From Installed to Stalled: Why Sustaining Outcomes Improvement Requires More than Technology
Igniting change and outcomes improvement within a healthcare system requires three unique elements. It’s well accepted that an analytics platform, with its enterprise data warehouse (EDW) and advanced analytics applications, plays a pivotal role in this process. But two other elements—adoption of analytics across the enterprise and converting data-driven discoveries to best practices—are essential for sustaining healthcare outcomes improvement.
In other words, effective analytics means integrating the project beyond the IT department and into the feedback loop of those who actually provide care to patients. Let’s explore these two elements and the common challenges organizations face as they attempt to achieve analytics adoption and implement best practices.
Outcomes Improvement Requires Adoption and Best Practices
The analytics platform is crucial for driving any outcomes improvement initiative, especially over time, but here are a few scenarios that show why it must be supported by an adoption methodology and best practices implementation:
Without adoption: Focusing on best practices and analytics is insufficient. Without a focus on adoption, the organization ends up with solutions that never quite scale to create a lasting impact across the organization as a whole. We’ve seen various symptoms of organizations that don’t have a robust adoption methodology. For example, take research-focused organizations. The good intention here is to use research to feed the creation of best practices. While research surely has a strong position in healthcare, as it should, organizations that are purely best-practice-centric are prone to academically appealing ideas with little regard for practical application, preferring to publish papers on insightful topics that are difficult to implement and with findings that are difficult to maintain. Alternatively, organizations focused squarely on the technical and analytical aspects of improving outcomes tend to manifest symptoms of large report queues and dashboards. The temptation with these organizations is to focus all efforts on building and automating, because of the mentality that says, “if we build it, they will come!” But without input and buy-in of the right stakeholders, the organization creates lots of technical solutions that end up on the shelf. Finally, there are organizations that are able to adequately merge the two other systems of analytics with best practices, but don’t have the expertise to get widespread adoption. They create successful “science projects,” pockets of excellence that are limited, and they are unable to rollout improvements across the organization.
Without best practices: If the organization implements an adoption system and analytics, but doesn’t include best practices, it may end up with consistent practices, but not necessarily the right practices, i.e., a standardized status quo rather than continuous improvement. We affectionately call this the “paved cow paths” approach because processes are automated and adopted, but not truly improved. Interestingly, an improvement initiative could have very well implemented a standard of care that was sufficient or even groundbreaking for its time, but the practices are now out of date and the old processes and practices stayed in place. To avoid these symptoms, we advocate permanent teams consisting of individuals on the frontline with deep subject matter expertise who recognize the need to be innovative and who are always looking for ways to improve their domain. Another symptom could be clinician’s disengagement of management’s “flavor of the month,” where there is missing evidence in the data and missing proof of best practice. Either way, the organization would ultimately be outperformed by competitors dedicated to finding the best ways to deliver care.
Without analytics: Finally, if the organization implements best practices and a system to spread adoption, but ignores analytics, it has no way to measure the effectiveness of its efforts. Likewise, it tends to take on improvement projects based on who shouts the loudest (for example, a vocal physician who wants to improve a certain process) rather than focusing its resources on areas with the most variability in quality and cost. Organizations without a robust analytics system experience unsustainable improvements after only a couple of projects because they don’t have the automated capabilities to scale.
Challenges of Implementing the Two Elements
Most organizations understand the need for an analytics system, but many fail to implement an adoption and best practice system to feed and nurture the technical aspects of an improvement project. Let’s touch on the challenges of implementing these two critical systems and then I’ll reveal the characteristics of organizations that implement them successfully.
Technology cannot improve outcomes without the buy-in, onboarding, training, and enthusiasm of the frontline providers and staff who will be working with it every day. Without adoption, even the best data-driven improvement projects can stall.
Health Catalyst is working with Piedmont Healthcare, an integrated delivery system with six hospitals and close to 100 clinics in Georgia. It is a perfect example of how adoption is so important to process improvement, and it’s a great success story of sepsis care improvement. Before it could standardize improvement, Piedmont had to organize a cross-functional workgroup team consisting of innovative clinicians and frontline staff, supported by analytical and process improvement personnel. As the workgroup built out the solution, they were constantly feeding the improvements up the chain so that those improvements could be vetted and implemented across each hospital in the system.
Then Piedmont brought in Health Catalyst’s Sepsis Improvement application, but before it could be trusted by the physicians, nurses, and financial staff who would be using it, Piedmont went through a data validation process to gain a level of comfort with the data that was being presented. Data validation can literally paralyze a project when there is too much focus on getting things exactly correct, so the workgroup was careful to do just enough validation to garner a high level of ownership and engagement, aka adoption.
Based on the data and their own knowledge of clinical workflow, the workgroup made initial decisions about where to focus improvement efforts. This group was also tasked with determining how to implement best practices that would drive the desired improvement. To gain buy-in and receive valuable feedback, the workgroup presented its recommendations to peers. This physician-led process ensured that the perspectives of all key stakeholders informed any changes to a clinical or operational process. It also gave clinicians an opportunity to “fingerprint” new processes—to give their feedback and make the process their own. This is more than just coming up with goals and metrics; this is about using the knowledge and expertise of individuals within the workgroup to drive out waste and improve processes. For example, Piedmont wanted to do more than simply track how well they were responding to system alerts, so we put the data to work by establishing a baseline, identifying parameters for improving alert sensitivity, and implementing triage and system alert training. Piedmont also wanted to improve its 3-hour and 6-hour bundle compliance, so we used the EDW to track order set utilization and standard notes and assessments. None of this detailed work could have occurred successfully without the participation of these innovative end users on the frontline.
Over the years, we have seen these physician-led workgroups achieve incredible, sustainable quality improvements. On the other hand, we have found that when IT drives the quality improvement project, it flounders. Clinicians tend to view such projects as imposed from the outside by people who don’t understand clinical workflow. Physician leadership and enthusiasm are absolutely essential to driving adoption system-wide.
Implementing Best Practices
In U.S. healthcare, spending waste is estimated at $1 trillion a year. Best practices are a key to transforming healthcare and reducing waste. So why is it so hard to implement best practices throughout an organization? Why does it take an average of 17 years for a newly discovered best practice to become standard practice in healthcare?
One of the biggest reasons is simply a matter of human nature: most of us are comfortable with the status quo. I often hear people say that physicians like to do things the way they’ve always done them and don’t like to change. In truth, comfort with the status quo isn’t limited to physicians. Looking at the diffusion process of new technologies and ideas in society as a whole, it’s interesting to note that fewer than 20 percent of people tend to be innovators or early adopters. The rest of us are slower to adopt new practices.
It’s hard for us to overcome personal habit and preference. From the organizational perspective, it’s even harder to overcome institutionalized inertia. For this reason, it is essential to identify influential physician innovators and early adopters who can help spread information about, and enthusiasm for, a new best practice.
Referring again to the Piedmont sepsis improvement project, we also consider it a success story of best practices implementation. In this case, the improvement team was able to use the Sepsis Improvement application to show executives, managers, and frontline staff precisely how well the health system was adhering to the best practice 3-hour bundle, responses to early sepsis alerts, and how this performance directly impacted length-of-stay, mortality, cost, and patient outcomes. When clinicians were informed by this accurate, near real-time individual feedback, they were motivated to make adjustments to their practice.
But a lack of resources has historically been a challenge to implementing best practices in healthcare organizations. Some organizations are able to dedicate resources to discovering best practices. Others simply don’t have those kinds of resources. The best analytics vendors help organizations overcome that barrier. They come equipped with best-practice order sets and other analytic intellectual property for use by healthcare organizations. Physician-led workgroups that form into permanent teams can then take these standard best practices and adapt them for the particular needs of their organization.
Connecting All the Pieces of the Outcomes Improvement Puzzle
Organizations that successfully and sustainably improve outcomes understand that the analytics system is not an end in and of itself, but a means to an end. They understand that adoption and best practice systems are also essential.
These organizations’ improvement projects are not driven by IT. Executives (and especially the chief medical officer) are engaged and involved in the project. At the same time, these organizations have grassroots champions and clinician leaders on the frontlines of care who help elicit enthusiasm and drive adoption. Engaged clinicians keep patient needs in their sights at all times and don’t get distracted by the bells and whistles of new technology.
Finally, successful organizations do not have an “if you build it, they will come” mentality. They know that gaining adoption requires concerted effort and establishing new organizational structures. They’re not afraid to break from the status quo or try a new intervention to improve care processes. Instead, they’re willing to break down barriers and disrupt inertia.
Would you like to use or share these concepts? Download this predictive analyitcs presentation highlighting the key main points.