Is Your EHR Telling the Whole Story?
"The fastest and most efficient way to add beds to a hospital isn’t to build a bigger hospital… it’s to reduce variation in length of stay.”
EHRs are commonplace in most healthcare organizations today. Data is captured everywhere. So you should be able explore every procedure performed, by every care provider, across demographics and specialties, with measurements of cost, quality, and outcomes. For any point in time or set of conditions. It should help you to respond with the best course of action and care plan for your patients.
So what’s the problem?
If you aren’t doing this today or lack the capability, it likely has to do with three main causes:
- You’re looking to your EHR vendor for these answers.
- Your questions require data from multiple EHRs or other systems.
- You lack the right tools to enable your clinicians and leaders to consume the analytics that provide these answers, right at the point of care or in their workflow.
According to Dartmouth Atlas, in the USA, unwarranted clinical variation accounts for at least 30% of wasted health care spend. Furthermore, about a sixth of your physicians may be responsible for this waste.
A leading healthcare organization in Texas once told me “the fastest and most efficient way to add beds to a hospital isn’t to build a bigger hospital…it’s to reduce variation in length of stay.”
Let’s take for instance, orthopedics. In this example, we want to determine: Which operations have the most volume (the dots) and which ones have the most variation in length of stay (the bars)? The goal is simple: Find the top procedures with the most variation, find the most opportunities to improve care, patient experience, and costs.
If we focus on knee replacements, which is our most common procedure (has the highest dot), we see the variation that exists among the surgeons. Let’s drill in.
(Did you know the American Academy of Orthopedic Surgeons (AAOS) uses analytics to present benchmarks for joint procedures? See the AJRR twitter feed for the latest.)
Notice the length of stay variation that exists among the surgeons. It’s clear one clinician is a significant outlier. Let’s remove this clinician from our analysis so that we can see the variation and opportunities across the pack.
The analytics now reveals the average length of stay for knee replacement is 5.37 days across all clinicians (excluding our one major outlier, whom we bookmarked for later investigation).
To understand the potential saving for each clinician if the length of stay came down to the average figure, you’ll need an engine capable of calculating the savings over a large data set quickly. We’re not talking ask IT to craft a report for us and wait weeks. You need answers immediately.
Try it out! Let’s find potential bed day saving with knee replacements.
But don’t just take my word for it. Learn how the Children’s Hospital of Pittsburg (of UPMC) tackles clinical variation within Cerner. They’re presenting in the healthcare track of the Qlik Virtual Forum on March 30th.
