Intuition-Based Healthcare or Data-Informed Healthcare — Which Would You Prefer? (Guest Contributor: Jim Adams, Advisory Board)
But what if all decisions in life were made on intuition? What if a home builder made decisions about the weight the supporting beams could hold because of his gut instinct? Or what if a doctor told a patient he or she had a heart problem based on his gut reaction to the presenting symptoms? Only later, after tests, would the patient discover his or her heart was fine but high anxiety levels were causing the symptoms to resemble a heart attack.
Even though I’m not a clinician and I’m not on the floors saving lives, I appreciate the value of data and how it can be used improve healthcare. This is why I am a big proponent for using data — not just in healthcare but in everyday life.
For example, I am somewhat of a quantified selfer. I use technology like Fitbit and smartphone apps to measure and monitor my activity levels and other health-related data to make sure I stay on track. This helps when I’m traveling because it’s really easy to end the day by eating dinner and responding to emails. But when I look at the Fitbit or app numbers, I may see that I need to go for a walk to get in some steps. Even though my instinct makes me feel tired, the walk overrides my original feelings of exhaustion and gives me more energy to tackle those emails.
The healthcare industry is in a similar situation by beginning to use data to make quicker and better decisions. Yet even though healthcare has come a long way, there’s still a long way to go because of the many challenges healthcare leaders are facing.
Challenges Healthcare Leaders Are Facing
Leaders in healthcare understand the status quo of the industry is no longer an option. But it’s difficult to know which improvements need to be made and when to make them to gain the most benefit at a manageable level of risk for their health system.
They’re also facing other challenges, such as implementing advanced analytics solutions that may be able to help them decide which improvements to make or setting up good data governance or business intelligence (BI) governance. And then, even if they’re able to set up these programs, they still need to find the right technical and analytics experts to make use of the programs.
Changing a health system’s culture to be more enterprise focused and data informed is another hurdle. This is because there are still a lot of data silos and turf wars. While this worked well enough in the past, it won’t meet the needs for more holistic, integrated, and coordinated health and healthcare services.
Healthcare leaders are also facing the traditional instinct approach to decision-making — the most profitable service lines and best physicians may be identified more on gut feelings than data. While it may seem like healthcare has good evidence for how patients are treated, the reality is that there’s really only strong evidence for about a quarter to a third of treatments provided by clinicians. The rest is based on what the individual thinks works based on his or her personal experiences and the experiences of closely-associated clinicians.
Opportunities to Improve Healthcare with the Right Data
Despite the many challenges, there are also great opportunities for healthcare to improve — both in terms of bending the cost curve and improving quality. But to reach these goals, healthcare leaders need to be able to better understand, monitor, and manage their businesses — both the clinical and financial aspects — by knowing what data to use or what data to capture.
Data is absolutely critical for understanding a business. Not all data, however, will be useful. I believe in the adage from Albert Einstein, “Not everything that matters can be measured and not everything that can be measured matters.” Since not all aspects of complex decision can be accurately quantified, it’s more beneficial to use data as a tool to inform decision-making rather than as the only input into the decision. For example, when I was at Gartner, I reported to a person who had developed a massive spreadsheet containing various productivity metrics for the analysts. He used the spreadsheet as the sole factor for decisions regarding promotions and raises. I never could fully convince him that, while the selected productivity metrics were important, that other job-related factors should also be considered. In short, I advocated data-informed decision-making and he practiced exclusively “data-determined” decision-making.
Using Data and Advanced Analytics to Drive Organizational Change
What’s really needed is a way to see the whole picture from a holistic view by tracking outcomes and segmenting results to understand what’s really going and how the pieces interact. And basic analytics and BI are no longer sufficient for healthcare’s complex requirements.
Health system leaders recognize that data can help solve healthcare’s problems and also that the data can be more complex than the traditional dashboard or scorecard. But to make the most of the data, health systems need to know how to implement the more advanced forms of analytics, such as predictive and prescriptive — but these types of analytics require different tools, data, skill sets and even “mind sets” to be effective.
The benefit to learning how to make the best use of these analytics systems is tremendous: with sophisticated analytics, clinicians could have access to the right patient data and clinical knowledge to better manage patient populations. This data-driven approach to healthcare means the chance of getting the right diagnosis and right therapeutic treatments becomes much greater.
Healthcare has become so complex it’s virtually impossible — particularly with patients with multiple chronic diseases — for clinicians to continue to try and practice medicine in their heads. If analytics can help provide cognitive support at the point of clinical decision making, clinicians can make huge gains in the accuracy and completeness of the diagnosis and relevance of the treatment, particularly for the most complex of patients. More cases would move from trial and error medicine to evidence-based or more personalized medicine.
Personalized Medicine: My Hope for the Future
Right now health systems are focused on providing effective population health management for their patients. But once health systems have the right analytics and BI tools along with experts to analyze the data, I hope that in the not-to-distant future medicine can move beyond managing broad populations of patients based on evidence-based guidelines (i.e., what works best for most patients) to managing subpopulations of patients or individuals based upon their unique characteristics, needs, and preferences.
Subpopulation health management would provide more personalized care. For example, under traditional medicine, I might receive the same diagnosis as two other patients and we’d all receive the same medication at the same dose. But the traditional approach doesn’t take into account the variations in how people metabolize or respond to the medicine. The medication might work well for one patient but not work for another patient. Then it’s even possible another patient may have an adverse reaction to it.
The traditional approach to medicine is practiced this way because there’s enough evidence to suggest that a particular population of patients with this type of condition will respond well to a particular medication at a specific dose. In reality, however, the results vary widely, even for patients with the same disease. This is sometimes referred to as the heterogeneity of treatment effects. With sophisticated analytics to pour through data-intense records, however, it could someday be possible to personalize each patient’s treatment according to their unique genotype, phenotype, environment, lifestyle, preferences, and other key factors.
Think Globally, Act Locally
Given the rapidly escalating and changing requirements in healthcare, health systems are struggling with their analytics and BI initiatives. While it’s tempting to look only at short-term needs, they need to also think globally — and plan for their overall longer-term strategy as well.
Balancing these two different sets of needs will be critical. Ending up with a bunch of point solutions that don’t meet the organization’s escalating longer term needs could be a waste of time, money, and resources. On the other hand, just planning for the longer term and not delivering on shorter-term requirements is not acceptable. That’s why it’s critical for health systems to develop a longer term strategy and framework within which to develop and implement their shorter term solutions.
The nature of developing analytics solutions to meet rapidly evolving business and clinical needs has been and likely will continue to be a game of “two steps forward and one step back,” but having a longer term vision and plan can minimize the disruption and keep health care organizations from having to scrap or redo too much of their BI infrastructure as needs and tools evolve.