchmarking methodology and the use of national vs. regional trends, learning key lessons from its participation:
- Assessing financial and quality contract performance lags 6+ months after contract period ends.
- ROI measurement lags.
- It’s not enough to have data; know what it’s telling you and use it in actionable ways.
Health systems need to change behaviors and focus on the social determinants of health in order to change healthcare.
Breakout 27: Do No Harm: Reducing Hospital-Acquired Conditions Through Cultural Transformation, Analytics, and Education
Abby Dexter, Director, Business Intelligence and Data Warehousing, Children’s Hospital of Wisconsin; and Holly O’Brien, MSN, RN, CPPS, Safety Program Manager, Children’s Hospital of Wisconsin
As one of the nation’s top pediatric hospitals, Children’s Hospital of Wisconsin’s (CHW) is committed to their vision of “healthiest kids in the nation.” To realize this, CHW has set goals to improve outcomes and shorten inpatient length of stay, with a focus reducing the rate of hospital acquired conditions (HACs)—and planned to use analytics to do so.
According to Abby Dexter and Holly O’Brien, MSN, RN, CPPS, both of CHW, they faced an uphill battle when it came to implementing a data-driven program to reduce rates of HACs. They had poor access to data and lacked the cultural buy-in to support their initiatives.
CHW started by creating a solutions for patient safety (SPS) network to collaborate on their goal on driving down HAC. Dexter and O’Brien said that they hit a turning point when senior leadership bought into the initiative and committed resources to safety.
In addition to backing from the senior level, CHW put into play the following key strategies:
- An effective governance structure.
- Team-based approach (areas that needed to work together and leverage all skillsets)
- Data and analytics (to see results and how work was paying off).
- Education and cultural transformation (using data for improvement rather than punitively and train in tech to reduce HAC)
The team achieved a formidable 80 percent reduction in serious safety incidents.
Breakout 28: Partners’ Care Management Strategy: A 10-Year Journey
Sree Chaguturu, MD, Vice President for Population Health Management, Partners HealthCare; and Eric Michael Weil, MD, Senior Medical Director, Population Health Management, Partners HealthCare
Forty-two percent of those who attended Drs. Chaguturu and Weil’s presentation said their organization’s care management strategy was somewhat successful. When you consider that 7 in 10 death are related to chronic conditions, the lessons learned during this session could be life-saving. Attendees were encouraged to discuss topics such as: which patient populations do you pick for your care management programs; how do you identify those patients; brainstorm components of managing high-risk patients in a care management program; and finally, how would you modify care management strategies to manage different patient populations. Lessons from the discussion included the idea that control/comparison groups are key to measuring success, and most importantly, communicate, communicate, communicate!
Breakout 29: Turn Your Analysts into Data Detectives: Discvoreing Pttaerns in Dtaa
John Wadsworth, MS, Senior VP, Client Engagement, Health Catalyst
If you’d never heard of a data detective prior to John Wadsworth’s presentation, then you came away with a significant understanding of what a data detective is, what she does, how to identify the attributes and skills of an optimal recruit, and how to better support this role within your healthcare organization.
The drivers for this unique role have been the exponential growth in data capture, the occasional catastrophic events in healthcare, and the need for critical thinkers around big data. Wadsworth compared data detectives to other roles in the analytics realm, by examining three types of analytic domains:
- Reactive analytics have low complexity and require low technical skill and contextual understanding (common roles are report writer, business analyst, dept. analyst).
- Descriptive analytics have moderate complexity and require moderate technical skill and contextual understanding (common roles are data architect, data analyst, BI developer).
- Prescriptive analytics have high complexity and require highly technical skill and contextual understanding (common roles are data detective, data scientist, technical director).
At their core, data detectives are story engineers who work with domain experts to bring together seemingly unrelated data to collaboratively focus on a specific narrative. They make discoveries that others missed using the same data.
Data detectives ask a lot of questions and are self-aware when it comes to their own ignorance. This helps them identify missing elements. They let the data inform, keep bias at bay, and they don’t get married to an idea. Data detectives get to the heart of what matters. They figure out what the point is, not just the facts behind the point.
There were many wonderful lessons learned, but Wadsworth ended with this poignant message for current and future data detectives: Ours is the sacred stewardship to find better ways to deliver care.
Breakout 30: Powerful Ways to Use Hadoop in your Healthcare Big Data Strategy
Sean Stohl, Senior Vice President, Product Development, Health Catalyst; and Bryan Hinton, Senior Vice President, Platform Engineering, Health Catalyst
With the digitization of healthcare commencing, Hadoop offers the ability to effectively transform the various types of data (structured, semi-structured and unstructured) into actionable insights for healthcare systems.
The biggest challenges to Hadoop implementation include:
- Organizational: Many organizations are “stuck in the mud” with current processes and tools, or are currently distracted by the latest shiny new thing.
- Buying: Determining how to extract value from Hadoop is also a significant barrier to purchase. Leveraging cloud vendor options (Microsoft, Amazon, Google, etc.) is an ideal way to explore the capabilities and benefits of Hadoop with limited investment and risk.
- Administering: There are still fewer experienced people in Hadoop and a lack of established best practices. A myriad of open-sourced tools are available, which offer the pros and cons of “assembly required.” Security of these relatively new tools can also be a concern.
- Using: You’ll be using SQL in some fashion, so you’ll just need to decide which technology to implement. The key is to remember that this is not a “rip and replace” scenario, but one of convergence. Specifically, RDBMS vendors (Oracle, SQL server, Teradata, etc.) will converge with Hadoop solutions (Hortonworks, Cloudera, Mapr, Cloud, etc.).
Big Data is an inevitable reality for healthcare, so now is the time to investigate Hadoop. Choose a cloud vendor (Microsoft Azure Cloud, etc.) and with a couple clicks of a button, you’ll have a small Hadoop cluster in which to play and prove the value to your organization.
Breakout 31: Leading Adaptive Change to Create Value In Healthcare
Val Ulstad, MD, MPA, MPH, Chief Engagement Officer, Partners at Cascade Bluff, LLC
What does it mean to lead adaptive change to create value in healthcare? At its core, the concept is simple: in healthcare, adaptive leadership empowers people in health systems to do the hard work required to improve outcomes.
Adaptive leadership describes group dynamics and strives to keep people in the productive zone of tension. When people are above the productive zone, the heat is too high. When people are below the zone, the heat is too low:
- Below the zone, people are disengaged (“Is there a gadget or pill to fix this?”).
- Above the zone, people are overwhelmed (“I am so terrified I don’t understand what you’re saying”).
- In the zone, people are productive (“I understand what I need to do”).
Leaders need to understand the difference between adaptive and technical work. For example, data is technical and the conversations about the data is adaptive. The most common cause of leadership failure is treating an adaptive challenge with a technical fix. Health systems can use adaptive leadership to mobilize adaptive work and address the gap between the way things are and achieving the Triple AIM:
- Practice dialogue.
- Conduct stakeholder analysis.
- Develop influence.
- Build up the emotional bank account.
Ultimately, exercising leadership to do adaptive work means disappointing people’s expectations at a rate they can tolerate. Adaptive leadership demands conversations, a passionate focus on people, and creativity. It mobilizes the individual parts of systems to work together as a whole.
Breakout 32: Population Health Management – Driving Improved Outcomes in Women’s Services Through Collaboration and Analytics
Stephen Poore, MS, MD, FACOG, Medical Director of Women’s Health, MultiCare Health System; and Maureen Faccia, MBA, Director of Women’s and Retail Health Services, MultiCare Health System
While there’s a movement toward standardizing care in order to improve outcomes and lower cost, Stephen Poore, MS, MD, FACOG, and Maureen Faccia, MBA, (both of MultiCare Health System) acknowledge that asking providers to change long-held patterns of behavior is no easy task. In their breakout session, Dr. Poore and Faccia described how they met the challenge of changing provider behavior, with a specific look at women’s medical care.
Women’s services are an important candidate for improvement within a health system due to the high rate of women who undergo surgical gynecological procedures. One in three pregnant women, for example will have a cesarean, and another high proportion will undergo hysterectomy. Clearly the opportunity is set to improve and outcomes and lower costs within a health system by focusing on women’s services.
They set a goal to reduce cesarean rate to 9 percent less than national average in their health system, which they met. In addition, the also saw reductions in the following:
- Abdominal hysterectomies
- SSI rate in cesarean
- Third or fourth degree perineal laceration rate
Data was essential to seeing these outcomes improvements (and consequential cost savings), as analytics helped accelerate quality improvement and sustain the gains. With analytics, Dr. Poore and Faccia were able to:
- Identify organizational structures to effectively set up and empower successful clinical improvement teams.
- Define strategies to engage and support providers for effective practice changes.
- Highlight performance and opportunities for improvement.
- Demonstrate the positive impact that standardized clinical content can have on patient outcomes.
- Show improvement, compare results, and correlate interventions with outcomes
Breakout 33: Improving Outcomes in a Value-Based Environment: Holistic Care Management for Complex Medical Conditions
Kyle Grunder, MBA, Director of Operations, Courage Kenny Rehabilitation Institute, Nasseff Spine Institute, Allina Health; and Jill E. Henly, MSW, LCSW, Manager, Care Coordination and Social Work, Courage Kenny Rehabilitation Institute, Allina Health
Kyle Grunder and Jill Henly of Allina Health shared their experiences working in Courage Kenny Rehabilitation Institute’s (CKRI) transformative holistic care management model as it looked to improve outcomes in today’s evolving value-based healthcare environment.
CKRI serves a very complex population – the average patient has seven to nine medical conditions in addition to one or more disabilities; and 80 percent of its patients live below the poverty line. Traditional patient support for the disabled was less than ideal due to the siloed nature in which care had been provided.
CKRI sought to refocus its entire culture on a whole-person care approach, backed by a system-based support of more comprehensive services and skilled providers both within and outside its organization. To better serve this complex population, CKRI created a Certified Medical Home – a fully accessible primary care clinic designed to help resolve the frustrating access issues for the disabled that so often results in poor outcomes and higher costs of care.
As part of this transformation, CKRI incorporated several innovative changes, including:
- Hour-long appointments to more fully evaluate patients
- Goals designed to meet the unique needs of individual patients
- A focus on early intervention
- Coordinated medical and social services
- Preventive services to manage chronic conditions
- A large referral network of community-based services
- Co-located physiatry, psychiatry and pain services
- Eleven specialty programs
Three-person teams led by nurse care coordinators were assigned to check with each patient and their families at least twice every week. In addition to care coordination, the teams help patients address non-medical needs such as housing, legal and financial challenges that often hinder their ability to adhere to care plans.
CKRI also invested in an electronic data warehouse (EDW) and analytics platform to integrate clinical demographic, cost, claims, functionality and satisfaction data that painted a more comprehensive picture of its patient care coordination. Aligned with the goals of the Triple Aim, CKRI was able to use these analytics to help identify, monitor and proactively manage care coordination.
The analytics platform has helped identify significant outcome improvements in key areas, including:
- 30 percent reduction in hospitalizations
- 66 percent reduction in hospital days
- 79 percent reduction in 30-day readmission days
As a result of those reductions, along with a decline in ED visits, Grunder and Henly noted that CKRI realizes annual community cost savings of $11.2 million.
Breakout 34: How to Measure and Get an ROI out of your Outcomes Improvement Projects
Bobbi Brown, Senior Vice President, Financial Engagement, Health Catalyst; Leslie Falk, MBA, RN, PMP, Senior VP of Customer Success, Health Catalyst; and Terri Brown, MSN, RN, CPN, Clinical Specialist, Quality & Safety, Texas Children’s Hospital
We’ve seen the numbers before: the U.S. spends $3 trillion a year on healthcare, of which we waste $1 trillion, yet there are still 400,000 premature deaths associated with preventable harm. The trio of presenters in this session helped us to understand the need for calculating benefit-cost ratio (BCR), the need for measuring ROI, and how to calculate both to address the healthcare problems brought on by these statistics.
Unfortunately, ROI is not something taught in nursing school, yet it’s an essential skill for all conditions. Fortunately, this session provided the formulae for calculation and conducted hands-on instruction.
Benefit-Cost Ratio (BCR) formula = Benefits/Costs
ROI = 100% x [(total benefit) – (cost of benefit)]/cost of benefit
Attendees also learned the 5-step approach to implementing an ROI project:
- Define the project and biz need
- Begin to quantify costs/benefits (don’t over complicate)
- Plan and implement
- Evaluate costs, revenue and benefits
- Communicate outcomes
It was noted that perhaps the most essential quality to move an ROI project forward is to believe in it.
Halfway through the session, the room buzzed with the activity of calculating BCR and ROI on a project to increase cervical cancer screening rates, an engaging educational opportunity that had the presenters floating the floor to gauge the progress at each table. Tables arrived at the solutions quickly, a shining example of the good things that happen when physicians, finance, and IT (the polled attendees in the room) come together
Terri Brown from Texas Children’s Hospital demoed a use case example of their evidence-based outcomes center BCR and ROI. She described the assumptions made going into the project, methodology used, shared sample analytics applications visualizations, reporting, and forecasting, as well as the results.
Breakout 35: Text Analytics: You Need More than NLP
Eric Just, Senior VP, Product Development, Health Catalyst
Much (up to 80 percent) of our healthcare data exists not in ready-to-use metrics, but in text. It’s therefore critical that we continue to develop effective systems to extract high-quality data from information available only in language form. Eric Just explained that one approach—natural language processes (NLP)—is doing some of this work, but that there’s opportunity for more capable technology.
Text analytics, Just explained, can be leveraged across organizations for more efficient, better, and larger information. But many organizations today are ignoring text analytics (due in part to the expense and resources required to build a system). Without this capability, though, they’re missing out on critical knowledge, such as better understanding their high-risk populations.
Successful text analytics will leverage medical terminology to carry out more accurate searches. For example, when you say enter “diabetes,” in a search, what do you really mean? With synonyms and distinct types of conditions, we’re likely picking up broader searches when we want specifics. In response, we need to provide a way to quickly identify synonyms in search engines and get more targeted results.
Breakout 36: Improve Reported Outcome Measures With Standardized Care Processes
Amber Theel, RN, BSN, MBA, CPHQ, CPHRM, Director of Quality Outcomes and Metrics, MultiCare Health System; and Jess Bouma, MD, Hospitalist, MultiCare Inpatient Services, Tacoma General and Allenmore Hospitals
MultiCare Health System recognized the need for standardized care processes and better outcomes for pneumonia, the 8th leading cause of death that accounts for 5.7 million inpatient days per year — and two of the publically reported CMS measures. Leveraging near real-time data to help guide their work, the organization introduced a multi-disciplinary clinical collaborative that established the guidelines and priorities for the organization-wide quality improvement initiatives.
The clinical collaborative achieved their goals – and beyond:
- 36% reduction in pneumonia mortality rate
- 21% in pneumonia readmissions
- 8% reduction in average variable cost per pneumonia patient
- 5% reduction in pneumonia patient length of stay
Lessons learned included:
- Support from senior leadership is vital. It conveys authority and respect across the organization, minimizing barriers and increasing willingness to change.
- Leverage the subject matter experts of the inter-professional team and allow them to work on items within their scope: nursing, radiology, respiratory, pharmacy and analytics.
- Do not make assumptions about the patient population. Time invested to perform thorough investigation and analytics provided new insight, resulting in the ability to better meet different patients’ needs.
- Physician engagement is critical to success. By incorporating feedback from multiple clinical specialties, the organization will achieve standardized, best practice treatments and enhanced patient outcomes.
Keynote: The Power of Not Knowing
Liz Wiseman, Best-Selling Author, Speaker, and Executive Advisor
While the pursuit of knowledge is often lauded as a worthy focus, especially in the workplace, Liz Wiseman makes a case for the benefit of not knowing. Whereas with knowing we run the risk of relying on what we know and thereby missing out on what others know, when we’re new to subject matter, we’re more likely to reach out to those around us. When we reach out, especially to more than one person, we vastly expand our knowledge base.
Wiseman applies this concept of not knowing, or “rookie smarts” to leadership. When leaders fall back on solely their own intelligence, they’re less likely to encourage the genius of others. They can even suppress the knowledge around them.
According to Wiseman, we tend to do our best work in this rookie zone. We don’t like not know and are therefore driven to find answers (asking questions, seeking help, staying alert). In rookie mode, we don’t establish patterns that keep us from seeing what’s really there. We make more mistakes, but we also work faster and are highly in tune with the need of your workplace.
How do we escape the knowledge trap? Liz offers five ways to lead from a place of inquiry:
- Operate from a place of not knowing; ask the questions rather than provide the answers.
- Admit what you don’t know.
- Throw away the notes; it can generate fresh thinking.
- Name the genius and the native know-how of each person on your team; do this even with those who seem most unconstructive.
- Seek new data sources; if you want to see new possibilities, you have to get beyond your assumptions, question your knowledge, and let this renew you.
Keynote: Closing Remarks
Dan Burton, Chief Executive Officer, Health Catalyst
The 2016 Healthcare Analytics Summit concluded as Dan Burton, Health Catalyst’s CEO, took the stage.
Keynote speakers that resonated most with attendees were Don Berwick, Anne Milgram, and Liz Wiseman. Don’s nine-step strategy for reforming healthcare, Anne’s parallels between the criminal justice system and healthcare, and Liz’s talk about the importance of being a rookie in the new game of work struck powerful chords with the audience.
The highest-rated breakout sessions were outcomes improvement governance, partners’ care management strategy, and leading adaptive change to create value. The Outcomes Readiness Assessment revealed surprising results: the most challenging areas for health systems were analytics and best practice, with leadership, culture, and governance as the highest-performing area.
Once again, the Summit created a myriad of opportunities for shared learning experiences. It created an environment that encouraged us all to prioritize learning over knowledge in our pursuit of becoming rookies for life, intent on admitting what we don’t know and seeking new data sources on our quest to scale outcomes improvement.