The use of reliable and well-integrated data from remote patient monitoring may prove to be one of the few bright spots that have come out of what we’ve learned in the recent COVID-19 pandemic.
Sending sicker and sicker patients home from the outpatient world and the emergency department, patients with multilobar pneumonia and hypoxemia we never would’ve not admitted to the hospital in earlier days, has helped us build systems that would hopefully inform us about who’s doing well at home, who is not doing well, and who needs to be brought back in. There simply has not been capacity to admit everyone with pneumonia and borderline oxygen levels, and we had to make hard decisions.
The EHR’s Role
When we send patients home from the emergency department or after an office visit with clearly suboptimal oxygen saturation, remote patient monitoring with pulse oximetry has allowed us to keep a closer eye on them compared with traditional follow-up phone calls. What we need to do now, moving forward, is build better systems for monitoring, better alerts, more timely and efficient integration into the electronic health record, and the development of tools that will look deeply at this data and help alert us that things are potentially about to go bad
Right now, we’re using the simplest form of monitoring — we have members of the healthcare team calling patients on a daily basis and asking what their temperature is, how they’re feeling, whether their symptoms are getting better or worse, and what their pulse oximeter reading is, both at rest and with exertion. Much better would be direct integration of this data — in real time — into the electronic health record (EHR), and then having that data be filtered by a system that’s keeping an eye on things, noticing trends, picking up patterns, noticing that levels, while potentially still normal, are heading in the wrong direction. This, I think, is the hope for the future, the true potential benefit, of using an integrated system-wide EHR that captures data and puts it into usable formats.
Example One: Renal Function
One of my favorite examples of how this might be done is monitoring of renal function in the outpatient setting. In every outpatient encounter in which we get lab testing that includes renal function, we take a look at the result when it comes back in our in-basket. Usually, we see what the number is, and instantly in our head we sort of get a snapshot of whether it’s good or bad, and if it’s what we expected. This often prompts us to take a look back at the last value, and maybe even the value before that. Is there a trend; does this value surprise me; are things stable; are going in the right direction, or in a very wrong direction?
But quite often we fail to use the functionality that allows us to graph out the data of all of the prior values for creatinine listed out over time to notice any trends. (Strangely — and this really needs to be fixed — the EHR does not adjust the X-axis for actual time between data points, but treats the time between each point as exactly the same.) Over and over, we’ve seen examples of patients whose values have remained in the normal range, but have relentlessly marched onward and upward, 0.7, 0.8, 1.0, 1.1, 1.2, 1.4. And then suddenly, we cross a magical mark (1.5!!!), the threshold where the system notes it as abnormal, turns the lab value bright red, and then suddenly we decide to do something about it.
Wait, has anyone else noticed that his patient has had a decline in their renal function over the past weeks/months/years? Wouldn’t it be great if the system was smart enough to notice these trends, because don’t we have enough to do with all the rest of the things that we are bombarded with? Not that it’s not our responsibility, but sometimes there’s just so much noise that we can’t see the signal that we need to help us make the right decision.
I envision systems like this could also start to work for monitoring home blood pressure (and so much more) at home. Blood pressure readings could go from the patient’s home device and automatically be entered into the right place in the EHR, and potentially linked up with data about compliance with their medication, physical activity, and even self-reported markers of stress in their lives and salt in their diet. Using systems smarter than us might allow us to make better decisions about when to change someone’s medication when that particular drug is not working for their blood pressure, and perhaps even help us figure out why it’s not working.
Example Two: COVID-19
For patients suffering from COVID-19 pneumonia, who are being managed in the outpatient world, perhaps noticing subtle trends in their clinical symptoms correlating with a ticking down of their oxygen saturation, can be used as an early warning sign that things are about to go wrong, and that the patient should come back in.
As we start to develop and design the new normal, the many new ways of managing medical conditions in the outpatient world, relying more and more on remote monitoring and patient-generated data, we need to make sure these are not just data dumps back to me, more entries in my in-basket, that I’m responsible for, that I need to do something about, without any additional assistance. When patients bring me in their paper blood pressure or fingerstick glucose logs where they are checking their BP and sugars five times a day, I find myself inundated with all that stuff, and it’s often really hard to do much beyond look at the gestalt, get a sense of how things are going. Do I need to make a change, add a medication, address compliance?
If we were operating in a fully integrated system, where this data went directly into the EHR, and then thoughtful, clinician-designed smart systems were able to detect the important signals hidden therein, perhaps remote patient monitoring could truly become the tool we need to provide all of our patients with the quality of care that they deserve.
The possibilities are better than remote.
Last Updated May 27, 2020
Source: MedicalNewsToday.com
