commit 592f0ca950ff2f446f06243d7c36b567eb242469
parent 0bcd6034b9d9b5d6b43d4b13fa43ad2f5eb78b41
Author: Beau <cbeauhilton@gmail.com>
Date: Sat, 23 Mar 2024 11:48:54 -0500
add post on talks website
Diffstat:
1 file changed, 165 insertions(+), 0 deletions(-)
diff --git a/site/posts/pp-tech-infra.md b/site/posts/pp-tech-infra.md
@@ -0,0 +1,165 @@
+# Patient-facing slide decks
+
+<time id="post-date">2024-03-23</time>
+
+<p id="post-excerpt">
+While I plan on being a physician-scientist forever, I'm not going to work in academia.
+Over the past few years I've put together very cool infrastructure for giving academic talks (click "talks" above),
+and it was making me sad that it might not get much use after I leave the university.
+But then, a ray of sunshine: Why not use my slide deck infrastructure to help in patient conversations?
+</p>
+
+## talks.beauhilton.com
+
+Academics give lots of presentations.
+I gave at least one 15min presentation per week in medical school,
+and still expect a few 15-45min presentations per half-year in fellowship.
+
+PowerPoint was long ago eschewed in favor of online-first presentation platforms,
+typically Google Slides.
+(Eventually Office365 came out and PowerPoint came on par with Google Slides,
+but I had already moved away.)
+No emailing around files and accidentally sending the old version,
+or running into an incompatible version of the software,
+and the presentation could be modified all the way up to
+(and, occasionally, during) the event itself.
+All you need is an internet connection.
+
+Eventually I wanted to move away from proprietary infrastructure altogether,
+so started using [Slidev](https://sli.dev).
+Slidev is geared toward developers,
+and as such uses all industry-standard web technologies,
+nicely packaged up to make it easy to go from a text file to a nice-looking presentation.
+
+The defaults are great, and, since it's all web standards,
+the sky is the limit as far as customization.
+
+The presentation lives in a single text file (markdown).
+This makes it easily versionable,
+tracked by git and backed up to [GitHub, GitLab, self-hosted git solutions, etc.].
+Since I keep a copy on GitHub,
+the ease of editing while on-the-go approaches the ease of editing a Google Slide
+(all you need is an internet connection and to be logged in).
+
+Once the presentation is online,
+I'm only ever a click or two away from accessing it.
+I make sure everything is HIPAA-compliant, and to the public internet it goes.
+
+It's been great for making talks in my academic program,
+and, when I thought I would stay in academia forever,
+it made sense to me to spend the time to build out my own infrastructure
+and subdomain for talks.
+
+## emphasizing the "physician" in "physician-scientist"
+
+Recently, I realized that I'd be happier shifting
+the physician-scientist balance more toward physician and less toward scientist,
+at least as measured by days per week dedicated to each.
+While I think I'll always see myself as as physician-scientist,
+my future oncology job (hopefully not "jobs," but who knows)
+will be in private practice, where I'll see patients 3 or 4 days per week
+and do research with the other 0.5-2 days that remain.
+
+(The following paragraph is an aside on how much time is spent on what in academia,
+and touches on why I decided to leave.
+I should probably move this out and expand it into its own post.
+It doesn't have much to do with the core of this one.)
+
+Classic academia is usually patient care 0.5-1 day per week, research the remainder -
+but that's somewhat misleading,
+since academics typically have a number of dedicated inpatient weeks,
+which may or may not put a damper on the ability to do research during that time,
+depending on how busy they get
+and whether it's July (with brand new trainees) or June (with seasoned trainees).
+Even accounting for this,
+it doesn't even out to the number of patient-care days per year in private practice,
+not by a long shot, but for most folks it isn't so freed
+from day-to-day patient care as it is advertised on the tin.
+There are a few academics who do a few weeks of inpatient care per year and that's it,
+no clinic,
+but that's the exception.
+You also pay your own salary out of your research funding,
+so there's a large time burden of not doing research,
+but writing grants to convince funders that your research is cool and worth funding.
+With the research I do and am interested in continuing,
+a university infrastructure adds little,
+where it isn't actively working against me,
+and I'd rather pay my salary, which will be much bigger, from taking care of people in clinic
+than working through the weekend
+(when I may also still be on call for the hospital) to meet grant deadlines.
+The university infrastructure is great for basic and clinical science,
+and some flavors of data science,
+but my scientometric stuff is truly better if done vigilante-style,
+not attached to academia's politics,
+and predictive modeling/etc. is available and important to private practice collectives (AON, USON, OneOnc),
+some of whom have more robust data infrastructure than the universities do.
+My most successful clinical data science project thus far was actually done under the purview
+of Business Intelligence for Cleveland Clinic,
+working with the Center for Clinical Artificial Intelligence,
+and the publication was the side project -
+this is much more in line with what I may find myself doing with clinical data science with one of the big three onc groups:
+do the thing, do it well, and if you feel like publishing it, gravy.
+The scientometric stuff is more classic research,
+where publication itself is the deliverable,
+but even here I'm much more interested in publishing a living website and tool
+than a paper in a journal.
+
+(And now, back to our regularly scheduled program.)
+
+Outside of academia, there's a high chance I won't give presentations nearly as often as I am accustomed to.
+At first, this made me simultaneously happy (presentations are so. much. work.)
+and sad (aww, my cute little talks website will just gather dust. Poor buddy).
+Then! I went to a dinner meeting to learn about renal cancer
+and get paid to hang out with friends and eat,
+and most of the physicians talked about bringing up primary data
+to help in discussions with patients.
+Most physicians do this when they think it will help,
+and it certainly wasn't the first time I'd heard people talk about it,
+but the timing was right.
+I realized that I have a great use for my talks website,
+and that I've already been using it this way for some time.
+
+## talks in clinic - data for shared decision-making
+
+At least 3 or 4 times in the past few months,
+I've been having discussions with patients about their prognosis
+and various options,
+and wished I had the graphs to show them without having to poke around Google
+and possibly fight with a paywall,
+then realized I just gave a talk on the same thing,
+the talk is on my website,
+and it has those graphs.
+
+It was very useful to go over the primary data with the patient
+(not everybody wants to see this stuff, but a significant number do),
+and having a talk I gave to reference immediately lent an additional modicum of trust to the conversation
+(most trust is built from showing you care,
+but another portion is built from showing you know what you're doing).
+
+Hearing the physicians at that dinner meeting
+talk about how they're using primary data in their shared decision-making,
+combined with my recent decision to go into private practice,
+and thoughts about what to do with my website,
+led me to the following idea:
+for diseases I'm most interested in,
+and for diseases I treat the most,
+it will be very useful to have a talk
+I made and maintain,
+geared toward patients,
+to immediately pull up and have available in clinic,
+and have the link to give the patient for review,
+particularly in those conversations with their loved ones
+where they're trying to remember all the stuff you said
+to make a group decision.
+It would also be easy and useful to have a slide describing the regimen
+and its schedule,
+especially if the way we're giving it doesn't quite match up with the published regimen
+(e.g. one of the many regimens that is written as "2 on, 1 off" but we're giving it every other week instead).
+
+I'm definitely not going to make a talk for every disease
+or use one in every conversation,
+that would be insane and impossible to maintain,
+but if I pay attention to the data I use repeatedly,
+a 30-60min session adding some of these to a slide deck
+with a little explanation
+would more than carry it's weight in future conversations.
