Among today’s biggest healthcare tech trends are EHRs and business intelligence. We’re investing millions of dollars creating “data engines” that track patients throughout the continuum of care. In many ways it’s similar to the internet in its infancy — we’re building a foundational layer that will transform how we deliver care and, over the next decade, care delivery will be clearly defined and data driven.
One question that often comes up in my discussions with peer OR managers and CxOs is. “How can we leverage all this data to improve OR operations?” Having spent more than a decade leading healthcare operations and struggling to balance physician needs and revenue targets, I know how important this is — at $100 to $150 per minute in operating room revenue, the stakes are high.
As a former perioperative business manager for more than eight years, here’s my perspectives on some of the big issues we as OR managers are facing today, as well as how data will help improve OR utilization.
- We spend too much time pulling data, and it’s still not enough to make objective decisions.
This is the fundamental problem most OR managers face today. Think about how much time we spend preparing reports for perioperative leadership to make informed decisions. How many reports do we need to pull from various business intelligence tools just to prepare one slide deck? In my experience, I would spend countless hours gathering data and preparing reports only to then have the committee request more information prior to making any final decisions. One time we came to a consensus on block time reallocation, only to begin a separate process of communicating the decisions to physicians which led to more and more conversations. I knew if the data was more transparent from start to finish with easy-to-read and interpret reports, we could make decisions faster and move on to other issues.
Now, here’s the bigger problem: despite all that hard work, surgeons often don’t believe in our numbers. It’s not their fault (or ours) — what they are looking for is “why.” Simply telling them their utilization was 65 percent isn’t enough; they want to understand why (and likewise have us as leaders understand why, because often part of the problem is not highlighted just in the two-dimensional report of utilization).
“I have a full-day block on Mondays, and I generally do two cases; why is my utilization low?”
Answering that question requires us to delve deeper into the data and understand their utilization patterns. The dashboard shows us their utilization based on the wheels in and wheels out and turnaround times. However, to really answer their question, we would need to look at what types of cases they have done, how long they took, how long they generally take, and show them why they are not properly utilizing their block time.
If we are able to say, “You’re utilizing your mornings well, but in the afternoons you’re taking up cases that end too early or take too long,” that’s more productive and convincing.
These types of questions require sophisticated analysis of OR block data and case data, something BI tools are simply not equipped to handle. What we need is a system that shows us exactly how surgeons are utilizing their time, what their “green” and “red” patterns are, and offers actionable insights on how we can make them better — insights such as: “Surgeon X is not utilizing Wednesday afternoons well, but Surgeon Y is scrambling to find time on Wednesday afternoons.” This allows us to have objective conversations with surgeons and opportunistically reallocate block time without frustrating them.
- While the concept of block allocation is good for planned available time, it often causes waste and downtime.
Let’s face it: our block allocation process isn’t very efficient. In many ways, it’s similar to how Congress passes bills — surgeons lobby for more time, we spend hours pulling data and proposing a change, and it gets debated and resolved in the OR committee sessions. It can take anywhere from a month to three months (to six months) just to make one change, not to mention countless hours of time we could have spent elsewhere. Does it really need to be so slow and complex? Developing a way to make the schedule more transparent so those providers can easily see available time to prevent the long list of add-on cases can better use this available or underutilized time, while in parallel the OR committee analyzes proposed block reallocation.
When was the last time you called five cab companies to book a cab? The smartphone has changed our lives — apps like Uber, Airbnb, Kayak, and HotelTonight have made most mundane tasks as easy as the push of a button. In a world where data drives block allocation decisions, block management should be as simple as that.
Surgeons and their schedulers should just be able to say, “I need a three-hour block in the next two days,” and the system should present them options based on their preferences. They should be able to easily find and reserve a block with one click. If a surgeon is planning a vacation or if they are not utilizing some of their block time well, the system should automatically prompt them to release it so other surgeons can take it, and they should be able to release blocks with one click.
- We need forward-looking tools to plan and execute better.
Metrics and dashboards give us a good “rear view mirror” into operations, but to plan ahead we need a more forward-looking view. It helps to look at room and block utilization, turnovers, and first case late starts in the past month or quarter, but that data doesn’t really help us plan for the future.
How often do we scramble to find nurses because of unexpected demand? How often do we wish we knew in advance how many cases to expect on Thursday at 5 p.m.? If an anesthesiologist takes Friday off, how would that impact our unit operations?
How many of you would agree we need less and more efficient meetings? I want to present to leadership forecasting models that are very easy to read, interpret, and act upon. So instead of spending the entire hour reviewing an underutilized block, we spend 10 minutes and are confident in what we are then recommending.
Having answers to these types of questions can save a lot of time, money and, of course, stress. The good news: data has answers to all these questions. Machine learning and predictive analytics can dig into data and predict demand ahead of time; they can tell us how many surgeries we can expect on Friday night, how many nurses and anesthesiologists we need to support that demand, and how much demand we can support if some staff members take time off. This is the kind of forward-looking information that will help us plan better and ensure we can actually meet our revenue and performance targets while balancing surgeon needs.
The OR is one area where more innovation is needed in short order. Given the enormous stakes in revenue as well as patient, surgeon, and staff satisfaction, any improvements we make can have a real impact on the bottom line.
What do you think? I would love to hear your thoughts.
Originally published in Health IT Outcomes.