Sometimes the question you’re asked, or the one you heard, isn’t the one to answer—and other lessons learned in the data chasm.

This spring, our team’s data analyst and I struggled with projects and queries that didn’t seem to yield results. Our data efforts were spinning out of control. We were investing large amounts of time but didn’t feel we were providing as much value as we hoped for our clinician and leadership stakeholders. We had fallen into the data chasm, and were surrounded by frustration.

The data stories that people tend to share are of successes; best practices; and administrators, nurses and physicians working in harmony to answer questions or solve problems that will improve or grow the practice. How often do we hear anecdotes where the data story was misunderstood or the team decided the question was either unanswerable or not worth the effort? How often do we think analytically about the failures and frustrations, or what we’ve learned from them? Here’s what our team learned (or relearned) from our efforts this spring.

More pre-work means less frustration

Consider this scenario: Clinician calls administrator with a request to mine the practice’s data to answer a question. If the administrator responds by rote, “Yes, we can do that,” it’s likely that frustration lies ahead for both parties.

The better reply is, “I’m sure we can help. Let’s go over some questions together so we can provide you with the best outcome.” Our team has five questions that we ask before launching any data-related query or project (see figure on page 33). The conversation that flows from these questions is often revealing; we may learn that the stakeholder needs something slightly (or very) different from what was stated (or what the administrator heard) in the initial request. From the pre-work conversation, both the administrator and the stakeholder may gain insights that will change how the population is defined, how parameters are set across databases or how the final report is presented.

Trusting the plan saves time & resources (it also reduces stress)

We use a project prioritization spreadsheet for data requests and wish lists. It helps us divide up work by categories, set priorities, maintain a clear direction and distribute our goals strategically. It keeps us on track for achieving requirements while also pushing us to be proactive and develop new reports that will challenge our service line.

In our haste this spring, we shortchanged the pre-work phase and fell into a reactive posture. Because of the volume of seemingly urgent requests, we lost sight of the priorities. The result was a cycle of nonstop intake, frustration and stress. We were in firefighting mode, rather than a strategic, thoughtful planning mode. When we returned to the plan, we found greater productivity toward key goals.

[[{"fid":"23066","view_mode":"media_original","type":"media","attributes":{"height":259,"width":622,"alt":" - Stakeholder Interview","class":"media-element file-media-original"}}]]

Sometimes there’s wisdom in walking away

Sometimes our data systems aren’t up to the job, and it’s usually because the problems or requests are too complex for the system. In these situations, it may be necessary to choose the most critical components of the project and sacrifice some of the detail. Remember, data that can’t be validated also can’t be shared and, therefore, won't result in impact.

Or a question may come along that requires an extensive audit, and the team may decide that “the juice is not worth the squeeze.” In our pursuit of the perfect data this spring, we were investing more time answering a request in its entirety than focusing on findings that could impact our business.

Few things are tougher for data-focused teams than walking away from an unanswered question or an incomplete project, but it happens. And when the team agrees to move on, the failure can make them stronger.

The story may not come together as expected

Often the data analysis team will hypothesize the conclusion we expect from a pivot summary or access query. When we’re right and the data story comes together as predicted, it feels like a celebration is in order. But sometimes being wrong, or partially wrong, is the right answer. For example, this spring, when we presented results that we thought were of little value, our clinical stakeholders saw a pattern directly related to their clinical practice. With their insight, we were able to identify opportunities for managing costs.

Other times the data hypothesis and story align, but it will take longer to achieve the impact. For example, our team reviews profitability by diagnosis-related group (DRG). Last year, we came to fast conclusions, with our clinical partners, about an intervention that would improve revenue. The intervention included EMR modifications and education for providers and coders. To our frustration, this spring, when we reviewed the data again, there had been little financial gain. We repeated the process, zeroing in on a simpler intervention that now looks like it will be successful.

This spring’s frustrations made our team more resilient. They reminded us to trust the process—from pre-work, through planning and prioritization, all the way to confirming the story with stakeholders. Turns out you can learn a lot when you’re in the data chasm.

Megan S. Berlinger, MHA, is the business administrator for the Heart and Vascular Center at Wake Forest Baptist Medical Center in Winston-Salem, N.C.