Sleep-Friendly Labs: Optimizing Testing and Timing

I recall being a chief resident years ago and being asked to explore why the morning labs were not on time or not drawn at all. Doing due diligence, I worked with the lab “Lost Specimen Task Force” to understand why. It turned out that so many labs were ordered at 4AM “STAT” that many of them were overwhelming the system since STAT labs were ordered through an extra system. The extra system enabled speedy alerts to get the lab drawn but were cumbersome for the lab to enter. So, as more routine lab orders were STAT, the system quickly broke down and resulted in fewer and fewer on-time labs. In essence, a workaround broke the system.

Fast forward many years later, and with the advent of Choosing Wisely® recommendations to avoid routine labs in stable patients and also to reduce unnecessary disruptions, our team was excited to tackle labs in the SIESTA project. We ended up improving a lot, including sleep and interruptions due to vitals and medications, but no one was ready to give up 4AM labs. We asked why. Interns said they did not want to use the 10PM labs to inconvenience their night float, and 6AM was going to be too late for morning rounds. This is despite attendings saying they did not need labs for morning rounds and that most stable patients did not need labs, but 4AM labs remained a popular option. While we could not fix labs then, we awaited an opportunity, and one came. Working with an interdisciplinary team and with the support of our Center for Healthcare Delivery Science and Innovation, we launched a survey to understand a fundamental question: How often should labs be done on stable patients?

The answers were all over the place, but there was a modal response – every 48 hours. Interestingly, the ask to go from labs every day to no labs at all did not sit well with our physicians. There was always the eminent threat that someone would go into renal failure or have a GI bleed, so the every 48-hour option would allow for surveillance to occur. It was at that moment, we realized we could marry two different ideas together – optimizing frequency of testing for labs with the timing of labs. If the patient is stable, and the lab is every 48 hours, there was literally no reason for it to occur at 4AM. It could easily occur at 6AM.

As a result, Order SMARTT (Sleep: Making Appropriate Reductions in Testing and Timing), the sleep-friendly lab option of q48h at 6AM, was born. To make it a bit more fun to order, a ‘sleep’ shortcut was created in our EMR; by typing “sleep” into the lab order, this special q48 6AM frequency would pop up. In addition, to help monitor how many patients on a census had 4AM labs, we introduced a 4AM lab icon on the patient list.

We were successful. Fewer labs were ordered, and those ordered were more likely to be sleep-friendly. Interestingly, the residents liked the sleep order shortcut, while the hospitalists liked the patient list option. One reason for this may be because resident ordering habits are easier to change whereas hospitalist services are likely to use a patient list to follow their patients’ multiple handoffs.

This leads to an interesting question in all practice improvement: Not every intervention works for everyone, and therefore, it may be important to test what matters most and to whom. Moreover, one quality improvement project’s failure is literally the needs assessment for the next project. In our case, we really needed to get a better understanding of lab culture to design an intervention that could get implemented in our hospital. Lastly, sometimes bundling two asks together is easier than one. Instead of separating the topic of sleep-friendly labs from reducing unnecessary labs, by combining the two into one intervention, we were able to achieve greater traction through Order SMARTT, our latest in the Journal of Hospital Medicine. While we could stop there, we did not fix everything… plus a pandemic hit. So, stay tuned for our newest work on addressing hospital sleep issues during COVID.