Statistics resources for clinicians

Another week, another Quora question.

What is an online resource for learning statistics needed for clinicians explained in a language that could be understood by doctors?

There are many biostatistics courses available on Coursera. Living in Baltimore, I’m biased towards JHU’s offerings. “Case-Based Introduction to Biostatistics” by Dr. Scott Zeger is a good one. If you prefer text to video, here are three good resources:

If I had to pick one, it would be Dr. Brush’s book. He is a cardiologist writing for other physicians in a language they can understand. Also, Dr. Lehman recommended it, which is more than good enough for me.

more ...

BRAF mutations in non-small cell lung cancer

This was the last tumor board of the year, and the first in which I did not discuss a hematological malignancy. Lung cancer is bad, doubly so when it’s metastatic, but looking at this patient’s history you can see it doesn’t necessarily mean months-not-years left. This particular patient is worthy of a case report, and one is indeed in progress.

more ...

Why be a chief resident?

For the first time since joining Quora, I found a question to which I can meaningfuly contribute. Thought you might like to see my answer.

Why would someone choose to be a chief resident (in internal medicine)?

Why indeed.

The cynical answers would be “out of a misguided sense of loyalty to your program”. The correct and not very useful answer is—it depends.

Most positions entail primarily administrative responsibilities, with some teaching and clinical duties, and a salary just slightly higher than that of a PGY-3. So, you can expect your patient care skills to languish unless you work on maintaining them, your teaching skills to be slightly improved—or at least no worse if you’ve had some prior experience—and your knowledge of hospital administration, people management, dealing with email, and making the most out of seemingly pointless meetings to go through the roof. If you have any interest in academic medicine, as a generalist and sub-specialist alike, this last skill set will be invaluable. It is also a stamp of approval of sorts for any fellowship program director looking at your CV if and when you apply.

You also have much more free time. Depending on how many chiefs your program has, it will be most or all weekends, and almost all federal holidays. This is a good time to study for the boards if you haven’t taken them already, write up the research you’ve been working on, or spend some time with your family (the chief’s maternity/paternity leave is usually more flexible, but that’s program-dependent).

The downsides: you will have one fewer year of attending-level salary, so if you have a large debt or other financial responsibilities think twice before saying yes; some friendships you made with the junior residents will be undone or temporarily put on hold, unless you are very careful about not playing favorites; you may lose some respect for your higher-ups, as it goes whenever you peek behind the curtain; you will need to develop a thick skin, if you don’t have one already. Some would say these last two are actually pluses. It depends.

Visa issues complicate the matter, but I won’t go into details—bureaucracy shouldn’t play a role in determining a career choice, and when there is will (your own as well as the program’s) there is a way to bypass any obstacles.

Hope this helps.

more ...

A few good pens

Helping out Georgetown University fellows with their oncology consult service for a few days reminded me of how important it was to have at least three good pens with you at all times. By that, I don’t mean grabbing a handful of disposable Bics from a Staples shelf just because you know you will lose a lot—though you should certainly plan for theft and absent-mindedness. There are at least three different use cases you will face daily, and no pen will be ideal for all of them. I’ve tried out enough, and listened to plenty of podcasts on the topic, to be happy with my choices.

Role 1: The task keeper

Intern or fellow, this will be your most-used pen—the one you pull out to quickly jot and check off to-dos, make note of pertinent lab values and vital signs, and write down other important bits of information. If you are anything like me or my colleagues, you will be doing this on the signout, or some other piece of paper packed with information you might need.

This is why your pen should be:

  • as thin as possible, since you’ll likely write on the margins, but
  • not too faint, so it wouldn’t blend in with the rest of the text,
  • quick to use, since you’ll have to pull it out during rounds, patient encounters, and other situations in which fiddling with the cap would lead to losing both time and the cap, and
  • affordable, since you will misplace one every couple of months.

For the first two, the Uni Jetstream wins hands-down. The tip is as smooth as a 0.5 mm can be. It doesn’t skip, spill, sploch, or splat. The price is right—just $2.99 on It is the best-in-class for every thing save one.

Zebra Sarasa Push Clip is not too thin, but thin enough to be scratchy and slightly annoying. Even with that small flaw, I choose it before the Jetstream. Because of the clip. The wonderful, magical clip.

You see, after four years of rounding, the act of pulling out a pen becomes a reflex. You hear something important, you have a thought, you blink, and you have a pen in one hand and your Very Important Paper in the other. You write something down, you blink, and your hands are free again. You are one with the sign out, and the pen.

To do that, you must at all times know where those two things are. The Very Important Paper is hard to miss, but the pen needs to be not just in the same pocket, but in the same spot in your pocket at all times. For me, the whitecoat-less fellow, that’s the inside of my left front pants pocket. This requirement rules out any clipless pens—goodbye, disposable Bics—but the regular clips don’t fare too well either. Too tight a clip, and you spend too much time fiddling it into the spot you want. Too loose, and it’s easy to put in, and easy to lose.

Which is why the clip is magical. You open it wide on entrance, and clamp it shut once you have the pen where you want it. It won’t budge after hours of walking up and down the hospital stairs. And, unlike one of my Jetstreams, it will be very difficult to break.

Alas, it only comes in one color. This is enough for the mild-to-moderate inpatient workload of a fellow, but during internship I needed the typical gunner pen to stay organized. Zebra Sarasa 3 is the high-end guner pen—one color fewer, but with the Zebra clip. For me today, it is just too bulky, I default to black anyway, and I’d just get annoyed with it running out way before the other two. But for me four year ago, it would have been perfect.

Role 2: The note writer

My choice: Ohto Graphic Liner Needle Point Drawing Pen - 02, 0.5 mm black Runner-up: Pilot Petit1 Mini Fountain Pen - Fine Nib, any colour

As much as I appreciate the ammount of writing I can cram onto a sign out with a thin pen, using one to write in paper charts, or for making notes on an old H&P while seeing a patient, creates an unreadable mess. The faint black lines of your pen blend in with the small type and the gray ruled lines of the progress note. Also, you don’t need as quick an access—a morning note-writing session may seem hectic, but you are the one who initiates the process. A couple of seconds looking for the pen or opening the cap won’t make you lose any information.

The Ohto liner leaves a consistent, dark line, lasts for ages, and is the right size for me. The ink is water-proof and archival safe—which is what you want for a medico-legal document. There’s a cap you need to worry about, but I’ve yet to lose one. And at $2.50 it won’t be the end of the world if I do.

This role can also be filled by a nice fountain pen—and you’ll see some attendings using one. I have had horrible luck finding a fountain pen that I won’t be afraid spilling in my bag or pocket, and most cost too much to carry around the hospital while sleep-deprived. The Pilot Petit1 pens have the right price, nice nib, and are easy enough to use. But I’m still too scared to put it in my pocket.

Role 3: The backup

My choice: Uni-ball Signo DX UM-151 Gel Ink Pen

This is the one you sprinkle around the house to be there just in case. The one you give out to friends and colleagues. And the one you use if you lose any of the others. If you needed to have just one pen, this would be the one, since it’s both thin and dark. Not perfect for either notes or task lists, but good enough.

At slightly less than $2 per pen it is afordable enough, though if you just want something to give out to others—and don’t care about them or their fingers—you can get a box of 60 horrible little stick for the price of three Signos. I guess you can give those out to your enemies and watch them writhe in pain and frustration.

Bonus: A pencil case

My choice: Kokuyo Will Stationery Actic Mini Pencil Case

This is entirely optional, but it saved me a surprising amount of time. It comfortably fits 4-5 pens and refills, and has good build quality. If your bag is small or you don’t mind fishing around for the pen you want, you can certainly do without it.

more ...

How to say “I don’t know” like an intern

A key skill to have during oral exams back in med school was never to admit not knowing. Avoid the areas you’re uncertain of, dodge the examiner’s field of expertise as much as you can, and never ever say “I don’t know”.

These sage words were passed on from generation to generation, propagated by everyone, including me. Only, this wasn’t what I or any of my friends actually thought. It was a poke at the climate of intellectual dishonesty at our school, not a guide to success in medicine.

Starting residency, though, flips the sarcasm switch somewhere and the funny guidelines become instructions to be followed verbatim. The knowledge in question is different—patient data instead of textbook medicine—but the idea is the same. Observe the modern American intern’s vocabulary:

  • Not that I know of (means I don’t know).
  • I wasn’t aware of that (means I didn’t know).
  • I don’t think it is (means I don’t know if it is).
  • I belive so (means I have no idea, but yeah, maybe).
  • It probably was (means I don’t have a clue but I did a D6 roll in my head and it was a 5).

I used all of the above, and more, during internship, but still get frustrated hearing it from others1. If you are an intern, or anyone reporting patient data to a person above you in the pecking order, try using “I don’t know, but I can find out in a second” instead. Then start practicing your EMR skills to trully make it a second.

  1. That makes me a liar and a hippocrite, yes, but at least I’m being honest about it. 

more ...

Lymphoma, a post mortem

Six weeks of anyhing will get tiresome near the end, doubly so if it involves taking a 6am train and not getting back home until 7. Combined inpatient/outpatient juggling twice a week didn’t help either, and neither did the high turnover1. Still, as inpatient rotations go, this one was reasonably busy—enough for me to learn quite a bit, but never overwhelming. More importantly, the attendings were awesome, the cases interesting, co-workers friendly, etc.

As the slide set I’ve linked to above shows, lymphoma are tricky in that many are either curable or indolent enough to be kept at bay until something else gets to the patient; but when it’s bad—and with current technology we can’t with certainty say which ones those are at diagnosis, though we’re getting there—it can be catastrophic. Thankfully, those cases are rare.

Though most chemotherapy I was ordering was EPOCH-R or, for NCI’s new primary CNS lymphoma protocol, TEDDI-R, there were several patients with non-hematological malignancies getting IL-15. If you hadn’t heard of this particular interleukin being used for treatment before, it’s becuse the results of the first-in-humans phase I trial done by NCI have just been published in JCO. The maximal tolerated dose was somewhat low when administered as an intravenous bolus, so the patients are now getting it either as a 12-day continuous infusion, or a subcutaneous injection. I’d rather not steal Dr. Conlon’s thunder—or break any government regulations—by writing how these patients did. Just keep an eye for a follow-up paper. And maybe a tumor board presentation from me in the next few months.

  1. I am sure a medical transcriber somewhere is cursing under his or her breath every time they hear my nasal baritone—the last time I had dictated so many discharge summaries was during my intern year. 

more ...

Myeloid sarcoma, a slide show

This tumor board was almost two month ago. The case is fascinating, but the presentation had too much patient-related information for me to be comfortable posting it here. I ended up removing most of the interesting slides, though it remains a decent introduction to GATA2 mutations/MonoMAC syndrome, and a nice overview of chloroma/granulomatous/myeloid sacroma.

more ...

On medical euphemisms

Observe George Carlin discussing how euphemisms are invading the English language:

I first heard a version of this years ago, back in Serbia, while I was still a med student. It hadn’t left much of an impression, but I can imagine myself nodding my head and thinking ha ha, yes, stupid Americans, ruining their own language, or something comparably obnoxious.

Well, I’ve, erm, matured since then. True, some euphemisms now inspire rage instead of vague amusement, like my two favorites:

  • I just wanted to let you know” instead of “I’m telling you”, and its relatives “Please let me know”, and “Thank you for letting me know”. Physicians are particularly fond of this, for we are the gatekeepers of knowledge, and the only reason you know something is because we are letting you. Don’t worry though, it’s not just you, we say that to each other all the time.

  • I don’t feel comfortable doing xyz” instead of “I don’t want to do xyz”, as mentioned here.

Most of them, though—particularly ones we use with patients—have a good reason to exist. The Radiolab segment which inspired this post made fun of “making someone comfortable” being used for dying ICU patients. Instead of… what, exactly? Euthanasia? There is a difference between giving someone drugs usualy meant for comfort—opioids, primarily—in order to kill them, and giving them opioids for pain and comfort knowing it may shorten their life.

Then there are turns of phrase used because they are euphemisms. “You should get your affairs in order”, “your time is becoming limited”, “at this point we should concentrate on quality of life, not quantity” are all ways of saying “I don’t know when you’ll die, but it will be soon, so start planning the funeral”. I am sure Mr. Carlin would appreciate getting it straight, but not every patient is as stoic. We can easily be more blunt if asked to do so, but you cannot un-hit a patient with a sledgehammer like that. So the default is to err on the side of softness.

Then again, most of the euphemisms we use with patients also make us more comfortable with the sitation. What I wrote above may then just be my rationalizing it away with a convenient it’s-best-for-the-patient mantra. In truth—to use another common phrase—euphemisitis is a multifactorial condition (as in, I have no idea what the reasons are, but it’s probably a little bit of everything).

more ...

T-cell/histiocyte-rich large B-cell lymphoma

Another month, another tumor board talk. The slides aren’t self-sufficient, I’m afraid, but the references might be a good starting point for learning more about TCHRLBCL[^wtf], or diffuse large B-cell lymphoma in general. I’d start with [this excellent review article on DLBCL treatment][ccr], written by several NCI attendings[^attnd].

more ...

History of advance care planning in the US

Here is a slide show I had to make during my [pain and palliative rotation][ppc], which ended up being an updated rehash of [a review article I wrote for Cancer Journal on the history of end-of-life care][eolreview]. The article itself is behind a paywall, but you can find the PDF on [ResearchGate][rgate].

more ...