It is always a pain clicking on a link to a journal article only to hit a paywall. It’s doubly painful when I know I have institutional access via my library’s proxy server, but have to jump through hoops to get it: go to the library website, log in, copy and paste the article name or PMID into its PubMed search box, and finally download the PDF. Arduous, and—turns out—unnecessary.

Enter Alfred 2 workflows. Here’s a nice article I found on Twitter today. The NEJM link in the top right corner leads to an abstract, but I need a special archive subscription for the full PDF. No matter—I can just highlight the PMID and hit my special Alfred 2 keyboard combo:

Since I’m not already logged into the Welch library proxy, I hit a login wall. It’s nothing 1Password can’t solve, but you can also just type in your username and password yourself, like an animal.

And Bam! The ugly but magic button is where it should be. Your institution might have a prettier one.

To make it clear—this simple workflow will do a PubMed search of any selected text anywhere in OS X, all through your institutional proxy server. Finding an interesting reference while reading an article, highlighting its title, and hitting ^⎇⌘P to get to the PDF always feels like magic.

You can download the workflow here.

AuthorMiloš Miljković
Why doctors shouldn't use Google services

Anyone using Gmail or Google+ can now email you without knowing your address. You can disable this “feature” in the settings, but having it be opt-out shows yet again how little Google cares about privacy.

Not that there’s anything wrong with that—privacy is a relatively modern invention that younger generations might not care for as much as we do. But you should understand the implications when patients and random strangers start leaving messages in your personal inbox. Suing doctors is not a modern invention.

This is why I stopped using all Google services—search included—years ago. The company has become so large, with so many users, that it doesn’t need to cater to fringe interests. Yes, when a business has billions of users, doctors are a fringe group—one that hates change-for-change’s-sake[1], having to re-learn an interface “just because”, and not being the true customer[2].

Also, the number of people at Google who may access my data is huge. FastMail[3], my email provider of choice, has fewer than 10 employees. Gmail alone has hundreds. Not that anyone would be interested in me in particular, but if I ever inadvertently send or receive private patient information through my personal account, I’d rather as few people as possible see it.

Email is fine, but why abandon search? Because I had at one point googled enough ailments and substances, common and obscure, that the add algorithm thought I was an elderly female recovering heroin addict with more than one paraphilia. The adds it served me were, in that sense, appropriate. In addition, the only valuable first-page results I got were Wikipedia entries. Everything else was a hodgepodge of useless Livestrong, Huffington post and five-pages-per–500-word-article-AND-behind-a-login-wall Medscape links. Duckduckgo and, yes, Bing at least eliminate the first problem while not making the second one any worse.

Google calendar is the only service still standing. It is fast, reliable, omnipresent and easy to use. There is, however, that constant nagging fear that they will find some way to integrate it with Google+ and yet again sacrifice functionality to force people into its circle[4]. This is why I use Apple’s iCloud calendar, its horrendous web interface and all.

Also: Reader. I use FeedWrangler now, but man.

Physicians’ concerns aside, Google is all set to become the network TV[5] of the internet—large, bland, and largely not relevant to the people who are. It is already two-thirds of the way there.

  1. I like change and think a certain amount is necessary to make life interesting, stave off dementia, etc. But I get enough just keeping track of the latest JNC, ATP, USPSTF, and—coming soon—NCCN guidelines. I’d rather not have to remember where a random designer decided to put the Forward button this time.  ↩

  2. This one in particular, as it keeps reminding me that doctors are second-class citizens in the tech world. Electronic health records are made with the billing departments in mind—we are there to provide content. Google services are created to sell adds—we are there to provide eyeballs.  ↩

  3. Yes, it’s an affiliate link.  ↩

  4. crickets  ↩

  5. Or Microsoft.  ↩

AuthorMiloš Miljković

Marco Arment has just discovered an old article in The Atlantic pronouncing the triumph of New-Age medicine. It’s been a while since I’ve read it, but the introduction reminded me of what I thought was its biggest fault in reasoning:

… But now many doctors admit that alternative medicine often seems to do a better job of making patients well, and at a much lower cost, than mainstream care—and they’re trying to learn from it.

Alternative medicine does not make patients well. It makes them feel well. The difference is huge.

Here are two graphs from an excellent free-to-access NEJM article that compared four methods of treating asthma: conventional medicine, placebo, sham acupuncture, and doing nothing[1]. The first one shows how well the patients in each group felt after 2–4 weeks of treatment.


Ah ha! Albuterol was no better than placebo medicine or sham (sham!) acupuncture. Alternative medicine wins! Or did conventional medicine lose? At the very least it’s a draw.

Not so fast. The second graphs shows the amount of objective improvement, measured in FEV1—the volume of air you exhale during the first second of breathing out:


If this were the common cold, it wouldn’t have been a big deal. But asthma is not the common cold. People die of it every day, not because they didn’t feel well—though being unable to breathe is doubtlessly uncomfortable—but because their airways were too tight to get any air out of the lungs.

This is why alternative medicine can be dangerous in the wrong hands, with the wrong patient. Improving quality of life is important, but so is curing disease.

  1. Adding real acupuncture to the interventions would have made the study perfect. Some other time, perhaps.  ↩

AuthorMiloš Miljković

January 1st seemed to be a good day to install Brett Terpstra’s Slogger. Every night, its army of gnomes will go over my tweets, blog posts, completed to-dos, etc. and record them in a Day One journal entry. Not a replacement for a real journal, true, but better than anything I could do on my own.

It’s a Mac-only app that runs from the command line—not user friendly at all. Even so, the installation instructions are straightforward, with some caveats for the not-too-bright, like me:

  • Plugin configuration is done in each individual plugin.rb file, not slogger_config.
  • All config strings (URLs, file paths, usernames…) should be in quotes (“…”), even when in an array (i.e. in square brackets). The Twitter plugin instructions wrongly give an example without quotes.
  • The Instapaper plugin doesn’t work since RSS feeds for folders are no longer supported. I’m still not switching to Pocket.
  • You will need an IFTTT account to log your Facebook posts, using this recipe.
  • Instructions for logging RunKeeper activity are convoluted, but work.

Slogger’s default time for sucking in your data is 11:50pm, when my laptop is usually in sleep mode. The scheduler should still be smart enough to start the app on wake-up. Nevertheless, it’s one more reason for me to get a used Mac Mini. In 2015, perhaps.

AuthorMiloš Miljković

In the last two months, two of my must-listen podcasts, Systematic and Mac Power User, have had medical professionals on as guests. I don’t usually listen to medical podcasts—Twitter and saved PubMed searches are big enough firehoses—so I thought it would be interesting to hear how my more experienced colleagues use technology. Two of the three episodes were underwhelming, one was stellar.

It started with Brett Terprstra and Dr. Pamela Peeke on Systematic. She has several books targeted towards lay public, and the episode went in the same vein—broad advice on nutrition, well-being, etc. I cringed more than once, but that was to be expected—public health information relies on overplaying the risks and simplifying facts to the point of absurdity[1]. The one thing I could agree with was how important meditation can be, as mindful meditation might decrease physician burnout. Negative points for not mentioning Mindfulness in Plain English as essential reading, though I haven’t read Dr. Peeke’s own recommendation, The Miracle of Mindfulness.

I had higher hopes for Episode 169 of MPU, since Katie Floyd’s and David Sparks’s guest, Dr. Jeffrey Taekman, has an excellent productivity blog. Alas, McSparky spent more than half of the show being fascinated by the minutiae of what doctors do. Which is better than what followed—long periods of uncomfortable silence while the unprepared guest clicked through every app in his menu bar to see if there is anything worth mentioning[2]. There wasn’t.

Then another episode of Systematic came on, with Dr. Don Schaffner, a microbiologist[3]. It was outstanding. Brett was a better interviewer than David, and avoided getting too side-tracked by his guest’s interesting work. But ultimately, the show was good because Dr. Schaffner had useful tips and app recommendations that did not simply regurgitate the latest round of MPU/Mactories/Macdrifter/etc. sponsors. His paper review workflow gave me several ideas I will work on during the holiday downtime. He also suggested a promising contender in my quest to find headphones that will survive more than 8–12 months of intensive use.

One more thing for me to do during the downtime: promote Zotero. Between the developers fumbling Papers 3 and Mendeley being taken over by an evil corporation, Zotero coupled with a few extensions is the best reference manager on any platform. Coming in 2014.

  1. Much like weather forecasts.  ↩

  2. OK, it was not total silence. You could hear Katie fuming in the background.  ↩

  3. PhD, not MD. Wonder if that explains why the show was better.  ↩

AuthorMiloš Miljković
Ten common residency idioms and phrases
  • I don’t feel comfortable doing that.—I don’t know what you’re asking me to do (nurse to intern); I’m too lazy to do it (intern to resident); I think it’s a stupid idea and there’s no way you can make me do it (resident to attending); You’re not paying me enough to do this --- (attending to administration).
  • It’s a light elective—You don’t need to show up.
  • Needs to read more. (on a written evaluation)—I have no idea how much medicine this person knows. I barely know any myself.
  • The family is reasonable.—Family members don’t ask too many questions and will agree with anything you say.
  • The patient has xyz.—I’ve read in an old discharge summary that the patient has xyz, but have no idea how they established the diagnosis, what stage it is in, or what the hell xyz even is.
  • The head is normocephalic, atraumatic. Pupils are equal, round and reactive to light and accommodation. Sclearae are nonicteric.—If I were to report the physical exam I actually did it would take five nanoseconds, so take these fillers to make it seem like I’ve put in some effort.
  • Thank you for the thorough presentation.—Why did you waste my time with all that useless information?
  • That’s an outpatient work-up.—Administration is already breathing down my neck because of this patient’s length of stay and you’re worried about a mild anemia and a positive hemoccult!?
  • That’s her new baseline.—Her disease is worse and we don’t know why, so I guess she’s stuck with it.
  • Please let me know if you have any more questions.—This is the end of our conversation, so please stop talking. I shall now leave.
AuthorMiloš Miljković

…to me, this looks flaky. Yes, Scott Adams (of-Dilbert-fame) is right in saying your best bet for success in life is being pretty good in several skills rather than trying to be the best ever in only one. So, a plan like this:

  1. Step one: become a decent entrepreneur
  2. Step two: become a decent MD
  3. Step three: ???
  4. Step four: profit!

might indeed be a good idea. However:

  • Medicine implies altruism. Entrepreneurship implies greed.
  • Programs want their residents to be 100% dedicated to medicine in general and the program in particular. Can you do that with a small business on the side?
  • Physicians in academia, i.e. those who conduct residency interviews, forgo 300k+ salaries so they could dedicate themselves to research and education. Are you sure telling them about your latest money-making scheme is a good idea?
  • As a resident, do you look at each patient as an opportunity to help them and learn from them, or to figure out how to build a business around them?

Residency programs exist to train physicians, not CEOs. Residency slots are already in short supply. Would program directors give a position to someone who is more likely to end up not practicing medicine at all?

AuthorMiloš Miljković

After 18 months of intensive use, here are some of the apps left standing on my iPad 3, sorted by category. I like to think I’m a semi-advanced user, so for some of them I have also listed simpler alternatives. It goes without saying that you should download all the free iWork and iLife apps.



  • For: all medicine residents
  • Recommendation: strong
  • Price: free (if you bought online MKSAP access)

MKSAP question bank. No-brainer if you are studying for your internal medicine board or MOC exam. Less page-flipping and instant gratification. Unfortunately, it doesn’t allow you to highlighting or annotate the explanations. Also, it can’t make custom quizzes, can’t review unanswered/wrong questions, and doesn’t allow you to copy any of the text to your notes. Lot’s of cants, but it’s the only MKSAP app available. Free if you purchase the electronic version of MKSAP 16.

Download MKSAP 16 from the app store here.

ACP Guidelines

  • For: all interns
  • Recommendation: ok, sort of
  • Price: free

It seems like a good idea, and the content is great, but it is more of a branded PDF reader than anything else. Doesn’t have search or favorites, and you have to download each recommendation one by one. The download is fast, but good luck getting what you need without internet access. So, good for night-time reading, particularly if you’re an intern, but not a good POC tool.

Download ACP Guidelines from the app store here.

Stanford 25

  • For: everyone
  • Recommendation: just OK
  • Price: free

If you haven’t heard of Stanford 25 before, see this TED talk and see the blog. It’s another good, if ugly, night table app.

Download Stanford 25 from the app store here.


Things for iPad (or Omnifocus)

  • For: everyone
  • Recommendation: strong
  • Price: $19.99 for Things, $39.99 for OmniFocus

Whether you’re a GTD fan or not, this or it’s more powerful and more expensive sibling OmniFocus are a must-have for anyone shuffling between more than two areas of responsibility. It still hasn’t been updated for iOS 7, but is very functional. Only two missing features for me, really: there are no nested tasks/dependencies, and you can’t filter by more than one tag.

I’ve been thinking about switching to OmniFocus, but this works well enough for me that the hassle of complete overhauling my system wouldn’t be worth it. Not to mention the >100$ price tag.

Download Things for iPad from the app store here. You can find OmniFocus for iPad here.


  • For: everyone who gets more than 5 emails/day
  • Recommendation: strong
  • Price: $0.99

The best mail client on the iPad. Apple’s was OK until I realized I spent way too much time scrolling through my list of 20 IMAP folders whenever I wanted to move an email. Boxer works with Gmail, IMAP and Exchange accounts, has smart email sorting, and integrates with Sanebox.

Download Boxer here


  • For: everyone
  • Recommendation: strong
  • Price: free

If you use Dropbox on your PC—and you must—then this is a no-brainer.

Download Dropbox for iOS from the app store here

iThoughts HD

  • For: nerds
  • Recommendation: OK
  • Price: $9.99 for either

I found Tony Buzan’s book on mind mapping as a first-year medical student and used the hell out of it for my biophysics, chemistry and genetics coursers. As the material got more complicated, shuffling huge stacks of A3 paper became unwieldy, so I went back to plain old Cornell notes for biochemistry et al. This app is what got me back to making maps, this time when writing review articles and planning out other research. Also good when contemplating the GTD 50,000 ft view.

Download iThoughts HD for iPad from the app store here. It’s prettier new cousin Mindnode 3 is available here

Calendars 5

  • For: all busy overachievers
  • Recommendation: ok
  • Price: $6.99

The default calendar used to be ugly and impractical. With iOS 7 it’s just the latter. This is a good replacement. Fantastical for iPad would be nice, though.

Download Calendars 5 from the app store here.


  • For: advanced users
  • Recommendation: strong
  • Price: $3.99

Quick note-taking and automation rolled into one. I use it as the default inbox for anything and everything, mainly by appending a dump.txt file in my Dropbox. There is a separate iPhone version that is just as useful.

Download Drafts for iPad from the app store here.


  • For: pack-rats
  • Recommendation: ok
  • Price: $1.99

Pinboard is an excellent almost-free bookmarking and discovery service. There are plenty of iPad clients available, but Pinner seemed to be the most cost-effective. I haven’t regretted the purchase.

Download Pinner from the app store here.

GW Mail

  • For: anyone who is forced to use GroupWise
  • Recommendation: meh
  • Price: $9.99

I have to use GroupWise email for work. This is the only decent client I found for iOS. Stopped looking for a replacement since my last day of residency is less than six months away.

Download GW Mail here


Reeder 2

  • For: serious feed readers
  • Recommendation: strong
  • Price: $4.99

I’ve been using RSS feeds since the days of Bloglines (circa 2001) and switched to Google reader after the first big redesign. It’s sad that Google decided to murder it instead developing its potential as a social service. Feed wrangler is a good replacement, Feedly is a free one. Reeder 2 is the best iPad feed reader there is, and works well with both.

Download Reeder for iPad from the app store here.


  • For: everyone who reads
  • Recommendation: strong
  • Price: $3.99

If you read any text that’s longer than 500 words with any regularity, you need a service that will keep track of the articles and remove all the annoying cruft surrounding the text. Instapaper is the first one of its kind, and the best way to read articles on it is on an iPad.

Download Instapaper for iOS from the app store here.


  • For: everyone
  • Recommendation: strong
  • Price: free

Ten good articles hand-picked by an expert hand-picker and delivered (almost) every weekday. My only source of news for the past six months.

Download NextDraft from the app store here


  • For: dabblers
  • Recommendation: ok
  • Price: $4.99

The second book from Tony Buzan that I read was on speed reading. This app will flash words from any article you find online or in your Instapaper/Pocket queue one-by-one at a set rate. Good for those who are too lazy to swipe.

Download ReadQuick from the app store here


Day One

  • For: everyone
  • Recommendation: strong
  • Price: $4.99

A journaling app. I don’t use it for the Dear-Diary types of texts—though I have no doubt it would be perfect for that. Instead, I use it to keep an archive of meeting and lecture notes (usually started in Drafts and sent to Day One), with an occasional milestone in between. Feature request: multiple journals.

Download Day One for iOS from the app store here


  • For: beginner iPad writers
  • Recommendation: ok
  • Price: $4.99

If you want to write a long text on an iPad and don’t need automation, text expansion et al. then this is the app for you.

Download Byword for iOS from the app store here


  • For: advanced users
  • Recommendation: strong
  • Price: $4.99

If you want to write a long text on an iPad and like mucking about with workflows, text snippets and Python scripts—which I most certainly do—this is your only choice on any platform. This will become essential next July when I start my long commute.

Download Editorial from the app store here


Twitterific 5 or Tweetbot

  • For: everyone
  • Recommendation: meh… you might want to wait for the newest version of Tweetbot to come out
  • Price: $2.99 for Twitterrific, $2.99 for Tweetbot

If you are on Twitter—and if you are a physician you really should be—please get a decent iOS client. The official one is definitely not it. Tweetbot used to be until iOS 7 came and made it look and feel ancient. Twitterrific is a good—if slightly annoying—substitute, with the added benefit of being universal (i.e. iPhone and iPad with the same purchase). I’m using the old version of Tweetbot and waiting for the new one, since Twitterific tended to make a mess of my position in the stream.

Download Twitterrific 5 here and Tweetbot for iPad from the app store here


  • For: everyone who uses Facebook (why?)
  • Recommendation: my wife likes it
  • Price: free

OK, I guess, if you’re into that sort of thing.

Download Facebook for iOS from the app store here


  • For: everyone away from family
  • Recommendation: OK
  • Price: free

This is the international default for long-distance communication, I guess. It gets choppy and drains the battery, but it’s the only thing my mom knows how to use so I’m stuck with it.

Download Skype for iPad from the app store here



  • For: everyone who can spell
  • Recommendation: strong
  • Price: free (with in-app purchase)

An excellent turn-based word game. The only multiplayer game I play with any regularity. You need an in-app purchase if you want to play more than two games at the same time, but it’s well worth it. I have five going on right now.

Download Letterpress from the app store here


  • For: nerds
  • Recommendation: strong
  • Price: $1.99

Bejeweled meets a 2D RPG. Hours of fun, even when you get to 100000000000 or however many points.

Download 100000000 from the app store here


  • For: adventure gamers
  • Recommendation: strong
  • Price: $4.99

A 2D side-scrolling action-adventure game set under the sea. At my pace I will finish it in about two years, but it’s great even in 15-minute increments.

Download Aquaria for iOS from the app store here



  • For: serious shoppers
  • Recommendation: strong
  • Price: $4.99

Forward an email containing a tracking number to a special email address. Boom, you can now track your package through this app, with push notifications if you’re into being interrupted whenever a case of -diapers- Wild Turkey is delivered to your front door.

Download Deliveries for iOS from the app store here


  • For: serious eaters
  • Recommendation: OK
  • Price: free

Good app for ordering food in the Baltimore area. Don’t know about rest of the country.

Download Eat24 for iPad from the app store here


  • For: world travelers
  • Recommendation: strong
  • Price: free (you pay for the plane ticket, though)

The best flight comparison engine there is. Find the most affordable and least annoying plane route. Also does hotel rooms, which I haven’t tried.

Download Hipmunk for iOS from the app store here



  • For: everyone with a Netflix subscription
  • Recommendation: OK
  • Price: free

I have used this app exactly once, to watch a couple of episodes of Buffy the Vampire Slayer while waiting for an Amtrak train. Well worth it, though.

Download Netflix for iOS from the app store here

AppleTV Remote

  • For: everyone with an Apple TV
  • Recommendation: strong
  • Price: free

I don’t have my original remote any more. We assume Dora ate it. This app is even better, since you don’t have to muck around with the tiny remote buttons when entering your wifi password or searching Netflix.

Download AppleTV Remote from the app store here


  • For: everyone who reads comics
  • Recommendation: strong
  • Price: free (the app, not the comics)

The only way to read comics on an iPad.

Download comiXology from the app store here

AuthorMiloš Miljković

From people at Press Ganey:

Reliably providing evidence-based clinical care is essential to reduce patients’ suffering — but it is not the only way.  Indeed, excellent clinical care is “necessary but not sufficient”. As discussed, care givers must also build trust and relieve anxiety.

Skeptics may wonder if qualitative improvement in the control of anxiety, confusion, and fear is possible. In fact, such improvement is already well underway, as demonstrated by patient experience data collected from patients receiving care from hospitals, ambulatory groups, and other providers.  This progress seems to be driven in particular by improvement in nurse communication, pain control, and care coordination.  Nevertheless, these data also demonstrate marked variability among providers in these measures, and opportunities for improvement for all.

Don't want to throw the baby with the bathwater, but the whole area of patient satisfaction scores being tied to reimbursement is one very filthy bathwater that has a lot to do with the recent opioid abuse epidemic.

AuthorMiloš Miljković
Using only one device isn’t liberating; it’s just the opposite. It means putting yourself through unnecessary discomfort and friction, even though a better option is available.


AuthorMiloš Miljković
The total number of applicants to medical school grew by 6.1 percent to 48,014, surpassing the previous record set in 1996 by 1,049 students. First-time applicants, another important indicator of interest in medicine, increased by 5.8 percent to 35,727. The number of students enrolled in their first year of medical school exceeded 20,000 for the first time (20,055), a 2.8 percent increase over 2012.

Nice, but two questions come to mind: how many of these were students who would have gone to Caribbean schools anyway; and if some of them are indeed brand new future doctors, where will they get their specialty training in four years? CMS doesn't seem likely to open up any more residency spots.

Also, FMGs are slowly getting pushed out. For better or worse. 

AuthorMiloš Miljković

They let us peek into the sausage factory last week.

Nominally, the lecture was about RVUs1. An accountant type in a pinstripe suit explained why the government came up with the concept and how more RVUs translate to mo' money for the hospital. Then he showed us a table. This is how much RVUs an average ophthalmologist makes in an hour. Here is an orthopedic surgeon. See here at the bottom? That's an internist. This is how much you're worth to us, scum2.

Then there was a chart. This is the last fiscal year. This solid line here are monthly RVUs for an average hospitalist. The dotted line is for a single physician in the practice. See how it's always above the solid line? That's good. We love that person.

We had medical students and interns just three months into training listen to this. It was blood-curdling.

Not because the hospital organized the lecture, mind you. It is a very good thing they did it, and it is good for doctors in training to realize as early as possible in what kind of a healtcare system they are expected to work. What is frightening is that there needs to be an entity, let's call it administration, which views the hospital as a production plant and physicians as line workers who need to maximize outputs, optimize efficiencies and do other newspeak claptrap.

Administration usually lies—appropriately—on the ground floor, far removed from that other sausage factory of actual patient care. It looks at pie charts and histograms and RVU tables and keeps coming up with new and exciting ways to increase production while wondering why those bumbling doctors at the bottom of the list can't do whatever the top performing docs are doing to keep the hospital in the black.

It's modern medicine, it's complex, it's expensive, it requires that level of organization and detachment—you might be tempted to say. Yes, you could indeed say that, if not for the lonely example of every other country in the developed world which does it differently than the US.

But never mind that. With all the shenanigans the Congress has been up to this week, that end of the equation is unlikely to change. What administrators should do—and I understand the banality of the following advice—is see real physicians interacting with real patients for at least and hour each week. Interns being bombarded by page after page—from critical to comical—while trying to figure out a 15-minute window to eat, get coffee and use the restroom3. Residents finishing a 24+ hour ICU shift that started with three codes and ended with a difficult end-of-life care discussion, with central lines placements and intern supervision—but no sleep—sprinkled in between. Attendings getting yelled at while trying to explain to family members why they need to pay for the medications out of pocket or bring their own4.

One hour. Each week. Mandatory. To put things in perspective.

  1. Relative Value Units

  2. Not his actual words. Actually, he sounded very apologetic when explaining it. Still stung though. 

  3. At the same time? — a thought will come to them, to be quickly dismissed. 

  4. Hello, observation status. 

AuthorMiloš Miljković

Electronic patient notes, the way they exist now, are dangerous. As physicians wiser and more experienced than myself have noted, they are made for billing, not story-telling and communication between healthcare professionals; and as anyone with even basic literacy in the English language will notice as soon as they read one, they are a barely comprehensible, intelligible, muddled word salad that looks computer generated because, well, in most cases it is.


For one, they are ridiculously easy to create. Click on a checkbox and every admission note you start will come pre-populated with what the EMR thinks are the patient's current home medications, prior surgical procedures and such. Have trouble accurately documenting the dozen medications your 72-year-old with systolic heart failure, diabetes, CKD and vascular dementia has? No big deal—the e-patient has at least something listed from an ER visit 9 months ago. You'll make sure to go back to the admission note later and append it with the correct list when you get it from the family member, right? Right.

They also save you from having to type. Click click click, and the review of systems is done. Too much clicking? There is a solution: spend 5 minutes to create a macro, and you will have all your common questions pre-answered as No on all the notes, shaving off seconds of additional clicking. Because asking all your patients the same questions and expecting identical answers is just plain common sense, amiright? Oh, and of course tachycardia is a symptom. It's right there on the ROS list, waiting to be clicked.

Most of all, electronic notes are the one cure for writer's block. While in the distant past1 you had to spend agonizing minutes staring at a blank admission note trying to form a coherent story on why the patient came to be seen, and then had to put it down on paper without feelings of guilt and shame, you learn from EMR that it is OK to sign a medico-legal document that contains this brilliant turn of phrase:

The reason for visit is: pt missed hd, high bp, n/v. The course was: constant. The exacerbating factor was: none. The alleviating factor was: none.

But why, when medicine residents are all moderately-to-ridiculosuly smart and ambitious people who should know better?

Well, some of them actually don't know any better. Even in the olden days1 you had a couple of interns who were not the best ever history-takers2 and wrote poor-quality3 notes. Electronic notes, unfortunately, help them obfuscate their deficiencies. It is very easy to see in a one-page note how much useful information the resident has actually obtained. Not so much with computer-generated six-pagers.

Then there is your typical smart intern who is just finishing putting in orders for her fourth admission that day, after discussing each one with the supervising resident, all while answering a barrage of pages about the 30 patients she is cross-covering. The first two admission notes are almost done—she has to change the plan after talking with the resident—but the other two will have to wait until she updates the sign-out and hands off all the patients to the night float. This is arguably much more important than notes, as it directly affects the care those newly admitted patients will get overnight. The admission note is not really needed until the following day, during morning rounds. She's smart enough to prioritize.

She's also smart enough to know what is expected of her. Most of what she knows about writing admission notes she learned from her peers, particularly seniors—who concentrated on efficiency —and that lady at the billing department who gave a noon conference talk on the importance of complete documentation for coding. So The Man wants me to be efficient-yet-thorough, and then he gives me this electronic tool with auto-population, templates, macros and such. Hmmmmm… OK!

Yes, she might get in trouble if her notes are so horrendously bad to significantly impede patient care. From my very limited experience this just does not happen. Or rather, if it does, appropriate documentation is a single bullet in the long list of areas of improvement during an M&M.

What to do?

Getting a queue from, let's say, an attending might help. Or rather it would, if not for the exceeding rarity of attendings who can lead by example. They do know what is supposed to be done, but with time constraints and a similar set of competing incentives that the residents have, they just cannot get around to it. Decades ago, we had physicians who knew what a good note looked like and always managed to write one. Now, the attendings usually know what they should do but are prevented from actually doing it. Slowly, we are slipping into a generation of doctors who know less and less about good documentation, never mind actually executing it.

Is this a bad thing? Of course! Not because of the notes themselves—they are just another victim of technology's war against spelling, punctuation and grammar. It's the skills you need to create one, like getting a clear and concise history and formulating a coherent plan, and what the purpose of the note is—coordination of care between all the interested parties4—that are taking the hit. Until we find a better way of doing this, quality of notes should be of concern.

A better way of doing things

A thought popped into my head.

We need something capable of taking a good history, physical exam, assessment and plan as an input, integrating it into a readable and eaily understandable format that can be accessed by every member of the healthcare team, and the patient.

It should also be easy to enter and change information without having to duplicate work that somebody (PCP, ER attending, the previous team) had already done. Knowing who entered what and when would also help.

Sometimes—often—different physicians disagree on the diagnosis or treatment. There should be a way to mark the controversial parts of the record in a way that would shall all sides of the story.

It should contain both static or rarely changed facts about the patient, like family and social history and, to an extent, the past medical history, as well as the ongoing narrative about his or her health, like ER and PCP visits.

Of course, that something would stem from, not be hampered by, technology.

So, how about a wiki?

Just a thought.

  1. Or in my program, six months ago.
  2. They would be the ones calling the patient “a poor historian”, and were usually correct, although not in the way they intended. Patients are the ones giving a (hi)story, the physician is the historian.
  3. OK, I'll call them what they are: crappy.
  4. Not billing

AuthorMiloš Miljković

After a couple of months on the floors, the new interns are learning the value of the bazillion pieces of advice that pop up around July 1st—all good, by the way. Here is what helped me (and if you don't have time to read through the rest of the post, here it is in nine words: Set your priority. Be deliberately slow. Know the Systems.).


Don't worry about life after residency. Learn how to be a great internist. If you already have a subspecialty in mind, great, less time spent soul-searching. But working on all those CV-buffing research projects is a lot easier when you become competent enough on the floors to be able to do a good job quickly1 and not let patient care—or your interns!—suffer while you are going through your 217th chart for a retrospective analysis.

To become great at doing any job, you need practice. Cutting oneself off from the outside world and letting friends and family know not to expect much in the being-there department for the first few months of internship can work wonders2. Do not be lulled into Breaking Bad marathons by the new work-hour regulations. For pleasure, read a good book or three.

This may sound very harsh to the sensitive ears of our touchy-feely generation. It is also the truth. Twenty-first century physicians may not need Navy Seal or firefighter-class dedication any more, but the profession is still up there in the Demanding category for those who want to be good at it3.

Medium scale

OK, so Work is now your top priority4. You will soon find out, if you haven't already, that work itself has at least three competing sub-priorities: patient care, your own learning and improvement, and interests of the hospital5, only one of which can be the actual priority in any given situation. Most good residency program would rank these in the order listed, which is nice and proper, and makes everyone feel warm and fuzzy when they say it. Patients come first!

As a new intern, however, you don't have enough knowledge, skill or experience to provide safe and effective patient care. The best thing you can do for your patients is to reach a level of basic competence as quickly as possible. The supervising senior resident and attending should be there to ensure appropriate care. You will pay them back by being a good senior resident supervising interns of your own next year. Of course, having your own improvement as a priority does not meen doing it exclusively, to the detriment of the other two. But situations will frequently arise when you will have a clear choice: spend an extra 30 minutes dictating The Best Discharge Summary Ever instead of just an adequate one, or read about the Stanford 25; play a game of phone-tag with the patient's primary care doctor to get her hemoglobin A1c from two months ago, or learn what your hospital's policy on inuslin drips is and why. Unfortunately, many dilemmas will be a lot more subtle.

The path to competence is thoroughness6. Yes, good patient care requires efficiency, but there is efficiency for its own sake, which leads to horrible med reconciliations, missed diagnoses and unreadable point-and-click notes, and efficiency through expertise, which is what you should be aiming at. If you put efficient patient care above your own expertise you will be admired by your co-interns and well-liked by your seniors for signing out on time and not making them stay late, but you will be a piss-poor clinician. In an ideal world the admission cap would be three and you would not have to compromise anything. Queue sad music and a line on real life, becoming an adult, etc.


Why go through all those years of med school if they can not prepare you for internship? Because they focus on exactly one of the six ACGME core competencies, namely, medical knowledge7. Out of the remaining five, systems-based practice is the single most high-yield area for any intern—and the most difficult one to grasp. Which is unfortunate, because it is actually quite simple: it's how you get stuff done in the hospital. Getting IR to embolise a bleeding mesenteric vessel at 3am; figuring out which surgical service to call for your cancer patient with an incarcerated hernia; differentiating between a case manager, social worker and a discharge planner8. Learning how to do this early in the intern year—a process that always involves at least two more compentencies, communication skills and practice-based learning, and often many more—will free up enormous amounts of time later on.

This is a particularly large hurdle for foreign medical graduates, but culture varies from program to program and many American grads will find themselves saying “This is not how we did it at XYZ” a lot. Even the seemingly best possible scenario, in which you did your core clerkship and sub-I in the program where you are now an intern, carries a huge risk of going into internship with a false sense of knowing the system and doing things wrongly or inefficiently because of it.

But why bother learning all that if it will only apply to that one hospital? It is very unlikely that you will stay there. Aside from the above mentioned spending a month or two to learn this will free up huge ammounts of time in the next N years, you will likely spend much of your professional life working in different hospitals and other practice settings. Knowing the different variants and learning how to learn the system will be even more valuable then, when your time gets an actual dollar ammount attached to it.

By the time you've figured out the systems, you should also be able to zero in on the next competency you need to concentrate on. Peer and attending evaluations from the first few blocks will help there. Be it practice-based learning or patient care9, just make it a priority to deliberately become better at it. Do not be surprised if medical knowledge never bubbles up to the top of the list during the first year, and you don't get to open that copy of MKSAP gathering dust on your night stand—that is what med school was for.

So there you have it, my 100% fool-proof plan for making your own and the life of your loved ones miserable. Or so it will seem. Intern year has a steep learning curve any way you approach it, but it ends eventually. You might as well make the most of it, and the benefit of close supervision that comes with being an intern.

  1. Good. Not great, not perfect, but good. A great job cannot be done quickly, and prioritizing is an essential skill to master during internship.
  2. Another essential skill to master, for life and career: setting—and lowering—expectations.
  3. It has also never been easier to coast through by being just barely competent enough not to grossly missmanage, but likeable enough to get good HCAHPS scores.
  4. Unless you have children.
  5. You will be hounded by administrators and nurse managers to put the last one first. Ignore them.
  6. Some may derogatively call this being slow. These people are idiots. Being deliberately slow is essential for any novice.
  7. It is pretty depressing to realize that all those things you were stressing over in med school are only one of six things you are evaluated on during residency, and not a very important one at that.
  8. Some may say this is scut work. This mode of thinking is probably a holdover from med school mentality, in which any activity that doesn't directly involve patients, lecture halls or large ammounts of text is considered scut. On the other hand, the best definition of scut I can think of is work that somebody with fewer years of education than you could do equally well. Calling consults is not scut work, it's medicine.
  9. Not to be confused with actual patient care, the competency of patient care are the skills you need to care for your patients, like getting a good history, performing physical exams and keeping track of test results. Some of those skills you improve by performing patient care, some thorugh reading, observing and trying out different things.


AuthorMiloš Miljković

Yes, the original research articles JAMA pushes out any given week are… less then memorable. But they do have pretty good opinion pieces. Alas, both behind a paywall.

Hey, it's Dr. Verghese! Here's how they do it in Stanford:

  1. Use real patients
  2. Observe's residents' practice
  3. Demonstrate clinical skills, in workshops and at the bedside
  4. Critique and develop the evidence

We don't have much choice about no. 1 at Sinai, but are actively working on nos. 2-4. The biggest challenge with direct observation? Staffing.

Table 1, showing 1975 and 2010 incidence and mortality of seven different cancers, is a great teaching tool that we will be using for one of our morning reports. Not to thrilled about calling some cancers IDLE, though.

Bonus piece: This awesome video (link) on rotator cuff disease

Now if they only brought back the old cover page style.

AuthorMiloš Miljković

Danielle Ofri, in a NYT piece on medical errors:

When you were juggling patients with acute heart failure and rampant infections, it was hard to get worked up over a demented nonagenarian who was looking a little more demented.
The trick to surviving was to shuttle patients to another area of the hospital as quickly as possible. This patient was a perfect candidate for the intermediate care unit, a holding station for patients with no active medical issues who were awaiting discharge. 

This part is very true. Then comes the conclusion to the story.

The next afternoon the doctor tracked me down. Without mincing words, she told me that she’d been called overnight by the radiologist; the patient’s head CT showed an intracranial bleed. The patient was now with the neurosurgeons, getting the blood drained from inside her skull.

The rest of the piece is about medical errors, checklists and such, and is perfectly reasonable if a bit dull after everything already written on the subject, especially by Atul Gawande. What's not mentioned is cost-conciousness--surprising considering it's become a buzzword since the ACA. Demented ninety-year-old nursing home residents with headbleeds getting the full neurosurgery monty is by no means the standard of care in those first world countries the US is looking up to when it comes to healthcare costs.

What's often given as a reason is that families want something to be done for closure, even if there is little chance of meaningful recovery. In cases like this, something done should be hospice. Neurosurgery shouldn't even be on the table.

AuthorMiloš Miljković

Dr. Deborah Dowell in a JAMA opinion piece:

Before prescribing opioids, a more useful and important question than a patient's likelihood of dependence is whether benefits of opioids in relieving pain are likely to outweigh the risks of the drugs. For pain control at the end of life, the answer to this question is often yes. If the indication for opioids is chronic noncancer pain, the answer to this question will be no much more often than many physicians may realize. Despite widely held views about the efficacy of opioids for pain control, systematic reviews have not found sufficient evidence that long-term opioid use controls noncancer pain more effectively than other treatments.

No, not while there's an HCAHPS survey with the question "Has your pain been well controlled" and penalties if the patients circle anything other than ALWAYS.

AuthorMiloš Miljković

Apparently, not lower than 10. From "A No-Prophylaxis Platelet-Transfusion Strategy for Hematologic Cancers" in NEJM:

Patients were randomly assigned to receive, or not to receive, prophylactic platelet transfusions when morning platelet counts were less than 10×10^9 per liter.


Patients in the no-prophylaxis group had more days with bleeding and a shorter time to the first bleeding episode than did patients in the prophylaxis group. Platelet use was markedly reduced in the no-prophylaxis group.

Yes, if your strategy is not to give platelets, you will end up giving less platelets. Brilliant. 

AuthorMiloš Miljković

The Guardian asks:

This question of decriminalisation and legalisation does seem critical…

No, it's not. It only seems to be critical to people on the outside of what's happening. Let me put it this way: drugs are legal in Baltimore. Right now there are 60,000 addicts in Baltimore. If they could lock up 100 or 200 of them a day, they would have a record number of arrests. But they can't, that's too many. By the numbers, not by the fucking law, drugs are legal in my city.

 David Simon is awesome.

AuthorMiloš Miljković