Published in the October 2013 issue of Today’s Hospitalist
AT CHRISTIANA HOSPITAL, which is part of the Christiana Care Health System in Newark, Del., every patient bed is telemetry-enabled. But because there are only a limited number of cardiac monitors, some patients who need telemetry some days have to wait longer in the ED for a monitor, which backs up the ED.
On such days, hospitalist Surekha Bhamidipati, MD, who has worked at Christiana for four years, used to find herself wondering: Do all these hospitalized patients really need cardiac monitoring? The answer, just as she suspected, is that they don’t. Patients often get started on telemetry in the ED and then never come off it, even as their needs change over the course of their hospital stay.
Part of the problem, Dr. Bhamidipati, discovered, is that despite their beeps and clicks, monitors are ironically often invisible to busy clinicians. Just like supplemental oxygen and Foley catheters, she says, once the monitors are ordered, it’s easy to forget about them or to assume that the equipment is still needed.
A few years ago, she decided to try to make the invisible visible. If physicians were prompted, she thought, they would likely question why monitoring was ongoing. Her solution was a “visual cue”: a red heart sticker on a patient’s chart to remind physicians to review their telemetry orders every day.
That very low-tech fix ended up reducing the average number of days that patients were monitored on one medical unit by 28%, dropping from 2.6 days to less than 2. Dr. Bhamidipati’s abstract describing the intervention was accepted at the 2012 Society of Hospital Medicine conference and published in an April 2012 supplement to the Journal of Hospital Medicine.
In an interview with Today’s Hospitalist, Dr. Bhamidipati notes that daily tele-order reviews continue in several Christiana Hospital medical units, although not in the form of chart stickers. Instead, the review is now an item in the script for bedside interdisciplinary rounds.
Besides the fact that unnecessary telemetry is costly and contributes to bedflow problems, what problems did you want to address?
Cost and bedflow are significant system-level problems, but as a unit director, I wanted to focus on unit-level problems first. And the amount of distraction that nurses have with every patient on a telemonitor is alarming.
Any time a patient has to go to the bathroom or for a test, the monitor has to be removed. Every time the monitor is removed or reattached, the nurse has to call a centrally located unit that continuously tracks all the hospital telemonitors.
Then there’s this problem: Every abnormal rhythm is reported to nursing staff, who are required to report it to physicians. If you are a covering physician, especially on nights and weekends, it is difficult to put a reported abnormality into context without knowing the patient. That can lead to unnecessary testing or consultations and ensuing complications.
If we don’t want to act on the results of a test, we shouldn’t do that test, and the same thing is true with telemonitoring. If we do not intend to evaluate the rhythm, we should not be monitoring cardiac rhythms.
Your project focused less on the initial order and more on telemetry duration. Why did you take that tack?
We analyzed the process flow for medicine admissions with respect to telemonitoring. We learned that there are institutional guidelines on which patients need telemetry and for how long, but those guidelines are not easily accessible and are based mostly on expert opinion. Plus, the guidelines are not always followed.
An equally significant problem is that patients initially monitored during their ED stay are sent to the floors with a telemonitor. The floor nurse sees that telemonitor and assumes the admitting physician wanted it, but just forgot to write the order. So nurses leave the monitor on, which leads to unintentional monitoring.
Another problem was that telemonitor orders defaulted to 72-hour monitoring. So patients with chest pain who should have been taken off a monitor after acute coronary syndrome was ruled out were monitored for three days if they stayed for other reasons.
Also, we did not want to question a physician’s judgment and ask, “Does this patient really need a telemonitor?” We just wanted to get the conversation started and decrease the average number of days that patients were on telemetry once the monitor was ordered. And we wanted to get rid of those monitors that an admitting physician didn’t actually order.
Why stickers on charts?
We wanted to remind physicians and nurses that this patient has a telemonitor, and prompt nurses to ask physicians, “Do you think this patient needs a telemonitor today?” Nurses would ask the question if they met the physician around the time of rounding. We thought that the back-end problem of not paying attention on a daily basis was maybe even more of a problem than unnecessary ordering.
And you found that to be true?
We saw much more improvement than we expected. Many physicians responded back to the nurse by asking, “Does this patient have a monitor? I didn’t know that.” And then they’d say, “Why does this patient have a monitor? Get rid of it.” Or in some cases, the answer was, “I think the patient needed it yesterday, but let me look at it now.” Physicians appreciated the fact that they were paying attention to unnecessary telemetry because it causes unnecessary work.
I understand you’ve stopped using the stickers. Why?
We started doing bedside interdisciplinary rounds every day, and we decided to include the question on whether the patient has a telemonitor in the safety timeout. Around the same time, another physician in my group did a project focusing on the front end, asking residents to follow guidelines for ordering telemetry at the time of admission.
Now, there are guidelines built into our electronic ordering system. So if I want a telemonitor, I have to choose among a list of diagnoses that meet telemonitoring criteria, and the diagnosis chosen dictates the duration of telemetry. But it’s still important to address it on the back end. We still have to bring it up on subsequent hospitalization days because the patient’s working diagnosis can change.
The same type of sticker project could be done for 10 other things that we do on the medicine units all the time but probably don’t need to. Does the patient need supplemental oxygen? Daily CBC if his or her blood count was normal on admission? IV fluids?
Choosing wisely in terms of tests and interventions can’t just happen on admission. It has to happen every day. What’s a good choice on admission may not be a good choice 24 hours later.
Deborah Gesensway is a freelance writer who covers U.S. health care from Toronto.