Healthcare organizations know that they need to an effective clinical data analytics strategy to improve and survive in today’s challenging environment. In order to make these necessary improvements, healthcare leaders need to establish clear goals for their clinical data analytics initiatives.
Achieving these goals requires clinical teams to clearly identify problems and plan for how to achieve them. This article walks improvement teams through sometimes confusing process of identifying problems, setting clear, achievable goals, and common pitfalls along the way. Topics covered include:
• Six categories of clinical data.
• Three types of goals: outcome, process, and balance.
• How to write an outcome goal.
• Internal vs. External Benchmarks.
• Mitigation strategies.
• Getting clinical buy-in.
Download
Healthcare organizations know that they need to use clinical data analytics to improve and survive in the new, demanding world of value-based care. In order to make necessary improvements to patient care, per capita costs, and patient and clinician satisfaction, healthcare leaders need to establish clear goals for their clinical data analytics initiatives. Although this sounds like a simple concept, it is often difficult for healthcare organizations to execute.
Clinical data can be broadly defined as any information that is a collection of observations about a patient or population. Clinical data falls into six major types:
Leveraging clinical data has the opportunity to assist an outcomes improvement team in transforming patient care by improving health outcomes, lowering costs, and improving patient and healthcare professional experiences.
Because of high EMR usage, healthcare organizations capture vast amounts of clinical data. Unfortunately, most of this information is stored away never seen or evaluated. In order to transform healthcare, we are going to need to move from the era of “big data” to “meaningful big data.” Meaningful big data is information that can be used to inform change. We must at the same time, begin asking what data is not informing change and then develop processes to streamline documentation. This is why an effective clinical data analytics strategy is needed.
According to an MIT Sloan Management Review study, top-performing companies in their respective industries are three times more likely to be savvy users of analytics compared to lower performing companies, and the top barrier to leveraging data is a “lack of understanding of how to use analytics to improve the business.” Cultivating success requires healthcare systems to form interdisciplinary outcomes improvement teams that are actively engaged in the transformation process. To be successful, these teams must be led by clinicians impacted by the work and who believe in making decisions based upon data.
Goal setting is one of the most important steps in implementing organizational change. For any outcomes improvement project to succeed, the impetus must stem from a problem describing the reason for change. This should be a concise statement that clearly identifies the problem and the implications that will result if gone unanswered. The problem statement should not be confused with symptoms of the problem or seek to identify the exact processes that need to be in place for the problem to be corrected.
For example, a community hospital identified that the rate of patients leaving the Emergency Department without being seen was higher than it should be (problem), suggesting issues with overcrowding and long wait times (symptoms). The problem statement should also not be confused with a solution–improving wait times in this example.
Once the problem statement is created, the outcome goal can be defined. Outcome goals should be seen as an extension of the problem statement. In looking at the work from Simon Sinek, every organization on this planet knows what they do; In healthcare, it’s taking care of people. Some organizations know how they do it, but very few know why they do it.
In creating a problem statement, an improvement team creates their esprit de corps. The outcome goal should be a specific extension of that statement. With the why and its extension clearly defined, outcomes improvement teams can next identify the how (process) followed by defining the what (intervention). Lastly for teams to be successful, it’s crucial that evaluation is a result of a broad review. Balance measures enable this enhanced view and provides team members with insight in determining if the various processes and interventions are generating the intended result and that there are no unwarranted consequences.
Outcome measures are the broad quality measures that healthcare organizations are trying to improve, such as mortality, readmission , and variable cost per case. Defined by the World Health Organization as a “change in the health of an individual, group of people, or population that is attributable to an intervention or series of interventions,” outcome measures are long-term goals that take the longest amount of time to achieve and are often influenced by multiple processes. It’s also important to note that outcome measures in healthcare are often surrogate measures that are a result of gaps in data availability.. For example, when an outcome measurement such as mortality is used, typically in-hospital mortality is the measurement reported rather than overall mortality – there are limitations in accurately measuring mortality events occurring outside of the hospital or healthcare system.
Process measures are detailed statements, often easier to measure, and can be achieved in shorter fashion that their outcome counterpart. They are linked to achieving the outcome goal and indicate how key parts of the system are improving.
For example, if the outcome measure is LOS, a process measure associated with that outcome might be reducing the time between when a provider writes the order for a patient to be discharged and when the patient leaves the facility. Other examples of process measures include the rate of patients with sepsis who have an antibiotic given within three hours of their arrival or the rate of patients with a diagnosis of heart failure who have a documented ejection fraction. These measures are the specific steps in a process that lead to a particular outcome metric and are important in order to reduce the unwarranted variation in patient care and thereby improve the quality of care being delivered. They use best practices to standardize improvement efforts.
Lastly, there are balance measures. These can be either process- or outcome- related measurements that help ensure the changes being made are moving the needle in the right direction and are resulting in no unintended consequences. Balance measure results are nearly always driven upon implemented processes. If a clinical team is trying to decrease the utilization of emergency department services within their system and they decided that implementing a care management program within their system in order to see if they could decrease patients from using unnecessary ED service, a balance measure might include looking at utilization of out-of-network services or inpatient hospitalizations. When thinking about balance measures, it may be helpful to think of them as surrogate safety measures. Make sure to spend an appropriate amount of time when thinking of these measures in order to keep the patient, clinician and healthcare systems safe.
For a measure to be transformed into goal it needs to be SMART: specific, measurable, attainable, relevant, and timely. Science tells us that people who write goals are significantly more likely to accomplish them. In a study of students in the Harvard MBA program, only three percent had written goals and included plans to accomplish them once they graduated. Ten years later, those three percent who had written down their goals were now earning 10 times more than the other 97 percent of their class combined. So remember to be SMART in writing down your goals.
See Figure 1 below.
Examples of SMART outcome goals are: reduce the mortality rate of patients with [diagnosis] from x to y by [date]; reduce the variable cost per 12-month episode of care for patients with a [diagnosis] from $x to $y by [date]; or increase patient satisfaction scores for patients with [diagnosis] from x to y by [date].
In order to write SMART goals, clinical improvement teams require baseline data (the “x”). Baseline data provides insight into the current-state and establishes a starting point to measure against once improvement efforts have begun. Baseline data can also be used for motivational purposes after the work has started to remind team members how far they have come and to realize that their hard work is indeed making a difference.
Before beginning improvement work, both healthcare leaders and clinicians frequently look to gather information in order to understand historical trends and the necessity behind the need for change. If expert clinicians are involved in the work, and have been seated at the table early, often their expertise can be utilized in confirming initial directions and identifying processes that require attention. These insights may not be summarized quantitatively however, this qualitative data is just as rich and frequently can be leveraged in preventing project delays and in facilitating widespread adoption. The benefit in having an enterprise data warehouse (EDW), is that once a measure is defined, a baseline value can be established at a later point, when the data becomes available. The data can then be used to validate the consensus of the team.
The “y” component of a good goal, be it a process or outcome, allows for identifying a desired level of performance. A good “to” goal should be ambitious, but realistic in order to promote change, deliver improvements in quality, productivity, and efficiency, and, in turn, bring innovation.
In order to set a “to” goal, an improvement team needs to understand why they are setting the target. Understanding the why, will likely inform the team on whether to use internal or external benchmark target(s) (using a tool such as Touchstone™). There are specific advantages and disadvantages to each.
Internal benchmarks compare performance within an organization, department, or service line.
Advantages to using internal benchmarks include:
Disadvantages of internal benchmarking include:
External benchmarks compare performance measures with other organizations or against data published in the literature. The advantage to using external benchmarks is that it can provide an opportunity to learn from other experts from across the industry. But, there are a number of disadvantages to keep in mind:
Ultimately, teams need to weigh the advantages and disadvantages and determine what measurement is best for them in setting their goal.
Setting clearly defined, strategically aligned goals is a critical part of a successful clinical data analytics strategy and a necessity for outcomes improvement. While it can be difficult for health systems to do effectively, clear goals are an important part of change management.
Below are some questions to ask when setting goals that will help tell a complete story and increase buy-in:
Clinicians are often concerned with the measurement of quality healthcare. Concerns arise with flawed methodologies and inaccurate data that in turn are used for managerial and cost-cutting purposes on daily routines. Concerns are compounded when publicly reported outcomes are involved. Having clinicians involved in the early stages of the improvement work is a crucial first step and involving them in creating definitions based upon accurate and valid clinical criteria is paramount. Remember that change happens at the speed of trust.
A good example of the importance is when a hospital system sets a goal around moving from a femoral to a radial first approach in cardiac catheterization. While the femoral artery is a larger target that allows for easy access, the radial artery approach reduces bleeding and has a lower risk of complications. Some cardiologists initially resisted this change because they argued that their patient population was sicker than other populations and as a result, speed was the more important factor. The improvement team had to break down the regions to show that patient populations had a similar breakdown of comorbidity. They also had to show that the data was being collected and documented in a consistent way.
Although this won’t necessarily guide the goal-setting, preparation around these questions helps clinicians understand why a change needs to occur and that they won’t be disproportionately penalized.
An effective improvement team should begin with identifying a problem statement and cultivate a story around the necessity for change leveraging both clinical experience in conjunction with analytics. With the “Why” identified, improvement groups can then set SMART outcome and process goals. Now the Improvement journey is focused, with a clear line of sight on key process and outcome indicators, the team has the navigation to drive toward their goals.
