Why we built Journal Club into Stream.
A primary care physician would need about 29 hours in a weekday to keep up with the literature relevant to their patients. That isn’t a figure of speech — it’s the estimate from a 2004 analysis in the Journal of the Medical Library Association, which put the volume of potentially relevant primary care literature at over 7,000 articles a month and the effort to actually evaluate it at roughly 3.6 full-time physicians. A separate 2022 simulation study landed in the same impossible territory from the other direction: providing all the guideline-recommended care for a typical panel would take about 26.7 hours a day, before you’ve read a single new paper.
We have 24 hours. Most of them are already spoken for.
So every clinician quietly runs the same triage: there is evidence out there that should change what I do, and I will almost certainly never see most of it. We’ve learned to live with that gap. I’d argue we shouldn’t have to. And when I look closely at why the gap persists, it comes down to three things that have very little to do with how much any individual clinician cares.
Information scatter. The things I need are fragmented and they never meet. On one side is the medical record — my own notes, my problem lists, the diagnoses I’ve documented across a week of patients. On the other is the literature — millions of articles, indexed by their own logic, sitting somewhere I have to go and retrieve them from. Both are about my patients. Neither knows the other exists. Bridging them is manual, every single time.
Time scarcity. Even when I know exactly what I want to look up, the search is a task — and it’s a task that competes with documentation, inbox, and the next patient. It loses that competition most days. Not because it’s unimportant, but because it’s never the most urgent thing in front of me.
Decision fatigue. This is the one I think gets underrated. It isn’t only that reading takes time. It’s that deciding what to read, and when to go find it, is itself a cost. By the end of clinic, the part of me that makes good judgment calls is depleted. Asking it to also choose, from an undifferentiated ocean, which three papers are worth my evening is asking for the decision that never gets made.
Journal Club is our attempt to remove all three at once, and the move is simple: stop making the clinician go and pull. Bring it to them.
Each week, the feature looks at the conditions I actually saw — the diagnoses I documented, the ages of the people I was managing them in — and assembles a short reading list from reputable journals, with titles, abstracts, and a link straight to the source. On Monday morning it’s just there. I don’t decide what to read. I don’t decide when to look. I open the digest.
And I trust it, because of how it’s built. This isn’t a feed guessing at my interests from clicks. It’s derived from my own panel — the problems I’m managing, or at least the ones I’m documenting. The relevance isn’t a promise; it’s a property of the design. The information is finally moving toward me instead of waiting for me to come find it, which means the scatter — across the record and across the literature both — starts to collapse in the one place it matters: in front of the clinician who’s actually treating the condition.
That’s the part I find genuinely exciting, and it’s bigger than convenience. There is a long, well-documented lag between when evidence is established and when it reliably changes everyday practice. A lot of that lag is just dissemination failure — the right finding never reaches the specific clinicians for whom it’s most relevant. If relevant evidence routinely lands with exactly the people managing exactly those conditions, you can imagine practice patterns shifting a great deal faster than they do today.
I want to be honest about where this version stops. v0 is deliberately simple. There’s no commenting, no clinician feedback, no discussion. It reads your week and hands you a list — that’s it. But you can see what it opens up. Once the right findings are reaching the right clinicians, the natural next step is letting those clinicians respond to them — annotate, discuss, flag what changed their mind. That turns a reading list into something closer to real dissemination and communication of evidence, reaching the clinicians, and ultimately the patients, for whom it matters most.
For now, the ambition is smaller and, I think, worth it on its own: the reading should come to you. This week, it does.