HJBR May/Jun 2026

THE FUTURE OF HEALTHCARE has been checked at least once in accordance with current guideline recommendations, and calculate the patient’s 10-year atherosclerotic cardiovascular disease risk using the pooled cohort equations. Ideally, this would be done using the newer PREVENT risk score, which has been shown to offer improved risk estima- tion in certain populations, but it has yet to be meaningfully integrated into most electronic medical records. This preparatory work allows me to define appropriate LDL and non-HDL cholesterol targets that are specific to each patient — values that must often be manually calculated and entered — as well as determine whether statin or other lipid-lowering therapy is indicated. As with other aspects of care, this process depends heavily on manual review, cal- culation, and synthesis, rather than being sup- ported natively by the system itself. And even as I contemplate a final chapter of my career in which I imagine an eventual return to full-time patient care, I do so with some de- gree of trepidation — mindful of the workload demands and the very real possibility of expe- riencing burnout a second time. However, if I were ever to return to full-time patient care, my hope is that it would not simply be a return to the same system with marginal improvements — but to something fundamentally better de- signed. The future of healthcare technology cannot be defined by its ability to document more, code more accurately, or store more data. It must be judged by whether it meaningfully re- duces cognitive burden, clarifies clinical think- ing, and strengthens the connection between patient and physician. The highest-functioning systems in other industries do not ask their us- ers to do more work; they quietly remove fric- tion, anticipate needs, and allow the human interaction to take center stage. Healthcare should be no different. Technology should operate in the background, organizing infor- mation, surfacing what matters, and enabling clinicians to spend less time documenting care and more time delivering it. Systems of Record vs. Systems of Action Current electronic medical records function quite effectively as systems of record, but far less so as systems of action. They are highly ca- pable of storing and enabling retrieval of vast amounts of clinical data, yet they struggle to translate that information into timely, action- able insight at the point of care. This limitation is not trivial. It is often cited that it takes, on average, approximately 17 years for evidence- based medical advances to be widely adopted into routine clinical practice — a gap that re- flects both the complexity and fragmentation of healthcare delivery. If the next generation of clinical decision support could instead bring these advances directly to the front lines we would begin to see a meaningful shift. In such a model, tech- nology does not simply archive information; it actively participates in care delivery, helping to translate evidence-based science into real, measurable improvements in outcomes. A true system of action would not rely on manual re- view, fragmented data gathering, or clinician recall; instead, it would continuously synthesize the patient’s information, calculate risk in real time, and surface the next best action — bring- ing evidence to the point of care rather than asking clinicians to go find it. And if both our payment and delivery sys- tems were intentionally designed to measure and improve health outcomes, rather than sim- ply reward activity, such a technology platform would become not just useful, but essential. Consider a common scenario in primary care that I recently encountered: a patient with diabetes, albuminuric chronic kidney disease (CKD), and elevated cardiovascular risk. Re- search has shown that optimizing blood pres- sure and using agents such as SGLT2 inhibitors can slow CKD progression, while statin therapy reduces cardiovascular events in patients with CKD. And yet, in practice, these therapies re- main underutilized; real-world studies suggest that only about 10% to 33% of eligible patients with diabetes and CKD receive an SGLT2 in- hibitor, despite strong guideline recommenda- tions. As I reviewed the patient’s chart, I saw that a physician had helped the patient improve her A1c from 9.0 to 7.5 while on an SGLT2 inhibitor — an important achievement. But this patient also had albuminuric chronic kidney disease, a condition that markedly increases the risk of kidney failure, cardiovascular events, and mor- tality. And yet, in the note, the SGLT2 inhibitor was framed primarily through the lens of dia- betes treatment and potential side effects, with “For technology to fulfill its promise in healthcare, it must do more than digitize work; it must eliminate the mundane, restore attention to the patient, and create space for the human connection that sustains both patients and those who care for them.”