How to Measure Health Outcomes that Matter to Everyone
Today’s healthcare industry is consumed by an urgency to measure and improve outcomes. The switch to outcomes-based healthcare is incredibly positive, but it also has many organizations scrambling to put together data-driven improvement programs. This rush, though necessary, often results in reactionary and haphazard plans for outcomes improvement. Using a few key ideas, organizations can improve how to measure health outcomes and take steps toward real improvements in quality and cost.
The Distinction Between Regulatory Requirements and Outcomes Improvement
As the industry accelerates its focus on improving outcomes, a virtual flood of new regulatory requirements continue to arise. Recently, CMS instituted a new bundled payment program for hip and knee replacement that will hold hospitals accountable for the quality of care they deliver to Medicare fee-for-service beneficiaries from the time of surgery through the initial 90 days of recovery following discharge. While initially restricted to a limited number of geographic areas, it has the potential to become the standard method by which CMS will hold hospitals accountable for delivering high-quality care in the near future, including care areas beyond the hip and knee patient population.
At times, regulatory requirements can create a national focus on common goals and standards. However, achieving the benchmarks established from regulatory requirements should not be seen as the final organizational destination on the journey to continued outcomes improvement. Rather, regulatory requirements should be seen as just a single leg of a multi-leg journey, where detours created by the individuals involved and obstacles experienced along the way require change and adaptation.
For example, there are national standards for stroke care improvement and for providing certain types of care within a specific time window. With stroke, time is crucial, and depending upon the type of stroke and other factors, tPA, an intravenous medication, may be used to dissolve a blood clot that is blocking a vessel in the brain. The national guideline is to begin administering tPA within 60 minutes of a patient arriving at the emergency department. If an organization achieves this measure, it doesn’t mean the work is over, or that the outcome has been achieved. Instead, the team should begin looking to the next leg of the journey and seeking how to shorten administration time even further so the majority of patients have a door-to-needle time of, say, less than 45 minutes.
In healthcare, we often use surrogate outcome markers and process measures (e.g., did a provider follow an evidence-based protocol? Did a patient receive the right educational materials at the time of discharge?) for improvement because some outcome metrics are often difficult to collect. When measuring outcomes, we often focus on length of stay or mortality rates; these metrics are often readily available and easily calculated. But do these represent true outcomes? No, our best outcome metrics speak to the health and functional status of the individual: How quickly was a patient able to return to work after surgery? After a prolonged hospitalization, how long did it take before a mother was able to carry her baby up the stairs without discomfort? The difficulty arises in collecting this type of data, as it often requires allocating additional time and resources. Ultimately, when clinicians and providers are able to capture this type of information, the results often are highly valuable and insightful into transforming healthcare. In 2003, Dr. Margaret Herridge and her colleagues at the Toronto General Research Institute, published a landmark article in the New England Journal of Medicine that looked at one-year outcomes in survivors of acute respiratory distress syndrome (ARDS). The study went beyond identifying hospital mortality as a primary outcome and recognized that patients often have long-term sequelae that were a result of an initial hospitalization. This insight resulted in many organizations reevaluating their care for hospitalized patients with ARDS by focusing on improving the associated long-term outcomes with interventions that included daily sedation vacations and mobilizing patients while on the ventilator.
We need to look at both outcome and process metrics in order to establish quality-improvement programs. The need to react to new regulatory requirements and to focus on metrics with an associated incentive or penalty won’t go away anytime soon. Hopefully though, as healthcare organizations become more analytically capable, we’ll transition from reactionary outcomes measurements to a thoughtful, proactive process for selecting outcomes measures and begin to identify regulatory measures as being just a leg in the journey to meaningful outcomes improvement.
Avoid the Outcomes Measures Graveyard
An outcomes measures graveyard is full of metrics designed with good intentions, but are of limited utility when it comes to significantly improving quality and cost. It results from imprecise execution. With so many human, financial, and operational resources required to improve outcomes, we need to ensure all efforts are strategically aligned with an organization’s goals and values while also considering how much of the data that is being collected is useful and necessary.
It’s understandable and easy to get excited about the possibilities inherent in data. However, too often we fail to consider the strain that data capture can put on already busy healthcare teams. For the most part, data collection for outcomes measures requires providers and clinicians to document an evidence-based process or action. This additional time spent documenting often results in time not being spent at the patient’s bedside providing direct care.
Unfortunately, so much of that documentation ends up going to waste. Clinicians and staff spend time documenting and recording, and we not ending up doing anything with it. Precise outcomes measurement requires planning. It begins with asking the right questions, such as:
- By requiring this to be documented, are we truly improving patient care?
- How many items of information are we requiring providers or clinicians to document?
- How often will we really use the information?
- How many of these pieces of information actually matter to the outcome?
If the documentation doesn’t serve a particular purpose, remove it. The consequence of too much documentation without enough thoughtful planning is frustrated and distressed healthcare team members… and an outcomes measures graveyard.
Don’t Let Perfection Be the Enemy of Progress
When it comes to measuring outcomes, we shouldn’t let ourselves be paralyzed into inaction by a desire for perfection in our outcomes definitions and measurements. It’s all about balancing perfection with progress. Yes, we must put thought into which measures deserve our focus. However, it is also vitally important that we don’t let perfectionist planning prevent us from launching our improvement efforts. George F. Will, summarized this idea well when he said, “The pursuit of perfection often impedes improvement.”
Many health systems get bogged down as they attempt to start their data-driven improvement programs. Planning and governance are important. Data quality in and of itself can become a project that easily drains resources and budget in the pursuit of perfection, but at some point we have to say, “We know we have a quality issue that we need to improve. We’ll work on defining and refining the metrics as we go.” The key is to put processes in place for iterative improvement in order to make adjustments and refinements as the initiative progresses.
Consider the Perspective of Different Stakeholders
When creating a quality improvement initiative, it’s important to consider how different metrics are relevant to different stakeholders. Someone with a finance background may see data through a different lens than someone with a regulatory or purely clinical background. In fact, the richest and most useful outcomes measure results from bringing different stakeholders together to understand the purpose of the project or rather, the “why.”
One such example comes from a blood utilization project I was involved in where we evaluated waste associated with the ordering of blood products. While evaluating the workflow, we discovered that clinicians routinely ordered a complete blood count (CBC) following a transfusion, when a hemoglobin would have been sufficient. It was only after being afforded the opportunity to work with an interdisciplinary team that we were able to gain the perspective and insight. We were then able to relay this information to other team members and update order sets, thereby reducing costs.
This example illustrates why we should leverage interdisciplinary teams to share knowledge with each other as we try to understand the “how” of measuring outcomes. Because each stakeholder comes to the table with unique knowledge and perspective, the likelihood of the group arriving at the best measure increases. A fortunate byproduct of this interdisciplinary approach is enthusiasm and engagement in the process. Rather than imposing new measures and processes on clinicians and staff, this approach engages members of the healthcare team to shape the improvement initiative at a grassroots level.
At the very least, organizations should consider soliciting input from providers and clinicians when designing improvement projects. Not only will they provide vital feedback regarding how well a measure is designed, they often have some of the best ideas on how to improve overall quality. And because clinicians are typically the stakeholders most affected by an improvement initiative, their engagement and alignment with the organization’s improvement strategy is the ultimate key to success.
Metrics Are More than Numbers—They Represent Individual Lives
Sometimes, as we deal with measuring outcomes to meet regulatory requirements or to keep the organization in the black, we get caught up in jargon and politics. Ultimately, we must remember that these metrics represent actual patients and real health outcomes; we can’t afford to lose sight of that as we move forward.
We must remember the why. The patient needs to be at the center of our actions. Over the last 10 years, I think of some of the best groups I have worked with and how those groups recognized the value of a patient’s story and perspective. I’ve seen former patients coming to see current patients and sharing personal experiences. I have been part of classroom discussions and conferences where patients have shared unique experiences. And, more recently, I have begun to see organizations asking patients to sit on improvement teams.
We need to focus on outcomes metrics that speak to the health and functional status of individual patients. We need to understand that this is a progressive, iterative process. We need to involve the stakeholders who have a vested interest in outcomes measurement and improvement, who are catalysts for continuous, long-term change. Finally, we need to remember that patients are why we need to continuously improve how we measure health outcomes.
Would you like to use or share these concepts? Download this predictive analyitcs presentation highlighting the key main points.