Addressing a two-headed monster of access and quality, representatives from our institution’s accountable care organization came to us with a proposal for a way to try to address a metric that we continue to do poorly on throughout the practices.
According to the data in their reports, across our institution our diabetic patients get annual screening for retinopathy done at a rate of approximately 30%. As you can imagine, this is well below the threshold that most organizations think is “best”, and is thus worth putting a lot of energy and effort into trying to improve.
When they presented this data to us, we had a lot of discussions about why the numbers are so low, and they listened to us when we said that many of our patients have ophthalmologists outside of the institution because of access issues, or fail to show up for the appointments within our own institution where we refer them most often.
When one of our patients goes to an “outside” eye doctor — and even in the best-case scenario, when we get a report back from them showing their findings — the information that this diabetic patient does not have retinopathy only makes it into the chart as a checkbox if we actively not only scan the report into the system, but we also go to the health maintenance tab in the electronic health record (EHR) and indicate the day that they had the screening done.
Due to these limitations, 70% of patients for whom the intervention is recommended don’t have the screening done, at least according to retrievable data from our EHR.
An Interesting Solution
The solution that has been proposed to overcome these issues, which they are already trying at several practices throughout our institution, is an interesting choice, a different way of looking at things, and it opens up a couple of cans of worms as we think about its implementation in our practice.
It turns out that one of our ophthalmologists has been working with a medical optics company and a large unnamed technology company to develop a freestanding retinal camera that can produce a complete 180° image of the back of the patient’s eye without the need for dilation.
Their plan is to introduce these cameras into every practice that manages diabetic patients, and offer to have every one of our diabetic patients sit down and get a picture taken of the back of their eye, which is then sent over to the ophthalmology clinic where they can analyze the image and screen for retinopathy without actually laying eyes or hands on the patient.
Apparently, where they’ve already tried using this, it’s been quite successful, and about 90% of patients had no findings that required anything further beyond those initial photographic screenings. A small percentage had diabetic retinopathy, and there were a few scattered incidental findings discovered as well that needed follow-up and ongoing care.
But what was missing here?
To me, it feels like this new technological process, while it may lead to 100% compliance with the guidelines, may in fact lead to a patient feeling like something was missing from their healthcare. Namely, an eye doctor.
Proceed With Caution
Notice I did not say ophthalmologist. While eye doctor and ophthalmologist are clearly synonymous, I use the word doctor to reinforce the sense that if each patient has an eye doctor taking care of them, someone who talks to them, carefully examines their eyes, sees the bigger picture while helping the patient see better, I think that’s better for patients than a remote camera and a highly trained technician sitting in the dark somewhere else reading screening images.
Now don’t get me wrong — for so many patients who screen negative, there’s probably not a lot of added value to actually spending all that time over at the ophthalmologist’s office, and for many patients access is an issue, the cost of the care is an issue, and the valuable use of that provider’s time is an issue. But I think as we began to expand the use of more and more technology to replace more and more tasks that have been an integral part of healthcare, we need to proceed with caution.
One of the stated goals of the companies behind this product, this high-tech camera, is to gather an enormous amount of data, scores and scores of images of retinas, to provide input to an artificial intelligence system that will ultimately allow a computer to read these screening images, without the input of a physician.
As technology becomes better and better, and we learn more effective and efficient ways to incorporate it into helping us take care of our patients, we will undoubtedly find ways to replace a lot of what we do with intelligent neural systems capable of doing some of what we’ve always done.
There’s value in that, but there is of course always a loss of some sense of who we are and what we do that goes with this.
The Virtue of the Physical
Many years ago, as a resident working in the critical care unit (CCU), we would admit patients overnight with a cardiology fellow, and then in the morning, with the sun rising outside, as a team we would present our findings and recommendations to the CCU attending who was on service with us that month.
I recall one early morning standing at the bedside of a sick elderly patient, who came in the night before with a heart failure exacerbation, was appropriately triaged and treated, and was now doing fairly well in the light of day.
As part of his plan, the fellow told the attending that he would like to get an echocardiogram on the patient that morning to assess the degree of his cardiac dysfunction.
I can still see how the attending gently guided the fellow and myself through a repeat history and physical examination, as we once again in an unhurried fashion laid our hands on the patient, and told the attending in words what we felt and heard and saw, where the jugular venous pressure was, where the cardiac point of maximal impulse was, what the delay in the distal pulses told us.
The attending finally turned to us and said, “So, what do you know?” It turns out we knew pretty much exactly what the echocardiogram was going to show.
I’m not suggesting we go back to the days where we relied solely on the history and physical, nothing in me wants to abandon the armamentarium of technology we already have, nor ignore the potential of technology momentum coming down the pike.
If this new camera provides information that helps me take better care of my diabetic patients, by knowing who desperately needs maximal interventions to get their diabetes under better control, then it’s all worth it at almost any cost. But I want to make sure that patient still has an eye doctor, still have someone looking at their eyes as part of their body, as part of them as a patient, and helps make sure they can always see their way through the world they need to navigate in.
Maybe eventually a great deal of what we do will be replaced by these computers, these robots, these devices full of neural networks and artificial intelligence.
X-rays and CT scans and MRIs will be read by machines. Cameras will scan every inch of our skin to pick up the precursors of malignant melanoma and other skin cancers.
But I hope we keep the doctors around, because I do believe that the care we provide is more than a brute analysis of data, or a crunching of numbers. This technology can help us, but I’m sure that it can never replace us.
At the end of the 1940 Disney movie “Pinocchio,” when he has attained his wish and shouts out, “I’m a real boy!”, he has achieved the humanity that he was unable to find despite having feelings, the capacity for thought, and the ability to propel himself forward.
He was, before that moment, wooden sticks and levers and hinges and wires controlled from above by strings, manipulated at the whim of others. And then he changed, becoming something more.
Technology will never be that; it will never be human. We owe it to our patients to always keep that in the care we provide them.