Future of Data in Care

What could data do for our patients with eczema? Raj Chovatiya, MD, PhD, discusses the future of data and algorithms in atopic dermatitis and dermatology. Dr Chovatiya is an assistant professor of dermatology at the Northwestern University Feinberg School of Medicine in Chicago, IL, where he also directs the Eczema and Itch Clinic.

Transcript
Dr Chovatiya:  There's a lot of big outstanding questions that we have, not only in atopic dermatitis and dermatology, but medicine in general. I'll list a few of them that come to mind. How do we improve long-term longitudinal care? How do we limit unnecessary ambulatory visits? How do we target care for those people who need to be seen more acutely or in person? A lot of it came to light during the ongoing pandemic in terms of how we allocate those resources. When it comes to resources, just thinking about how we lower cost, and when we actually have reduced cost, put them in the right location. This is where I think patient-reported outcomes, clinician severity measures, essentially data itself, and other analytics hold great promise.

I think that if we can collect this kind of information in a routine fashion from our patients, the same way that we have been conditioned to think about pulse, blood pressure, temperature, we are going to be able to track our patient symptoms and disease course and reactions to therapy over the long-term. It can guide how we do our appointments, that somebody, let's say, is mainly having severe flares during one time of the year. It allows us to focus on them during that time, versus somebody else who is pretty clear throughout the year, vs someone who is severe all around.

It allows us to better dictate followups so people aren't coming in any more or less than they need to. We can actually have automated prompts built into the data we're putting into an electronic health record. Oftentimes as clinicians, we're constantly thinking at the back of our minds about patients we've seen, whether we're worried about this patient, less worried about this patient. This is the way to make sure that nobody is falling through the cracks, and we're able to really followup in the long run anybody that gets flagged that needs to be seen more acutely, and gets placed in the right place for care.

We've seen other specialties that have embraced the use of patient-reported outcomes and shown that it's possible to design predictive models that allow patients to understand, based on their baseline factors, how a certain treatment or intervention may impact them in the long run. A really good example is orthopedic surgery. They've adopted patient-reported outcomes like the PROMIS measures for joint replacement, and they've been able to help track long-term outcomes and give patients better understanding of how their function or symptoms might be impacted in the long run by some particular type of procedure. I think this is where the opportunity lies for us. This is what I'm really excited about for how we are going to incorporate patient-reported outcomes in our care. Given how many new treatments are coming out, there's many more in the pipeline, I think the thing that we all think about is how can we match the right treatment with the right patient and give them the most accurate estimate of what their course is going to look like, what their response is going to look like, any side effects they need to look out for.

I think important elements of any algorithm that's going to be built are going to be many of the clinical domains that I've talked about previously, comorbidities, symptoms, burden, and things like that. We can use those patient-focused symptoms, we can use clinician-focused examination of severity scores, alongside basic demographic factors. We go to figure out which ones of those are going to be most important. Age, race, sex, things like that are relevant to this model or is it less relevant. I think all that together is going to allow us to just be able to pull this up at our fingertips and give patients a better idea of where their care is headed.

I think that one of the things that really makes us want to pull our hair out as clinicians is there really isn't a streamlined measurement system for what makes somebody fall in this mild, moderate, severe category, particularly in the case of moderate to severe atopic dermatitis, when we want to prescribe a systemic therapy. Some insurers or payer system might focus on what therapies have they failed to get that therapy? Others focus on what is the body surface are of the disease? Others focus on specifically just lesional severity if you classify them as mild, moderate, severe. This misses a more holistic approach to understanding the overall disease they protect, each hammering in at, is that...This only looks at one or two specific facets. It means that there is probably people not getting the level of care or therapy that they need.

This is an opportunity for us to introduce codified standardized data so we're all speaking the same language, we all understand what severe looks like in terms of itch, in terms of skin pain, in terms of sleep, in terms of the actual clinical scores, in terms of comorbidities, in terms of previous responses people have had to treatment.

We already have an electronic system where this stuff can be built in, and there should be a universal approach and platform where we're all looking at the same thing, and we don't have to keep constantly reinventing the wheel and doing something different for every single patient just to get something approved. In one sense, we have a lot of the tools, but the challenge is now going to be how do we leverage those tools, get the electronic data we need, and streamline the process so it's less of a burden for clinician, patient, and payers across the board.