What can I say? I can be a complete idiot when I’m distracted. For the full story, listen to this clip [.mp3] [.ogg] from episode 88 of the Linux Outlaws podcast. As a side note, Linux Outlaws is a great podcast – it’s funny, informative, and opinionated (and if you’re new to Linux, episode 92 is a fantastic way to find out what all the fuss is about).

I recently switched to Arch Linux and ran into a vexing problem – Bank of America’s SafePass authentication applet refused to load. The flash-based widget simply demanded that I update to the latest version of Flash, even though I was (of course) already using the latest version; the situation remained unchanged when I switched to GNU Gnash and Adobe’s beta Flash plugin.

It turns out that the solution is much simpler. For whatever reason, Arch Linux’s Firefox 3 packages still identify themselves to websites using the code name Gran Paradiso. Since the BoA site doesn’t recognize Gran Paradiso as a supported browser, it refuses to proceed any further. The solution is simple: type “about:config” into Firefox’s address bar, filter for “user”, and change the “general.useragent.extra.firefox” string to “Firefox/” (leaving the version number intact). Works like a charm!

Edit: Apparently Firefox packages in most Linux distros ship without the branding, so they continue to identify themselves with cryptic code names long after they’re released; installing a branding package (probably something like firefox-branding) fixes this problem more elegantly.

I’m a few months late to this party, but I’m guessing a lot of busy medical students are too. If you’ve been ignoring your EvidenceUpdates emails (don’t worry, even the best of us get overwhelmed), the editors of MedPage Today have compiled a list of the most influential and controversial stories of 2008. The editors don’t limit themselves to published studies (often because the controversy surrounds unpublished data or alterations in ongoing trials), but the article is still a good read.

My editorial summary is (for once) short, as summaries should be: be wary of surrogate endpoints. We know that the deleterious effects of CAD are related to plaque thickness, but you’d be a fool to think that’s the end of the story; likewise with type 2 diabetes and HbA1c. So when you’re told that drug A is the most effective first-line treatment for disease X, remember to ask exactly what endpoints have been investigated – and maintain a healthy degree of skepticism.

As an aside, EvidenceUpdates can give you a list of the most-read published papers in your field(s) of interest over the last year. This service is pure gold for medical students and residents – get an account (they’re free) if you don’t already have one. And no, they don’t pay me to promote their site – there’s just no justice in the world, is there?

I’m supposed to be giving a 10-minute talk on the surgical management of ulcerative colitis this Thursday at teaching conference, so (as always) I’m posting my presentation [.html] online in case anyone else can use it. For all statistics, I’ve included references in the primary literature when possible and references to Sabiston otherwise (the footnotes are clickable hyperlinks). You can navigate through the presentation using the arrow keys, or you can mouse over the lower right-hand corner of the slides and some controls should pop up. If you’re so inclined, pressing “t” will toggle between presentation mode and my scanty speaking notes.

This is also my first attempt at using Eric Meyer’s S5 presentation format with John Manoogian’s Glossdeck theme, and they’ve been really great so far! S5 uses a nice mix of JavaScript, XHTML, and CSS to make simple, standards-compliant presentations that can play in any modern browser. Though I haven’t experienced this yet, I’m told that the images won’t scale well at resolutions other than 1024×768. I have to say that it’s really nice to be able to write a presentation in a text editor rather than something intensely visual like PowerPoint or Keynote. Like writing LaTeX markup instead of word processing, it lets me focus on content rather than style. It’s also nice to use when I’m trying to work in little snippets throughout the day, using whatever workstation I happen to be near. As a side note, I’d originally planned to test out Google Presentations, but the site was down for a while this afternoon and the thought of a server failure during the conference almost made me pee my pants.

Update, 10 Jul 2008: This talk was well-received by a group of surgeons renowned for their merciless pimping, so you may get some mileage out of it. The meta-analysis referenced on the DALM slide was particularly good fodder for discussion.

Looking back on the last three years, it’s really amazing how much knowledge medical school has crammed into my head. My undergraduate education didn’t focus on medical science; nearly all the biology I studied was molecular in scale. Despite having a degree in biology from a good school, I’d have been completely unable to explain how to evaluate and treat even something as basic as diarrhea. Three years later, it’s become difficult to talk about anything related to health and disease without using a barrage of jargon. And I’m still a medical student – who knows what I’ll be like after twenty years in practice. It shouldn’t be a surprise that most physicians’ explanations leave their patients completely bewildered (especially since the average patient’s health literacy is shockingly limited).

[quote]

Part of the medical student’s job on most services is to field questions from patients. Your attending and resident probably don’t have the time or inclination to explain what heart failure is, how it causes problems, and why the usual medications work, but those explanations are crucial to patient adherence. Ideally, the explanations you give should better equip patients to deal with their other healthcare providers, too. A nurse in the hospital at Vandy just turned me on to a great website for patients and doctors called AskMe3. The website is part of a larger initiative by the National Patient Safety Foundation to help facilitate communication between patients and providers. It recommends that patients ask (and physicians try to answer) three simple questions during each visit:

1. What is my main problem?
2. What do I need to do?
3. Why is it important for me to do this?

After all, these are the basics – and the last one is the most important and the most neglected. “You’ll go into respiratory failure” is more of a threat than an explanation, but “You’ll have a harder and harder time getting around the house and keeping up with your grandchildren” might work better. Do your patients a favor – try to give meaningful, simple answers to these questions and direct them to the AskMe3 website (or give them printouts). It doesn’t help much if you tune up your patients and then discharge them with no idea of how to manage their problems.

As a side note, the strongest predictor of poor health literacy is poor literacy overall. That should be obvious – if patients aren’t comfortable with reading, they’re likely to have an exceptionally difficult time understanding medication inserts and even online resources like WebMD. Initiatives like AskMe3 are great ways of coping with poor health literacy, but the foundation of good healthcare is a well-educated population – and that means better public schooling.

The Beauty of LaTeX

I’ve always been a little unsure why medical schools (and schools in general) don’t make more extensive use of free or open-source software (that’s free as in speech, not free as in beer). It seems like a win-win situation to me: schools don’t have to pay for huge installations of expensive software, and professors and students can keep their data in open document formats that will not become obsolete. The PDFs, PostScript files, and plain text files I wrote in the mid-1980s are still easy to open and work with – can you say the same for all those old WordPerfect files? After putting so much effort into creating decent lecture notes and presentations, professors shouldn’t have their hard work suddenly rendered incompatible with the latest tools. This issue is particularly important for medical schools and hospitals, where increasingly large quantities of patient information are being stored in proprietary formats accessible only through closed-source programs. Vanderbilt, for example, uses an electronic medical record system that only functions well in Internet Explorer. Thus, every workstation in the hospital must run Windows (which undoubtedly costs the medical center thousands of dollars a year), and the systems administrators must upgrade operating systems and reprogram the EMR according to Microsoft’s release schedule. This approach lacks foresight – while the information itself is (I hope) actually stored in an open-format database, access to that information depends on a notoriously unstable and insecure operating system and browser.

And beyond these practical issues, there’s also a compelling moral and philosophical argument to be made. Schools should be institutions that encourage curiosity, innovation, and the free exchange of knowledge – all ideals that run directly against the spirit of proprietary software. Proprietary software is locked down with copyrights and patents, preventing students and professors from reading or modifying its source code. Free software, on the other hand, generally comes with an invitation to re-use, modify, or improve the code to benefit the community. By using free software, even schools and students who are uninterested in programming can express their support for freedom of thought.

Now, I’m not suggesting that every school and medical center immediately switch to Linux and ditch all proprietary software (though that would be an excellent goal). Instead, there are some easy and painless substitutions anyone can make.

1. Switch the default browser from Internet Explorer to Firefox.
2. Install OpenOffice rather than Microsoft Office, and encourage faculty and students to make the same switch on their personal computers.
3. Install GIMP rather than Photoshop. Or install GIMP rather than nothing, which seems to be the current standard (who in their right mind uses PowerPoint to edit images?).

For the more adventurous (or the more tech-savvy):

1. Use LaTeX for document preparation. Your documents will be better-structured and more beautiful (the image in this post is an example of the gorgeous output LaTeX is capable of).
2. Use S5 rather than PowerPoint or Apple’s Keynote.
3. Use Ubuntu Linux for a good introduction to free computing bliss.

I gave a 5-minute presentation on the drug entacapone (Comtan) last month on the wards, and thought that the handout [.pdf] I made might be useful to some other folks. It talks briefly about the drug’s mechanism of action, its pharmacokinetics, and its clinical utility. Most of the info came from a great review article that’s available free of cost online. You’re welcome to use this to guide your own research or to prepare a short talk; as always, don’t distribute it verbatim without crediting the author.

For all the geeks out there, I had to write this handout at the school’s computer lab, where the draconian sysadmin won’t install LaTeX. I ended up being forced to use Microsoft Word 2008 for the first time, and I was pleasantly surprised. I still think that word processors are misguided right out of the gate, but the latest version of Word is leagues ahead of its predecessors. It allows at least somewhat convenient logical markup, has reasonable default font combinations, and has reorganized the menus to be much more task-specific (writing, editing, footnoting, etc.). I still much prefer Vim and LaTeX’s ease of use and visually superior output, but the thought of writing in Word doesn’t make me cry anymore.

It’s a sad fact that most medical school lectures stink. When was the last time attending lecture added something really valuable to reading a good textbook? Too many lecturers distribute “notes” that are little more than paragraph-by-paragraph summaries of a major textbook and then spend their classes reading these notes to their students. Obviously, lecturing gives professors valuable time to both impart information in a different sensory modality and highlight key points. However, most of a lecture’s impact is lost if it’s just a numbingly dull reiteration of both the notes and the textbook. Instead, lectures should be as lively as the material permits; they should be a time when connections between subjects and topics can be made explicit and overarching themes drawn out. Incidentally, lecturing effectively can usually reduce the number of class hours needed to cover each topic.

I’m overly critical of nearly everything and everyone, but I’m particularly incensed by terrible lectures. It’s not that hard to prepare a halfway decent lecture, especially since the same materials can be used year after year. I know: I both taught and wrote curriculum materials for a successful middle school. That said, here are some concrete suggestions for improvement:

1. Assign reasonable textbooks. If professors would recommend reasonable textbooks (e.g., Robbins Basic Pathology rather than the full Robbins) and require students to read them, they’d make everyone happier. Professors wouldn’t feel obligated to write exhaustive notes (essentially re-writing the textbook) and students wouldn’t have to cope with the amateur typography and inevitable errors that creep into 20-page handouts. If students need more detail, they can always consult the standard references in the library or on MDConsult.
2. Don’t cover the details in lecture (or in course handouts). Seriously – just don’t do it. Rather than trying to cram every detail into an hour, professors should assume (and require) that students read the relevant sections of the text before coming to class. Every possible exam question doesn’t need to be covered in lecture or in the handouts – if the required text and the exam questions are reasonable, students won’t complain more than usual.
3. Treat lecture time as the precious commodity that it is. If professors spend hours of lecture time re-hashing the same material that’s already in the handouts and the textbook, they shouldn’t be surprised if many students don’t attend. They shouldn’t be surprised that the students who are there send text messages or surf the internet. And they shouldn’t try to play the professionalism card – by not putting in the effort to prepare a decent lecture, they’re just as unprofessional as the class truants (especially if they’ve also plagiarized the lecture material and its exact organization from a textbook). Instead, use lecture to push students a little higher on Bloom’s taxonomy and encourage an integrated understanding of the material. At the very least, use lecture as a time to highlight the key points rather than trying to cover all the details.
4. Don’t plagiarize. Yes, professors and students are allowed to draw on textbooks and published papers according to standard fair use criteria. That doesn’t mean that professors can attempt a sentence-by-sentence conversion of the textbook into an outline and call the resulting mess a lecture handout. If a student did the same thing and passed it off as an essay, it would clearly be plagiarism even if he cited the textbook; why should the rules be different for professors? The point of notes isn’t to replace the textbook, it’s to highlight key points and themes. Don’t lose the forest for the trees.
5. Pay for curriculum development. Writing school curriculum is a learned skill, and it’s not one that every faculty member has the time or inclination to develop. Maybe medical schools can’t afford to pay lecturers for time spent in class and hours of prep, but they can certainly pay a one-time fee for curriculum development. If professors already had a good text chosen and decent slides and handouts written for them, they’d be able to focus on amending the materials to suit their personal preferences and delivering good lectures. And rather than hounding each faculty member, administrators who care about the quality of med school education could work with one faculty member and finish the course materials once and for all.

That’s just my take on the situation. Like every medical student, I’m always certain but often wrong. I’d love the chance to put these ideas into practice, though – maybe someday I’ll try to develop a set of materials for a single unit (e.g., cardiovascular physiology) as a demonstration.

I just read a great article on the clinical and educational value of the problem list. The author is a hospitallist at Duke, and he argues (quite convincingly) that documented plans should be problem-based (rather than system-based) everywhere outside the ICU. Part of his argument is that problem-based lists are actually more useful; he argues this point well, though he doesn’t give much attention to the opposing argument. However, the more interesting portions of his article argue that system-based assessments are contrary to the investigative spirit of internal medicine and that they reduce the educational value of each patient encounter.

He quotes extensively from Dr. Lawrence Weed, who invented the SOAP note in the 1960s and was a strong proponent of problem lists. Arguing in favor of integrative problem lists, Dr. Weed says,

It is this multiplicity of problems with which the physician must deal in his daily work…[T]he multiplicity is inevitable but a random approach to the difficulties it creates is not. The instruction of physicians should be based on a system that helps them to define and follow clinical problems one by one and then systematically to relate and resolve them…[T]he basic criterion of the physician is how well he can identify the patient’s problems and organize them for solution.

Interestingly, Dr. Weed places the burden of enforcement not on students and residents, but on their supervising physicians. He says,

The education of a physician…should be based on his clinical experience and should be reflected in the records he maintains on his patients…The education…becomes defective not when he is given too much or too little training in basic science…but rather when he is allowed to ignore or slight the elementary definition and the progressive adjustment of the problems that comprise his clinical experience. The teacher who ultimately benefits students the most is the one who is willing to establish parameters of discipline in the not unsophisticated but often unappreciated task of preventing this imprecision and disorganization.

The Department of Medicine at Vanderbilt strongly encourages residents to write problem-based plans. In fact, the electronic medical record here actually requires it. While we’ve been told that problem-based lists are preferred outside the ICU, I can’t remember anyone ever explaining the reasoning behind this push. After reading this article, I’m sold; if problem-based lists are going to improve my training, I’m all for it.

Your medicine attending probably only sees you on rounds. That means that his evaluation of you is based solely on the few minutes you spend presenting patients each morning (and possibly your written notes). Don’t let this opportunity go to waste – with a little effort, you can be a rock star!

Communication in medicine is an art, and it’s one that medical personnel stink at: studies suggest that poor communication is the most common root cause of medical error. As a medical student, remember that your clinical years are intended to teach you the fundamentals of practical medicine and how to function effectively as part of a team. Do yourself a favor now and focus on your communication skills, beginning with your presentations. If you need a model or aren’t convinced this is worth the effort, evaluate your residents: senior residents who can give brief, fluent presentations that highlight key information get more done and have fewer patients crump.

[quote]

1. Be brief: Five minutes is the absolute limit, and three minutes is better. Nobody (not even your fellow student) wants to listen to your ten-minute monologue, and long presentations make rounds take forever. Don’t try to anticipate and answer every question, and don’t compress your presentation by talking faster! Keep in mind that your attending works from a 30,000-foot perspective, so his information needs are very different from your intern’s.
2. Follow standard structure: Deviating from the standard presentation structure (CC, then HPI, then PMH/PSH…) confuses your listeners. If portions aren’t relevant to your patient’s hospitalization, eliminate them entirely.
3. Fight word bloat: Transitions between sections of your presentation are awkward and unnecessary (especially the dreaded “in terms of”) – just pause for a second and move into the next topic. While speaking and while writing, take care to avoid wordy constructions and phrases that lack an agreed-upon meaning. In particular, avoid saying “The patient describes X and may have felt a little Y, but doesn’t think that he had too much Z.” Instead, say “The patient had X and Y but not Z.”
4. Focus on the assessment and plan: Though usually the least emphasized part of the presentation in physical diagnosis courses, this is where the money is. As a student, this is where you show off your knowledge (well, your ability to filter out irrelevancies is also a measure of your clinical acumen) and demonstrate that you want to own this patient’s care. As a clinician, this is the section of the presentation that keeps everyone on the same page and directs the day’s activity. Again, remember to focus on the big picture – your attending doesn’t care about the minutiae.

I’m the first person to admit that I break these guidelines often, but at least I feel bad about it. Incisive presentations are a mark of incisive thought; since I aspire to the latter, I cultivate the former.