How MultiCare uses LEAN for healthcare data requests

Many healthcare organizations feel overwhelmed by massive data requests coming from throughout the organization. Data requests can take over time and resources, draining staff of the ability to engage in strategic and innovative initiative.

Healthcare data is typically locked away in electronic health records systems; thus, the ability to retrieve data in a meaningful way can be a daunting task. This inaccessibility can create difficult situations for analyzing, reporting and predicting, which could, otherwise, lead to improvements in patient care and reduce the cost of healthcare.

So why is it so difficult to meet data needs in healthcare?

Healthcare data is complex. One of the original architects of data warehousing, Ralph Kimball, states, “Some of the most complex models that we have ever worked with are from the healthcare industry.” The learning curve of the healthcare data model is extremely high. Years of learning and experience are necessary to provide effective service to the organization, because there are hundreds of tables with endless columns from which to choose. Therefore, it is critical to effectively utilize the precious knowledge resources on the highest value tasks.

Requirements are unclear. Many healthcare workers have no idea how to go about achieving their reporting or analytics tasks, nor do they have the knowledge of alternative avenues to finding reports for their consumption. Depending on the expertise of the requester in knowing what they want, our team could spend significant time defining requirements after receiving a data request.

Rapid changes in healthcare. The lifespan of reports or dashboards is about three years on average. This is caused by an influx of ever-changing issues: Leaders come and go, healthcare regulatory requirements change, the EHR system build changes, workflow changes, and the list continues to unravel. These often put IT into the situation where things start breaking in the background, and data requests start to trample other IT duties.

Lack of standards. The healthcare industry has not been good at standardizing processes, particularly around data governance and definitions. There are so many different versions of tracking providers, as well as different definitions of readmission and lengths of stay, just to name a few. IT continues to spend time rehashing, recreating logic, and needlessly developing strategies, each in a slightly different way.

What is the common thread to the question of why is it so difficult to meet data needs in healthcare? It all ties into “waste.”

Waste causes physical and emotional fatigue, increased frustration and stress, causes us to blame others, and it causes a significant toll on time and resources. This wasteful process and behavior are not LEAN. But there is hope. Here’s how our organization tackled our problems by managing “waste.”

Our team harnessed one concept of the LEAN methodology called “The eight wastes reduction.” It is commonly used for LEAN manufacturing, but we can easily apply the same concept here. There are eight different types of waste in LEAN, which include motion, waiting, transportation, defects, extra processing, overproduction, inventory and non-utilized talent.

The 7-step LEAN approach
1. Hire a triage specialist to focus on managing data request tickets.
Reducing waste: Non-utilized talent

Data architects and BI developers are generally not a good fit for triage work because they are not specialists in triaging, but excel at solving data and technical problems. Rather, a solution would be to find someone with good communication skills who is excited to do triaging work. We had great success by hiring a triage intern who is good at working with customers and excited to learn healthcare reporting and analytics. The purpose of this role is to triage and organize tickets that come into our queue from the help desk. The triage specialist contacts requesters within several hours, either requesting additional data requirements or supporting prioritization tasks for larger requests.

Here’s what to look for in the triage role:

• Excellent oral and written communication skills.
• Good customer service and documentation.
• Ability to learn reporting tools and teach others how to use them.
• Interest in process improvement.
• Shows curiosity and enjoys investigating existing solutions.
• Enjoys problem solving.

2. Hire technical interns to work on less complex request tickets.
Reducing waste: Extra processing and non-utilized talent

Have you had an experience where you keep seeing the same or similar request over and over? This indicates that you keep putting your time and energy into reviewing and working tickets that are similar. Extra processing is the fifth waste of LEAN.

Hiring several technical interns is a perfect win-win proposition. Our team normally has two interns at a time, and their assignments are to resolve various data and report requests to help contribute resolving simpler tickets. They are excited to learn from the real work environment by solving real problems, and they get to learn new skills. Repetition is helpful for interns to enhance and solidify the learning. Interns start seeing patterns in similar requests and can come up with a fresh solution that results in reduced repetitive requests or consolidation of similar reports. We assign simple tasks initially, and over time they can resolve more complicated requests.

What to look for in a technical intern:

• Understanding of SQL or has taken SQL classes.
• College student, focusing on information technology or computer science.
• Communicates well with customers.
• Enthusiastic about learning data, reports and analytics.
• Good at customer service.
• Excellent problem-solving skills.

3. Assign a rotating on-call person to assess requests and work on those determined to be “quick wins.”
Reducing Waste: Motion

Our team has data architects on call on a rotation. The on-call assignee manages incidents and maintenance requests, monitors processes, answers technical questions, performs production migration, monitors and works on small “quick-win” request tickets. The on-call designee is normally the go-to person for the triage and technical interns when questions arise. Because the on-call person has a technical background and experience, it is more appropriate for on-call to make an assessment where the tickets should be assigned. The clear expectation of on-call duty reduces the motion waste of tickets bouncing around between different staff members. If the request is a “quick-win” request, the on-call person may resolve the requests while they are assigned to that duty.

4. Develop a Service Level Agreement.
Reducing Waste: Waiting

Have you ever received requests without clear requirements from nonresponsive requesters? Waiting is a big and common waste in LEAN.

The development and its implementation of SLAs is a critical part of actively managing request tickets. Quite often, we receive data or report requests without any requirements. To work those requests, we need additional information and requesters work with our team to provide information in a timely manner. We close tickets if requesters don’t reply after several attempts, thus preventing empty tickets from dragging us down for any longer than two weeks. The SLAs are also helpful with customer expectations and experience, because we can get to their requests quicker.

5. Define the requirements upfront.
Reducing Waste: Overproduction, inventory and defect

Have you run into a situation where the requester continues to add a request—one after another? Overproduction, inventory and defects are types of waste in LEAN.

Have you also encountered a situation where the requester’s deliverable expectation did not match because it was never documented? Just-in-time delivery is the best-case scenario. If the request is delivered too late, the information may become irrelevant. If the request is delivered too soon, rework becomes an ominous possibility.

We frequently deal with unhappy requesters who say reports are wrong because of obscurely defined requirements. Regardless of right or wrong, if the displayed data do not meet their requirement, it is considered incorrect, and rework is required. Requirements need to be documented and signed off by a requester.

Our team created a data intake form to capture basic requirements for smaller requests. We also created a request intake form for larger requests to capture business values, based on a priority scoring tool. Our team meets weekly to score and come up with a high-level estimated effort before being presented at a prioritization committee.

6. Physically locate triage specialists near data architects and BI developers.
Reducing Waste: Transportation

Locating the triage specialist in a cubicle close to those who have answers promotes the efficiency of dealing with request/tickets, thus reducing transportation waste.

7. Designate a single person to monitor the progress of visual management.

In LEAN, there is a concept called visual management, which is the strategic placement of performance key indicators in plain view to improve communications. The visual aid ensures everyone involved can understand the status of the system at a glance. In our case, a chart was sent via email every week to those who were involved. The regular communication ensured that everyone was aware of the progress as to how we are doing. Visual management is a great tool to motivate everyone to let them know that we are making strides and making progress.

Goals are in place to improve the process and burn-down backlog. See the Request for Enhancement (RFE) Burndown Chart below as an example. This burndown chart is the actual tracking of our team’s progress in 2017. The key data points, “Begin counts,” “End Counts,” “Closed ticket counts,” and “Unassigned counts,” were tracked on a weekly basis and this chart was reported every Friday to the team and leadership.



Conclusion
The different types of waste took shape as an antagonistic dragon. Yet no matter how ominous the dragon became, we realized that data requests and requesters were not our enemies after all. Each problem identified presents enough challenges to spend years improving. However, we can make pragmatic and incremental improvements by tackling waste through LEAN and ultimately taming the “waste” dragon, thus making data request become a small part of our work.

Reduction of waste alone is not a panacea to our problems. However, this LEAN methodology, indeed, had its place in health care, and allowed us to be more productive, happier and boosted our morale in every aspect.