Everything in its right place
How electronic health records fail at collecting what matters most
Early in my life I remember participating in an Easter egg hunt. I, like so many of the other kids there, was so excited to be chasing after eggs, and the always and inevitable surprise inside. At least that’s what I thought. You see, it was one fateful Easter morning where I was preparing for an egg hunt with my friends when I learned that not all Easter egg hunts are made the same. In fact, this one, arguably the most egregious and offensive type, had a hunt where there was NOTHING inside the eggs. I mean nothing. While running after eggs was fun and all that, it was the true treasure of milk chocolate goodness I was after, in whatever form that took. I learned that sometimes what we collect doesn’t matter if what’s inside what we collect isn’t what we are actually looking for.
I think by now we all know the most significant things that impact our health are not related to health care. Sure, it’s great to have health care when you need it, but the countless factors that go into our health are so far outside the health system space that it’s a surprise we don’t spend more time talking about them and connecting them to health.
As I am sure 99.9 percent of you know, when we talk about social determinants of health (SDOH), we're referring to the conditions in the environments where people are born, live, learn, work, play, and worship that affect a wide range of health, functioning, and quality-of-life outcomes and risks. These conditions include factors like housing, food security, education, neighborhood and physical environment, employment, social support networks, and access to health care.
Research has shown over and over that SDOH has a profound impact on health outcomes. For example, where you live matters long term for your health. Another study found that food insecurity was associated with a higher risk of depression. Despite the clear impact of SDOH on health outcomes, these factors have traditionally been overlooked in health care especially in what we pay for and how we document care.
This is beginning to change, with a growing recognition of the importance of addressing SDOH; however, there are still significant barriers to integrating SDOH into health care, particularly in the context of EHRs.
Yes, those pesky Electronic Health Records (EHRs), which we all welcomed to the world with great optimism, as they promised enhanced communication, improved record-keeping, and better patient care. It was a new promise land, or was it? While EHRs have even been questionable at delivering efficiencies, a worrying trend has emerged - the focus of EHRs has become skewed towards "clicks" – data entry representing completed tasks within the EHR system. This shift has led to a concerning relationship between increased EHR use and physician burnout. And of course, there are data on this, too!
“What, what was that you tried to say?”
While burnout is a big issue, it’s not the only one when it comes to EHRs. It’s also the type of data we are collecting and why. Our emphasis on clicks has pulled our attention away from the bigger picture – the actual health of our communities. By neglecting social determinants of health within EHRs, we miss crucial factors impacting patient well-being and fail to address the root causes of health disparities.
Wait a second, Miller, why the fervent push to address this topic in such a specific way? Well, my dear reader, today I was able to present to a committee for the National Academies of Science, Engineering, and Medicine (NASEM), on this very topic. And yes, much of what I am saying here, I said to them. Here are my topline recommendations:
Standardization is Key: One of the key challenges as to why we don’t include more social data in EHRs is the lack of standardization in capturing SDOH data. While some EHRs have begun to include fields for collecting SDOH data, there is no universally accepted set of codes for capturing this information, making it difficult to analyze and share across health care systems. Additionally, there are major financial disincentives to collecting SDOH data, as the current fee-for-service payment system often rewards volume over value, which has almost become a cliche statement in 2024, but remains true in all the worst ways.
Embrace Technology: We know that many of the new and exciting tools with AI/Large Language Models (LLM) and Natural Language Processing (NLP) tools can help both streamline the input of SDOH data but also automatically extract SDOH data from clinical notes that may not be in discrete fields.
Patient Empowerment: We should also integrate patient-reported SDOH data collection into EHR workflows. These data may be the purest gold we can find, and also empower patients to share their social circumstances in a more authentic way. Honestly, without these data, I am not sure we every truly have a comprehensive picture of their health.
When I think about so much of the work I have done in my career to do things like integrate mental health into primary care, there are certain similarities that stand out. For example, practices and providers have to create clear and consistent EHR workflows for the collection and use of SDOH data. You have to incentivize and hold people accountable whether that be financial or otherwise to ensure that SDOH integration and appropriate interventions occur.
From 30,000 feet, incorporating SDOH data into EHRs will not only provide a more complete picture of patient health, but it will also enable health care providers to better address the root causes of health disparities. Taking a more comprehensive approach to health allows a more intentional and meaningful way to achieve positive outcomes for all.
I want an EHR that has what I expect inside. An EHR that has the type of information I really want people to know about me as the chances those data are relevant far outweigh what some random person is telling my provider to collect data on me for. I know this is possible - it’s just about creating the right conditions that allow this massive cultural shift to occur in health care. Then perhaps everything will be in its right place.
All of this is so spot on and points to a glaring need for patients and their medical providers to understand and give the best possible care and support. Our futures are all linked to the zip code we're born into.
To that point about zip codes. I wonder if simply gathering the patients zip code and using AI/LLM to sort through zip code data the physician can get a better sense of SDOH. Or have the AI based on Zip Code entered ask additional questions? I may be lost in my own private Idaho here. Zip Codes and AI are data rich.
Radiohead in the title! Sorry, haven’t read it yet but needed to say that immediately. Back to reading.