Using Clinical Metrics the Right Way: 5 Considerations Every Hospital Should Know
As healthcare reform continues, the industry’s headlong shift from fee-for-service to a pay-for-performance model means quality improvement projects have become more top-of-mind for hospitals and healthcare systems, and this means that clinical metrics are more important than ever.
Between government regulations that carry heavy penalties for poor or unacceptable outcomes and financial incentives for improving population health, hospitals and healthcare systems are finding increased urgency in evaluating how care is being delivered and determining the best practices to follow in the future. As always, follow the money.
What makes all of this possible, of course, is the wealth of clinical data now available. While electronic health records (EHRs) aren’t quite perfect, they have given us a way to take data that was once isolated and proprietary to individual physicians or practices and use it to see a larger picture of management. When clinical data is combined with financial, operational, and other data, organizations have a powerful foundation to use in driving quality.
Yet having data by itself isn’t enough. That’s like having a bicycle in the garage gathering cobwebs. It isn’t until you decide on a destination, map out a route, and then hop on the bicycle and start pedaling that you realize the benefits of owning that bike.
In healthcare, the destination is higher quality care to deliver better outcomes for patients. The route to get there is having the right clinical metrics to guide quality improvements.
Cohorts Versus Clinical Metrics
Before we get into the deeper issues, it’s worth taking a few moments to outline the difference between cohorts and clinical metrics. The lines can get fuzzy at times, which can cause confusion that delays quality improvement projects.
A cohort is a group of subjects who have something happen to them, such as a disease state or type of injury. For example, you can have cohorts of patients with heart failure, asthma, or diabetes.
You can narrow down the cohorts to focus on a specific area, such as only including heart failure patients who had problems on the right side. Or you can develop a cohort based on a set of comorbidities such as diabetics who also have high blood pressure. Of course, the same patient can belong to multiple cohorts.
Clinical metrics, on the other hand, are pieces of information upon which you can base measurements. For the heart failure cohort, you can measure whether they are taking a certain medication, or if they are following an exercise program, or the last time they visited their primary care physician. The metrics are facts about the patients in the cohorts that you can change, and to which you can assign a value. Tracking those values will tell you whether the quality improvements you have instituted are delivering the desired outcomes for those cohorts.
Launching a Quality Improvement Project
A quality improvement program can be initiated for many different reasons; two commons drivers are not meeting regulatory standards or concerns about escalating costs.
Whatever the impetus, at Health Catalyst we like to begin with a key process analysis (KPA) assessment to look for the reasons why things are happening. The KPA goes into the enterprise data warehouse (EDW) to aggregate data from the clinical, financial, operational, and other data marts as-needed to identify the biggest opportunities for quality improvement by minimizing variation and reducing costs.
Having an EDW, particularly a Late-Binding™ EDW that is geared to the way data is consumed in healthcare, greatly simplifies the data-gathering process because it is all available in one place—and free to be used the way you need it. It also allows the quality improvement team to visualize real-time data to determine whether the initiative is achieving the desired results rather than having to request a report from IT.
When the greatest areas of need have been identified, you can begin narrowing the options by looking at the readiness for change within specific areas. For example, your KPA may indicate that reducing unnecessary cesarean sections offers the highest opportunity for cost reduction, but if your obstetrical clinicians are highly resistant to change your quality improvement project is destined to fail.
In the early stages of your overall quality improvement program, you want to look for an area that will give you a quick win. Once the value has been demonstrated in the context of your own hospital or healthcare system it will be easier to recruit others to the cause. In fact, we have seen instances where there was more demand than capacity to fulfill it.
With the area for improvement identified, and management is bought-in, the next step is to build the team. You’ll need an executive sponsor as well as a workgroup led by subject matter experts, including a physician and a nurse leader in the area you want to improve. You’ll also need a data architect and perhaps other support personnel such as someone from the lab.
Defining the Clinical Metrics
Now that the area for improvement is defined and the team is in place, it’s time to build the cohort and start defining the clinical metrics. If you have an EDW, you should be able to build the cohort quickly by expanding on the parameters used for the KPA, refining it as-needed and running a query. Without an EDW, or an analytics application that can delve into your EHR, you are looking at a long, tedious manual process.
Then it’s time to determine what to measure. There are several ways to make that determination. Many clinical metrics are driven by government regulations. These are things you have to do, so they make a good starting point when you’re defining your clinical metrics. You’ll want to review evidence-based literature as well to see what the best practices for that area are and whether you’re following them.
You should also gather all your stakeholders together and ask them. Based on their experience, they will have many ideas for solving quality issues, and their buy-in is important.
The downside of all this investigation is once the ideas start flowing, you often end up with too many metrics. At this point you have two options.
One is to identify a goal for quality improvement and put all the ideas into a “bank” to draw on when you decide to move forward. If your goal is to increase the number of new mothers who breastfeed you may consider dozens of potential measurements such as the mother’s age, if it is her first child, who her obstetrician is, whether she breastfed before, her education level, etc. At some point, however, you will need to decide which of these elements to focus on, or the program will become too unwieldy.
The second option—and the one we highly recommend—is to start with a smaller group of clinical metrics based around an aim statement. Aim statements are powerful because they focus on a specific improvement in a specific population, and provide a timeframe to accomplish it. They give quality improvement teams a sense of direction and help them identify the steps that need to be taken to accomplish those goals.
With the aim statement defined, it is easy to determine which metrics support it. The workgroup can then use this information to determine the process changes and evidence-based interventions that will be required to remove barriers and achieve the desired quality outcomes. It can also help identify what information to examine to determine whether the new processes are being followed and make corrections to ensure the quality improvement project’s standards are being followed.
Ensuring Consist Clinical Metrics
One of the common challenges in a quality improvement project is ensuring you can get to the information you need. In the case of clinical information, while it is stored in the EHR it may not be entered there consistently. In other words, while there may be a field with a dropdown to include the information that will contribute to your metrics, that doesn’t guarantee that the information is there. It may have been entered in the free text notes or five other places instead. If that is the case, your clinical program team will need to determine where the best place to capture that information is and how to gain compliance among the clinicians you are working with to ensure they enter data in the same place consistently.
That is what happened when I was working with a team on an initiative to encourage more new mothers to exclusively breastfeed their infants while hospitalized after delivery. The team would have to go to 10 different places in their nursing chart to document the metrics we wanted. The program team received permission to create a breastfeeding “quick chart” that listed every metric they needed in one section. The quick chart not only got the metrics into the proper place in the EHR so we could pull it out of the EDW more easily – but the nurses loved it because it simplified the documentation process for them while giving them information they could use.
Practical Example of Clinical Metrics at Work
Here’s a real-life example of clinical metrics being used to improve quality and outcomes around heart failure patients in a hospital. The goal in this case was to lower 30-day readmissions after patients had been discharged back to their permanent residence.
The literature indicates that it may take numerous transitions of care interventions to prevent 30-day readmissions. One of these interventions is to make sure heart failure patients follow up with their primary care physicians within 48 to 72 hours after discharge from the hospital. Yet, when this hospital looked at their current state clinical metrics, they discovered that only 20 percent of heart failure patients had appointments scheduled prior to discharge.
The team identified actions to increase use of this important intervention. Clinicians educated on how important the follow-up appointment was to improving overall outcomes and were encouraged to explain this rationale to their patients. The hospital made checking whether the appointment had been made a part of the transition process when the patient was being released. In some cases, case managers or nurse practitioners even contacted the primary care physician’s office to schedule an appointment on behalf of the patient. Finally, clinicians were given a specific place to document all the information so the results could be easily measured – including whether clinicians were following the new protocols.
The net result was an increase in the number of conversations about follow-up appointments, an increase in the number of appointments made (up to 90 percent compliance), and an increased potential to decrease preventable 30-day readmissions. The program was deemed a success on all counts, and led the way for further care transition quality initiatives.
Clinical Quality Improvement: Keep It Simple – and Doable
Clinical—and other metrics for that matter—offer a wealth of information and guidance for hospitals and healthcare systems looking to make the transition to pay-for-performance. Yet, they can also become a road to nowhere if you try to do too much too soon.
Keep both individual initiatives and the overall program tightly focused around specific goals or aim statements and your chances of improving clinical quality, lowering costs, and increasing patient satisfaction will grow exponentially. And you’ll be delivering the outcomes that result in quality care.