Lessons from the Mayo Clinic on Using Data to Improve Surgical Outcomes
Many health systems have a valuable asset that they are underutilizing: clinical registries. These database systems collect information about every patient encounter, including co-morbidities, surgeries, postoperative complications, and patient-reported outcomes through standardized surveys.
But they are often not fully utilized. If selected information from the registries is displayed via dashboards, it can serve as a tool to provide a comprehensive overview of a clinic’s performance in terms of data quality, cost and volume for various procedures. Such registers can be used to identify areas for improvement, secure more favorable terms in contract negotiations with payers, and improve patient-physician discussions and decision-making. This is exactly what the Mayo Clinic’s neurological surgery department did.
In 2016, department leadership recognized the need to collect and accurately measure surgical quality in a systemic fashion that followed the expanding neurosurgery practice of the Mayo Clinic, which had expanded to six campuses with 51 neurosurgeons in different parts of the United States. We both ran a multi-site pilot project to create a solution.
Our pilot was successfully completed in 2019 and new data is now being fed into our registry and dashboards on an ongoing basis, with the exception of complication data, which data management analysts manually enter on a daily basis. . Our results demonstrated that a departmental registry is an effective tool for tracking surgical outcomes in disparate practice settings. Based on our pilot project, other departments of the Mayo Clinic are launching similar efforts.
Here are the steps we learned during the pilot project that can serve as a guide for other healthcare facilities interested in creating a similar system:
1. Identify the opportunity and the barriers to its success.
Our neurosurgical practice identified the need for a performance measurement system that would provide comprehensive insight into the value of our highly specialized neurosurgical care that includes complex surgery for brain tumors, cranial aneurysms, spinal deformities. and other neurological conditions. The mission – to design an effective clinical registry tool for a complex and highly integrated healthcare organization like the Mayo Clinic – was daunting.
The geographic distance between the various campuses of the Mayo Clinic complicated matters. Mayo has one of the strongest neurosurgery practices in the country, with 20 neurosurgeons in Rochester, Minnesota; 11 in Jacksonville, Florida; 12 in Phoenix, Arizona; and eight in community hospitals in the Mayo Clinic Health System in the Midwest. In total, our neurosurgeons perform approximately 11,000 surgeries per year.
2. Get a broad membership.
An effort to create a data-driven department requires a significant investment of money and talent and must gain buy-in from departmental leadership and the institution. Therefore, the first step in developing useful data reporting practices and dashboards is to get staff at all levels of the organization to recognize that their creation is essential to ensure that the organization provides health care. high quality and cost effective and remains competitive in the market. . We got buy-in from our department staff by running training sessions at the six locations in the department where we discussed the need to get such data and have user-friendly solutions.
3. Ensure adequate financial resources.
Prior to this project, the Mayo Clinic had made significant investments in the development of a clinical registry infrastructure in the neurosurgical specialty area. But it took a significant additional investment to turn that data into a dashboard designed to inform operational decisions and contracts with payers. This required a total investment – including time, personnel and active mining – of around $ 1.5 million over a five-year period.
4. List the variables of interest.
We envisioned that this neurosurgery practice database and dashboard would allow for a clear and insightful visualization of operational metrics such as case volumes, outcomes, and relative value units (UVRs), a metric used by Medicare, Medicaid, and commercial insurers to calculate reimbursement. The database captured variables such as postoperative readmissions, complications, returns to the operating room, mortality, length of hospital stay and disposition at discharge for all patients undergoing surgery. neurosurgery at the Mayo Clinic. The registry would also report and visualize financial metrics such as total hospital costs and fees for all operative neurosurgery admissions.
The dashboard would become an important feedback mechanism for neurosurgery managers at each site, facilitating multi-site discussions on performance and quality improvement projects to improve patient care. (The current version of our dashboard allows the user to view performance metrics by date, Mayo Clinic location, procedure category, and individual surgeon. The average cost associated with admissions for different surgical categories can also be viewed , which allows cost comparisons between different sites.)
The idea was to create a sort of âbalanced scorecardâ as has been proposed by Robert Kaplan, David Norton and others. Our team embraced it and recognized that we had the opportunity to create a non-intimidating and easy-to-use dashboard that would provide insight into quality and profitability.
5. Build a project team of talented people from all critical areas.
We have created a committee with expertise in clinical, administrative, financial, supply chain and IT areas as well as quality reporting and improving practices. All six Mayo Clinic hospitals where complex neurosurgical procedures are performed were represented, as were data analysts, statisticians and data display experts. It was essential to have both data experts and data novices on the committee to ensure that the end product would be widely usable. The commitment of these people required a personal commitment on the part of the department directors and their designated doctors and administrative managers. The committee members were highly qualified specialists who did not necessarily work together or did not know each other’s needs and processes. To make sure they worked well together, we embarked on a multi-month effort to build bridges of understanding and common purpose between them.
6. Create a feedback loop to improve and enrich the dashboard.
Following the optimization of data collection and storage mechanisms, as well as the accumulation of data on nearly 20,000 cases, we began to identify a methodological paradigm for the use of these robust data in real time. . This led to the design and development of a dashboard that would turn complex data into actionable information for quality monitoring, operational decisions and contracts with payers.
The scorecard framework described these as critical requirements:
- Data can be imported from multiple sources (neurosurgery registry, costs and UVR information from a separate organizational platform) in an automated fashion
- Data summarizing key metrics by site, surgeon, and procedure type and can be displayed visually
- Data that could be exported when a critical review of the underlying information was required, such as when a specific patient number is needed to review the exact details of a surgical case or group of cases
As it was developed, the dashboard was continuously reviewed by the stakeholders involved – department and project management as well as financial analysts – to assess its operational capacities, clinical relevance and its relevance to shape market strategy.
7. Include predictive analytics.
In addition to descriptive analyzes that allow you to interpret historical data to better understand the changes that have occurred, our team has developed a predictive component tool that we have called âthe neurosurgical risk calculatorâ. Based on real-time evidence from the practice of the Mayo Clinic, the calculator allows the neurosurgeon and patient to estimate the operative risk for a particular patient profile. This risk calculator provides another way for physician and patient to discuss options, estimate risks, including potential complications after surgery, and share decision making.
8. Use the data.
Hospitals collect data that often goes unused and unreviewed. To avoid such a scenario, we created periodic reports for the directors of the neurosurgery departments of the six sites, solicited research proposals from the different faculties of the departments and launched system-wide quality improvement projects. to address areas of relative underperformance. The ministry also used the data to secure competitive contracts with employers and insurers, including bundled care agreements. By showing them their patient outcomes and how they compare to national referrals, payers have increased the number of patient referrals to the service, resulting in an increase in volume. Data from this registry is also used in shared decision making with patients and to provide patients with expected outcomes after surgery.
After significant investment and commitment, we have developed a database and dashboard that achieves its goals of providing reliable data on quality, costs, volumes and results. It has helped us transform our practice.