Our Approach to Clinical Documentation: A Philosophical Shift

When people ask us what we do at River Records, it is tempting to start with the technology. Stream, our AI-powered documentation system, organizes clinical information by medical problem. The platform is modular, collaboaritve, and fast. Yet, the true origin of Stream lies not in its features, but in a persistent frustration, and a philosophy about what clinical documentation should be.

Copy-paste is often misunderstood as a shortcut, but in reality, it signals a deeper problem. In 2022, our team published a study on duplication in the electronic medical record, and the findings were eye-opening. Clinicians were copying large portions of text from one note to the next, not out of carelessness, but because they had no better way to preserve what was important. The systems available to them were brittle and inflexible, so clinicians created their own workarounds. They copied clinical details they could not afford to lose, duplicated histories and medication lists, and repeated plans, not because it was efficient, but because it was the only way to maintain continuity across time.

We see this not as a failure of individual clinicians, but as a reflection of the system’s shortcomings. If we want to fix documentation, we first need to rethink the way we approach clinical information.

Traditionally, the medical note has been a container, a time-stamped block of text tied to a single encounter. You write one, close it, and move on to the next. Yet patient care rarely fits into such neat boundaries. Problems do not resolve themselves at the end of a visit, diagnoses evolve, and clinical stories are always in motion. When documentation is organized around encounters, not problems, clinicians are left to reconstruct the patient’s story again and again, often with little more than a patchwork of templated prose. Context is easily lost, and redundancy becomes inevitable.

We believe it is time to move beyond this paradigm. Instead of anchoring information to the date of the encounter, Stream anchors it to the patient’s ongoing problems. Each issue, whether it is heart failure, depression, or a diagnostic uncertainty, is tracked in its own thread. This thread holds the relevant history, the evolving assessments, and the latest updates. It grows and changes as the patient’s condition changes, and it allows you to see both what matters now and what has come before, all in one place.

While the problem-oriented approach is not new, our vision is to bring it to life in a way that fits modern clinical workflows and supports better patient care. We are not here to replace the note entirely, and Stream can generate them when needed, but we no longer see the note as the central artifact of clinical work. The note is simply one lens through which to view the record, a helpful summary, but not the whole story.

Our true goal is not to help you write faster, but to help you think and understand more clearly. We want you to have the information you need at your fingertips, organized by clinical relevance and current context. This is why we built tools like Recap and Show Diffs, so you can immediately see what has changed since you last reviewed a problem. It is also why we allow for precise, problem-specific updates, rather than asking you to rewrite an entire narrative just to make a small change. We have let go of the assumption that documentation must start from a blank page every time, because clinical care is continuous, and so too should be the record.

There is a growing trend in healthcare AI to simply automate what already exists, but we believe faster is not better if it accelerates us in the wrong direction. We did not set out to build an AI that merely writes notes more quickly, but to build one that supports a new and better way of capturing clinical care.

With Stream, documentation is lightweight, flexible, and always focused on the problem at hand. Our platform is designed to support clinical reasoning, collaboration, and care continuity, not just compliance or billing. We want the record to serve clinicians, not the other way around.

This is the philosophy behind our work at River Records, and it is what drives us as we build the future of clinical documentation.