The Managers Field Guide to Actioning Healthcare Analytics

Gregory S. Nelson, MMCi, CPHIMS
Chapel Hill, North Carolina

About this guide

Discover what it takes to put healthcare analytics into action.

The data you collect every day has the power to help real people and your business' bottom line, but most healthcare organizations are struggling to turn patterns in healthcare data into actionable insights.

The Manager's Field Guide to Actioning Healthcare Analytics contains expert advice on how to overcome this challenge and start realizing the full potential of your data to inform patient care and improve lives. Here you will you'll learn about:

  • The current state of healthcare analytics
  • Common challenges and pains we experience around making analytics actionable
  • Solutions to those challenges, including the organizational capabilities and individual competencies needed to achieve analytics maturity
  • What's next for healthcare analytics
Introduction icon

Save For Later

Remember, if you'd like to save this guide for future reading, you can:

  • Bookmark the page in your browser
  • Save it as a PDF using the print option in your browser
  • Or, if you're on a mobile device, add it to your reading list

Healthcare Analytics: What’s the big deal?

Analytics is the backbone, the nervous system and the learning center of the health IT-enabled healthcare system.

Jonathan Weiner, Director Johns Hopkins Center for Population HIT

As mega-trends go, “analytics” is resilient. In large part, because of its ability to impact the way that we work, the decisions we make and the outcomes we achieve. Analytics is often seen hanging in the same circles as “big data”, “data science”, “informatics” and even “business intelligence”. (See this article for a summary of the differences between analytics and other common terms.) However, analytics should be considered a strategy and set of processes. In fact, here is how we operationally defined analytics for the purposes of this article: the scientific process for fact-based problem solving.  We don’t want to diminish those that would tie it to statistics, computational algorithms, data visualization, or massively large databases, but realize that (1) they aren’t the same thing and (2) while useful, these other disciplines are not required to engage in a data-driven, fact-based discovery and problem solving process.

We frame analytics as a process and is supported by these following observations:

  • …is not a destination, but rather the process of gaining insights to effect change. Analytics is the art and science of turning data into actionable interventions
  • …allows us to discover meaningful patterns in data and supports the examination of data with the intent of drawing a conclusion (acting)
  • …is not a technology, although technology is used to support the process
  • …is more than simply counting things or using simple math, but takes advantage of what we know about the past to predict and optimize the future
  • …can include but does not require computationally intense algorithms that can only be driven by the “data scientists” of Silicon Valley but rather by the curious, anticipators we call Data Champions
  • …necessarily creates an artifact that is used as input to a “decisioning” process. That is the process of analytics creates a data product – large or small, reusable or not – that feeds another process

Key Point

Analytics is a scientific process for fact-based problem solving.

Like many structured processes that have preceded modern analytics, we continue to evolve our thinking and raise our proficiency in how we make sense of the world.  We benefit from the accumulated knowledge of our prior work including such the scientific process, statistics, exploratory data analysis, data mining, artificial intelligence, data visualization, computational sciences, psychology and behavioral economics, Lean thinking, Six Sigma, and so on.  The world of analytics has benefited from each of their unique contributions as applied to thinking, learning, problem solving, deciding and behavior change.   

In this field guide, we offer a practical perspective on “actioning” analytics in healthcare. While not officially a word, actioning is used because it implies movement, that is, actioning is an organized activity to accomplish an objective or outcome

In the following section, we will touch on what we believe analytics is and its contribution to the modern healthcare organization.

Healthcare Analytics Is About Improving Outcomes

Things get done only if the data we gather can inform and inspire those in a position to make [a] difference.

Mike Schmoker, PhD, author, former administrator, English teacher, and football coach

We analyze to understand, frame and solve problems, make decisions, and create insights that can be used to drive change. We use what we know to make sense out of our worlds – that is, we “describe, discover, predict, and advise” (Blackburn, et al., 2015). But we would argue that advise falls short when analytics neither creates change nor produce outcomes.  Results are interesting at best. The litmus test we should use is whether analytics has real world impact. Fortunately, there are plentiful examples how analytical thinking and its resultant products create change throughout healthcare organizations.

Healthcare Analytics has the power to transform healthcare and is on the cusp of realizing tremendous benefits from analytics in the world of clinical, operational and financial areas.  As AHIMA reported in their 2017 survey of healthcare executives (AHIMA, 2017), they found that two-thirds of healthcare decision makers say analytics is a top priority.

Analysis icon
Percent (%)
Areas where health executives were planning technology upgrades: Security, Analytics, Patient engagement, Population health, Electronic health records

So why all the attention? To answer that, we look to some of the outcomes that we can achieve with analytics.

Examples of impact using Healthcare Analytics

Chart icon

Much of the early success in healthcare analytics included applications in financial and operational areas.

Within financial functions, healthcare organizations can:

  • Understand actual costs of care and total costs of care
  • Manage the reduction in reimbursements and associated risk
  • Shift to new payment models including bundled payments
  • Gain leverage in contracting with payers
  • Improve utilization
  • Optimize revenue cycle and supply chain
  • Reduce fraud, waste and abuse
  • Assess referral leakages out of network
  • Improve the bottom line

Within operations areas, healthcare organizations can:

  • Optimize processes—work more efficiently so that you can spend more time with patients
  • Mitigate operational risks associated with patient care
  • Improve compliance and quality
  • Enhance communications among stakeholders using data
  • Enhance the patient experience and their personal engagement in their health and well-being

While financial and operational analytics have importance, the transformative power of analytics can often be seen in applying analytics with strategic aspirations – those that integrate data from various parts of the healthcare enterprise to understand, model, and predict the entire system. For example, we see the true the opportunities for analytics in healthcare in three areas:

  • Contracting for and managing risk (shared savings, bundled payment, ACOs, PCMH, population health management, predicting excessive length of stay, anticipating readmission risk, chronic disease management, identifying care gaps)
  • Transforming care (integrative medicine, service line design, care redesign, patient engagement and commitment, predicting complications and other impacts to patient safety, enhance clinical decision support through prescriptive analytics)
  • Improving performance (design and execution of experiments, innovation, labs that help customers explore, reducing variations in care)

Key Point

Analytics can transform by linking effort to strategic aspirations.

A graphic showing strategic opportunities for healthcare analytics.
Figure 1. Strategic opportunities for Healthcare Analytics

In sum, we believe the biggest opportunities for analytics in healthcare are those areas where the need involves:  

  1. An integrated, unified view of data;
  2. A going beyond description and discovery of the unknown to prediction, prescription, and optimization; and
  3. A problem important enough to be considered urgent and solvable.  

All three are required and not merely a "nice to have” to ensure that analytics has its permanent place in healthcare.

Healthcare Analytics Is About Creating Value

We are already overwhelmed with data; what we need is information, knowledge and wisdom.

Dr. John Halamka, CIO Beth Israel Deaconess Medical Center

As we saw in the previous section, outcomes are a critical component of analytics in that we must create something that is worthwhile.   As we extend our discussion of analytics, it’s hard not to talk about value creation. After all, many view analytics in the same light as the countless Information Technology (IT) projects that they see fail. While there are lots of reasons why projects fail (see for example (Petranka, 2017).) As Dr. Jeremy Petranka notes, failed projects often relate to the fact that the linkage between the undertaking and organizational strategy is absent.

While the reasons vary, we find in our advisory work that projects often fail when the fundamental value proposition was never fully realized. That is, the promise of the value to be “delivered, communicated, and acknowledged” was never fully realized. 

Value has many definitions including the net result of [benefits – costs]. We prefer to look at value with a quality component, depicted as:

Value = (Quality + Outcomes) / Cost

The quality component is important in that analytics without quality presents risk, uncertainty and unrealized potential. Quality comes in the form of robustness, repeatability, reliability, and validity.  When the ratio is less than one – that is, when costs outweigh the quality plus outcomes – then we have failed to meet our fundamental value proposition (the reason we did analytics in the first place.)   However, full value isn’t enough. Instead, we expect analytics to be a “force multiplier” for organizations and as such, should be far greater.  Of course, return on investment (ROI) for analytics isn’t the only measure of value as we need to consider other things such as mitigating risk, avoiding missed opportunities, or committing ourselves to improving the lives of patients.

In summary healthcare analytics is about creating value in that we achieve outcomes through the scientific processes we refer to as the analytics lifecycle. This requires a multidisciplinary approach to achieving value.

Petranka, Jeremy. (2017, June 20). Duke University Fuqua Faculty Conversation. Retrieved from

Key Point

The key proposition for analytics is in its value creation.

Healthcare Analytics Is About Discovery

Even more than what you think, how you think matters

Atul Gawande, Author and Surgeon

If Business intelligence (BI) is about knowing the knowable, then analytics helps us with knowing the unknown. As the adage goes, “we can never know something until we discover it.” The power of analytics is that it supports discovery. We use our skills of reasoning and sense-making to unearth patterns in data and are, in fact, often pivoting between deductive and inductive reasoning when we “problem solve” with data. Generally, we find ourselves approaching the problem in the following manner:

1. Formulate a question or problem statement

2. Generate a hypothesis that is testable

3. Gather / generate data to understand the phenomenon

4. Analyze data to test the hypothesis / draw conclusions

5. Communicate results to interested parties or take action

However, when we are unsure of what we know and the what we don’t know, we can use an alternative approach:

1. Gather / generate data to understand some domain or phenomenon

2. Make sense of the phenomenon through data discovery and visualizations

3. Begin to formulate testable hypotheses

4. Analyze data to test the hypothesis / draw conclusions

5. Communicate results to interested parties and act

As you can see, this is a fundamental change from traditional data analysis approaches where inductive reasoning and exploratory data analysis provide a means to form or refine hypotheses and discover new analytic paths.

Healthcare Analytics Is About Change

Don’t rely solely on data to drive decisions; use it to help drive better leadership behaviors.

John W. Boudreau, PhD, Professor and Research Director, University of Southern California’s Marshall School of Business and Center for Effective Organizations

I know of few people who like and embrace change. Yet change is inevitable and perhaps the only constant organizations should count on. The impetus for change can come in many forms but for healthcare it is often a crisis such as disaster, deregulation, declining profits, government mandates, failed systems or public health scares.

We began this document with words like outcome and impact. For us, these are key to analytics – driving change to improve outcomes and creating value. If we look at some of the most celebrated cases of analytics in healthcare we see evidence of outcomes:

  • Cleveland Clinic uses advanced forecasting models to schedule operating room staff
  • Dignity Health predicts risk of sepsis in high risk patients
  • Carolinas Health System predicts risk for readmission in real time
  • UPMC predicts patients who need greater high-touch, preventative care
  • Intermountain Health modifies treatment guidelines for labor and delivery
  • Duke University Health System uses geospatial analytics to design care delivery systems and support community health partnerships
  • University of Utah predicts outbreaks of Respiratory Syncytial Virus (RSV) three weeks before it happens for high risk patients

In each of these cases, whether we use analytics to improve patient care, spur innovation, or redesign care delivery processes, change is inevitable as the impact to the organization necessarily involves altering how work gets done.

Change has afforded entire industries the opportunity to transform how they operate. Take, for example, the case of the Oakland A’s as depicted in the book Moneyball (Morris, 2014) and their use of analytics to drive competition. Major League Baseball has been transformed by analytics, and its decisions around players will never be the same. 1

While no one sets out to intentionally do this poorly, we see these types of scenarios play out in our experiences every day. At the heart of any successful change, whether transactional or transformational, how you lead the change effort will largely determine success.

1 Note: There are those that would certainly like to believe that Moneyball was all wrong and that gut has its place. See for example but the impact of analytics remains forever embedded.

Key Point

We do healthcare analytics to support change in an organization.

A case study in analytics and change management

Writing icon

There are so many reasons why analytics projects fail, but failing to plan for how the changes will affect how people do their work has to be near the top of the list.  Take a recent example of a real-world scenario in a large hospital system where a predictive model was being deployed to help clinicians prioritize patient cases. With a few changes, here and there to avoid potentially embarrassing anyone, we’ll highlight the key points:

  • The hospital’s infection rates were significantly higher than other hospitals based on their size and capabilities
  • The finance and operations teams identified this as having a significant financial impact on the organization as hospital acquired infections required treatment that was not being reimbursed by the payers
  • The goal of the model was to identify “at risk” patients for a specific hospital acquired infection
  • A physician champion was selected to work with the analytics team to develop a predictive model

In the context of a change management strategy, this had all the trappings of a great “change” project.

  1. There was a shared sense of urgency (and purpose)
  2. The team had developed a “guiding coalition” that included a highly respected and influential physician
  3. The model was developed with regular checkpoints along with way with and included feedback regarding the importance of several clinical indicators
  4. The model was developed using solid testing and validation principles

The sad news is that while the model predictions were outstanding, the model ran automatically with little impact to the infection rates. The reason this initiative failed was multi-faceted but there were two primary challenges that the project faced: 

  1. This was one of literally a hundred things that the physician champion had to do on a given day (competing priorities)
  2. There was a disconnect between what the model produced and how the clinical care teams operated.  That is, there was no clear linkage between the model output and the action that was required to effect change.

All too often we see smart people doing great work that ends up not living up to its potential or expectations, or a few heroes putting the project on their backs and forcing it through, despite the lack of a great change management strategy. This is a dangerous demotivator to staff and causes organizations to lose some of their best, most capable people.

Leading Change

The proper leadership of any change initiative is critical.  Without it, it begs the questions of whether the change is worth doing.  Change management assures that people, processes and technologies are in optimal working condition.  All too often, there is focus is on the technology without regard to what the impact to people and process might be – or worse, assuming people are smart enough to figure it out.

The management of change, “Change Management”, is more than just “training” or a communications plan, but rather a systematic approach that helps facilitate the transition of organizations and people from a current to a future state.  In our example above, the desired future state was one in which hospital acquired infection rates were significantly lower than the current state – a reasonable and measurable outcome.

It is unfortunate that the phrase “change management” has such as negative connotation for some people since there are such great examples of organizations leading transformation.  For a great example of change leadership in action, look at Google’s Project Oxygen.

Click above to watch the video

Change management is not about forcing change, but rather helping people prepare for the change (awareness), understanding the rationale and impact (knowledge), helping them with the tools that allow for skills and competency development (adoption) and supporting the integration of those skills (commitment.)

If we truly believe that analytics can have a transformational impact on organizations, then the “transformation” must be managed – or as John Kotter would advocate, “Lead” the change. (Kotter, 1996)

The process that we follow to manage change should depend on the impact of the change and should be scaled accordingly. While there are a number of change management methodologies available, both commercially and otherwise, we believe that the eight step process that John Kotter characterized in the seminal work Leading Change (Kotter, 1996) stands the test of time. 

A systematic approach is required that will facilitate transition from a current to a future state. We define organizational change as creation of a mindset that helps individuals transition between technology and processes in the context of organizational culture. Change is realized only when people do their jobs differently and make their own transition to a future optimized state. Change management by force does not work.  Successful approaches instead include the Giving Voice to Change:

  • Awareness: Helping people prepare for change
  • Knowledge: Understanding the rationale and impact
  • Contribution: Giving adopters a voice in the foreseen change
  • Adoption: Helping staff develop skills using new tools
  • Commitment: Integration of new skills into workflows

Key Point

Analytics drives change. Leaders lead the change.

For our work in healthcare analytics, we have developed six best practice areas that include:

Change Management Best Practice Areas Examples of what this means in practice
Create a shared change purpose
  • Create an aspirational future state
  • Establish clarity around why you intend to achieve this
  • Analyze who will be impacted
Establish a visible and engaged leadership coalition
  • Conduct stakeholder analysis
  • Collaborate to define the impact of the change
  • Engage change champions at the extremes (tipping point leadership)
Facilitate stalwart engagement and communication
  • Implement a robust communications strategy
  • Build awareness, knowledge, Contribution, Adoption, Commitment (see Giving Voice to Change above)
Support strong individual performance
  • Conduct knowledge & skill assessment
  • Deploy effective training & knowledge management strategies
Build a supportive organization and culture
  • Build a change network
  • Business process alignment
  • Manage change progress
  • Address change readiness
Create a measurement strategy
  • Design change governance
  • Track progress and issues/report status
  • Measure progress (bright spots, laggards)

Table 1: ThotWave’s Change Management Best Practice Areas

Note that not all projects will require every one of the processes contained within these six best practice areas. The processes should be right-sized according to the breadth, depth, impact and criticality of the change.

Competency model documents

Get Started With the Healthcare Analytics Competency Model©

The first of its kind, ThotWave's Healthcare Analytics Competency Model© is available to organizations and individuals, designed to help you or your team assess your analytics standing and pinpoint your development needs. Learn about the Competency Model and register for the Talent Development Program to get started.

Get Started

Efficient and Effective Analytic Programs

Everyone's talking about it, no one really knows how to do it — everyone thinks everyone else is doing it, so we all say we're doing it.

Deb Gage, president and CEO of Medecision, likening population health to teenage sex during a panel at the Becker's Hospital Review 5th Annual CEO + CFO Roundtable

In the previous section, we examined the “why” of healthcare analytics. Here we witnessed the power of transforming healthcare through programs that support the discovery process which help create value through organizational change to produce improved outcomes. So why don’t all organizations simply do more with analytics?

To answer this, we must first talk about what’s involved in analytics and secondly, what are the challenges. To help frame the conversation, let’s first talk about the difference between effectiveness and efficiency as this will help understand the potential areas for opportunity.

The Difference Between Effectiveness and Efficiency

Effective (adj.) – Adequate to accomplish a purpose; producing the intended or expected result.

Efficient (adj.) – Performing or functioning in the best possible manner with the least waste of time and effort.

In healthcare organizations, we usually seek to increase and improve the efficiency of data and analytic operations (data flow and management, business analysis, discovery, results interpretation, presentation and operationalization.) As is true with most organizations, we are faced with limited resources and would prefer to maximize the use of each of these resources – regardless of the type of resource (money, technology, time, human capital, quality.)

However, by pursuing efficiency at all costs, some of these companies are missing a valuable chance to take a step back and look at their overall effectiveness from a big picture perspective. That is, are we being effective in our data management efforts – are we creating data that is accessible, usable, trustworthy? Are we being effective at storytelling – turning data facts into insights that resonate with our stakeholders? Are we being effective at selecting projects - those that have the potential for greatest impact – and are we seeing these through to how clinicians and front line business users will use these to effect change?

Key Point

Effectiveness is doing the right things. Efficiency is doing them well without waste.

Analytic leaders strive to find an optimal balance between effectiveness and efficiency that ties directly back to value that we address above. One way to illustrate the difference between efficiency vs. effectiveness is with a common 2×2 grid shown below:

A graphic showing the difference between efficiency & effectiveness
Figure 2. 2 x 2 Effectiveness and Efficiency Matrix

The goal for most analytic organizations is to pursue the top right box – going after the right goals and being efficient, by making use of our analytic resources: people, processes, technology and data, not wasting time, and having better alignment and collaboration of between employees. While most organizations have the best of intentions in that they know what they want to achieve, often the silos make true efficiency and effectiveness allusive.

Often times it is easy to focus on efficiency as it is somewhat easy to measure (costs, time, effort) and is perceived of as more “controllable.” A great case study on the potential pitfalls of the easy path can be found in a classic Harvard Business Review case study describing the death-spiral of Bridgeton Industries, an automobile manufacturer.

Understanding the Healthcare Analytics Lifecycle

It is a capital mistake to theorize before one has data.

Sherlock Holmes in “A Study in Scarlett” by Arthur Conan Doyle

As a matter of practical explanation, consider the differences between effectiveness and efficiency in the analytics lifecycle. Below we note the various stages and activities in the pursuit of creating analytic products. Note that analytics product management is an overarching set of activities that help manage and support the analytic teams and their deliverables.

A Well-Executed Analysis Project Follows an Agile Life Cycle

Figure 3. The Healthcare Analytics Lifecycle (click to enlarge)

For each data product that is created, we go through the various stages of analysis, exploration, analysis, interpretation and operationalization. Note that this may seem linear, but we often go back and revisit activities to refine, clarify, and elaborate. Not all projects are created equal in that there are some projects may not require all the activities outlined here.  For example, for business questions that demand low fidelity, you may find that a rough order of magnitude answer may be sufficient.  Similarly, where the key stakeholders have high data literacy, the effort towards explanation and story framing may be less critical.  As we look to operationalizing data products, we may be find that some projects never make it that far (nothing interesting here to see) or the change requires effort beyond the value that it creates or is impractical to realize the true benefit.

Key Point

The analytics lifecycle is ever fluid and changing.

The difference between effectiveness and efficiency for analytic teams can be summed up succinctly – being effective is about doing the right things to create value and improve outcomes, while being efficient is about achieving maximal productivity with minimal waste or expense.

In the table below, we highlight some common examples of effectiveness versus efficiency in the analytics lifecycle.

Stage Effectiveness Efficiency
  • Prioritize the right problem to solve
  • Demonstrate empathy for the customer's challenge
  • Accurately define the problem, it's impact and value
  • Translate the problem into a question that can be answered with data
  • Manage and deliver on stakeholder expectations
  • Expend the right level of effort for the problem
  • Collect and manage requirements competently (for clarity and reuse)
  • Economically utilize stakeholder time and knowledge
  • Manage project
  • Translate the problem into the appropriate data sources
  • Accurately assemble the data
  • Assess the value of the integrated data
  • Formulate testable hypotheses about the relationships found
  • Capably accessing and acquiring the data
  • Proficiently integrating the data
  • Utilize the most appropriate tools and technology (eliminate waste, avoid redundancy/ duplication/ rework)
  • Identify opportunities for enrichment
  • Test hypotheses for their validity
  • Eliminate spurious relationships
  • Competently utilize statistical and visualization software
  • Create summarized information/ insights
Leverage and Communicate results
  • Create stories that resonate and influence
  • Select the most appropriate mediums for communication
  • Master storytelling for influence
  • Lead change management
  • Assess generalizability
  • Imbed into workflow
  • Assess product durability and opportunities for improvement
  • Create storyboards and visualizations
  • Document findings and other knowledge for reuse
  • Validate analytic models, calibrate/ retune, maintain and retirement
Analytic Product Management
  • Right-size the analytic processes depending on expected value, and breadth and depth of impact
  • Define program evaluation and measurements strategy for analytic products
  • Manage enterprise portfolio
  • Efficiently manage resources
  • Capture knowledge and opportunities for reuse

Table 2: Differences between effectiveness and efficiency in analytics

Perspectives On Effective Healthcare Analytics

Never confuse motion with action.

Benjamin Franklin, American Politician/Inventor

In Table 2, we outlined some examples of what we see as the difference between effectiveness and efficiency in analytics. However, note that there may be wildly different opinions about what constitutes efficient and effective analytic programs depending on the perspective.

In healthcare, we often see multiple, competing perspectives that can include both internal and external stakeholders.  Analytic leaders and analysts can have a one-sided view of the analytics lifecycle and view analytic process efficiency and effectiveness in different time frames and in different ways. This often differs from the “decision lifecycle” that other stakeholders utilize when attempting to capitalize on the analytic products. In short, each of these stakeholders ask different questions about the importance of analytic processes and their impacts (Table 3). Confusing the needs (aka incentives) of these stakeholders often results in misunderstanding, priorities and effort.

Understanding that there are these different perspectives can be useful in both service design and how people consume data products.  One of the tools that we use to help design both analytic products and processes is design thinking.

Key Point

What I think may not be true – for me or anyone else.

Role Question Time Horizon Incentives / Outcomes
Analytic Leaders Are analytics having an impact (creating value?) Quarterly, annual, 3-5 year Quality, benefits, costs
Analysts How can I get better (learn, grow, become more proficient?) Days, weeks, months Time spent on interesting, relevant, actionable, purposeful work
Front Line Data Champions What can I use to impact outcomes? Minutes, hours, days Integrated, actionable view of data
Executive Leadership How can we transform healthcare? Mid-term, long-term Long term implications for operations, financial and clinical domains. Driving innovation, learning and creation of new opportunities
Patients What does this "risk score" mean to me? Today, tomorrow, old-age Data that makes sense/ resonates, drives behavior change and is actionable
Third Parties How can I get better visibility into what's going on? Weeks, months, quarterly Data that is accessible, accurate, at the right level of detail and in a form that I can use

Table 3: Stakeholder Perspectives on Analytics Effectiveness and Efficiency

Challenges to Effectiveness and Efficiency in Analytics

The comforting conviction that the world makes sense rests on a secure foundation: our almost unlimited ability to ignore our ignorance.

Daniel Kahneman, Psychologist, Author and Nobel Memorial Prize in Economic Sciences (2002)

There are several potential challenges we face in healthcare analytics that range from implementing change to ensuring that we have the right skills and competencies to deliver on the organizational capabilities.

In Figure 4. Common Challenges in Healthcare Analytics below, we outline some of the most common challenges in healthcare analytics.

Key Point

Design thinking can be used to help design how analytic services are delivered as well as how data products are consumed.

Figure 4. Common Challenges in Healthcare Analytics (click to view interactive map)

While this graphic doesn’t capture all potential challenges with healthcare analytics, it certainly represents the most pressing of those challenges and impact our ability to be both effective and efficient in the delivery of analytic products to organizations.

At a higher level, this can be seen in the struggles that every analytic leader faces:

  • Strategy versus execution
  • Culture and politics
  • People and process
  • Data and technology

Key Point

Analytics challenges are rarely technical in nature.

Actioning Analytics – A Futurist’s Perspective

Why Operationalization of Analytics Fails

Once we know something, we find it hard to imagine what it was like not to know it.

Chip & Dan Heath, Authors of Made to Stick, Switch

As if the list of challenges above was not overwhelming enough, we also know that transformative change is hard. Even if you have the right strategy, the right data, the right people and the right technology, we can still fail. Below, we have outlined what we believe as some of the reasons why we fail to operationalize analytics.

Key Point

All things being equal, we can still fall short if we fail to understand the consumer and their application.

Why we fail to operationalize analytics

  • We produce data, not insights – organizations are overwhelmed with data as evidenced by massive number of reports and dashboard that continue to proliferate. Democratize the data, push data to the edges of the organizations where they can be vetted, enriched and utilized.
  • Insights are untimely – often the “time to decision” is in minutes and hours, yet analytic insights arrive well outside the decision-making window.
  • Decisioning process is not understood - The analytics lifecycle must be in sync with the decision lifecycle. When we fail to empathize with how people use the data to support decisions and problem solving, we set ourselves up for failure.
  • Technology is too general – When technology solutions (data warehouses, dashboards, reports, etc.) are implemented, we often fail to consider the how they can best support organizational decisions.
  • Trusted partners - Producers and consumers of analytics often do not have a strong relationship that allows the analyst to understand or anticipate the true needs.
  • Data fluency – there exists in most organizations a gap in basic data and analytics literacy. Data literacy must be evangelized and championed throughout the organization.
  • There is no silver bullet – in general, we fail to collect deep, non-quantitative, insights about patients and processes that could help position analytic insights into the proper context.
  • Thinking small – while big-bang approaches are ill-advised, we cannot continue to report on the past. Our mandate is to apply analytics where we can influence change – including transforming the business and delivery of patient care.
  • Poor stewardship – we fail to inspire confidence when we provide inaccurate, incomplete data that cannot be replicated. We must ensure quality in every data product we develop.
  • Culture of Have’s and the Have-Nots - when we fail to push analytics to the edges of the organizations, we fail our organizations. We need to create voracious consumers of analytics – those that solicit data, ask questions, push for insights, and take actions. We need to model the use of analytics and support our data champions.
  • Modest predictive accuracy – analytics are only as good as they help us influence change. Analytic results that perform no better than a nurse’s intuition (Spidey-sense) or produce predictions that are not impactable contribute to the negative narrative about ineffective analytics.
  • Unplanned or unintended consequences – when we fail to understand the how models are used in the context of the business and the systems of interconnectedness, we are susceptible to analytics being used for the wrong things.

What’s Required for Successful Analytic Change

An idea not coupled with action will never get any bigger than the brain cell it occupied.

Arnold Glasow, American Businessman

Healthcare is still in its infancy when it comes to the use of analytics.  On the backs of heroes, we see the potential through leading healthcare systems and these case studies will continue to drive our future aspirations.  But how do we effect change? How do we “action” analytics in our organizations? 

The opportunities can be seen in the challenges that we face.  It’s up to us to apply what we know about behavior change, motivation, problem solving, design thinking and so on to effect change where we can influence.

Key Point

Analytics requires a multidisciplinary approach.

While organizations differ in the specifics, to us it comes down to these set core elements:

Developing strategic intent and clarifying analytic aspirations (organizational capabilities)

Define the data strategy (outlining what data to collect and how we govern it)

Build capacity to capture and analyze data (including nurturing and developing analytic talent)

Implement the right solutions (people, processes and technology) 

Apply analytics to the right problems

Create an insights-driven, culture of innovation that reacts timely to insights

Note that many of these elements rely heavily on leadership, people, processes and data. Technology is an enabler and should not be confused with being “the” solution.

Analytic Competencies for Healthcare

Analytics competency relates to the knowledge, skills, abilities and disposition required to successfully turn data into actionable interventions.

Greg Nelson, Founder and CEO, ThotWave

As we have written elsewhere (see The Elusive Data Scientist: Real-World Analytic Competencies), a major challenge in building an analytics team is defining the blend of skills that suit team mission and the enterprise culture. To understand how to develop staff to achieve future capabilities, we have developed a competency model that maps analytic functions, skills, and competencies to specific organizational roles (Figure 5).

Key Point

We must create data champions throughout the learning health system.

A graphic showing healthcare analytics competency domains
Figure 5. Healthcare Analytics Competency Domains

Developed through a process of workplace analysis and expert knowledge, our model includes nine domains of knowledge, skills, and behaviors that need to be demonstrated within a healthcare analytics team. It is noteworthy that many of the competencies we recognize as being critical for analytics have several non-statistical and non-technical features. This is because our model seeks to address the entire analytics lifecycle.

To discover the people and skill you need on your team in order to develop the best analytics strategy for your organization, download our Job Families Executive Guide.


Actioning analytics is perhaps of the biggest challenges we face in healthcare. As we have outlined, understanding the barriers that get in our way and thinking about designing for change is critical to ensuring that our analytics matter. Leading change is part of our role as analytics leaders – to help support and influence change that helps facilitate the transition of organizations and people from a current to a future state. This necessarily includes best practices, process, tools, and techniques to help deal with the people, process and organizational side of change.

Modernizing data strategies and developing analytic mindset supported by technology changes is an important area where change management should be considered essential to success. While there is no simple solution or off-the-shelf answer, change management does not have to be expensive. It should be right-sized in accordance with the breadth, depth, criticality and impact to the organization.

Of all the changes one can make in healthcare analytics, one of the best places to start is to look inward and ensure that you (the reader; presumably the manager) has the right team with the right competencies needed to make analytics actionable. To do this, you need a true north—a compass that guides you. If you’re interested in helping evaluate your team’s competencies and readiness to make analytics actionable, feel free to schedule an appointment and chat with one of our experts.

Competency model documents

Get Started With the Healthcare Analytics Competency Model©

The first of its kind, ThotWave's Healthcare Analytics Competency Model© is available to organizations and individuals, designed to help you or your team assess your analytics standing and pinpoint your development needs. Learn about the Competency Model and register for the Talent Development Program to get started.

Get Started

About ThotWave

ThotWave Technologies, LLC is a Chapel Hill, NC-based advisory firm and market leader in healthcare analytics education. ThotWave’s primary focus is to help health organizations mature their use of data and analytics.  Through the strategic development of analytical aspirations, best practices for organizational design, purposeful talent development strategies and improvements to analytic lifecycle processes, our products and services are designed to help leaders achieve breakthrough performance in their use of data and technology for decision making.

Sign Up to Read the Rest of The Manager’s Field Guide to Actioning Healthcare Analytics