A discussion of heroic responsibility

[Origi­nally posted to my per­sonal blog, re­posted here with ed­its.]

Introduction

You could call it heroic re­spon­si­bil­ity, maybe,” Harry Pot­ter said. “Not like the usual sort. It means that what­ever hap­pens, no mat­ter what, it’s always your fault. Even if you tell Pro­fes­sor McGon­a­gall, she’s not re­spon­si­ble for what hap­pens, you are. Fol­low­ing the school rules isn’t an ex­cuse, some­one else be­ing in charge isn’t an ex­cuse, even try­ing your best isn’t an ex­cuse. There just aren’t any ex­cuses, you’ve got to get the job done no mat­ter what.” Harry’s face tight­ened. “That’s why I say you’re not think­ing re­spon­si­bly, Hermione. Think­ing that your job is done when you tell Pro­fes­sor McGon­a­gall—that isn’t hero­ine think­ing. Like Han­nah be­ing beat up is okay then, be­cause it isn’t your fault any­more. Be­ing a hero­ine means your job isn’t finished un­til you’ve done what­ever it takes to pro­tect the other girls, per­ma­nently.” In Harry’s voice was a touch of the steel he had ac­quired since the day Fawkes had been on his shoulder. “You can’t think as if just fol­low­ing the rules means you’ve done your duty. –HPMOR, chap­ter 75.
I like this con­cept. It coun­ters a par­tic­u­lar, com­mon, harm­ful failure mode, and that it’s an amaz­ingly use­ful thing for a lot of peo­ple to hear. I even think it was a use­ful thing for me to hear a year ago.
But… I’m not sure about this yet, and my thoughts about it are prob­a­bly con­fused, but I think that there’s a ver­sion of Heroic Re­spon­si­bil­ity that you can get from read­ing this de­scrip­tion, that’s maybe even the de­fault out­come of read­ing this de­scrip­tion, that’s also a harm­ful failure mode.

Some­thing Impossible

A wrong way to think about heroic responsibility
I dealt with a situ­a­tion at work a while back–May 2014 ac­cord­ing to my jour­nal. I had a pa­tient for five con­sec­u­tive days, and each day his con­di­tion was a lit­tle bit worse. Every day, I reg­istered with the staff doc­tor my feel­ing that the cur­rent treat­ment was Not Work­ing, and that maybe we ought to try some­thing else. There were lots of com­pli­cated med­i­cal rea­sons why his de­ci­sions were con­strained, and why ‘let’s wait and see’ was maybe the best de­ci­sion, statis­ti­cally speak­ing–that in a ma­jor­ity of pos­si­ble wor­lds, wait­ing it out would lead to bet­ter out­comes than one of the po­ten­tial more ag­gres­sive treat­ments, which came with side effects. And he wasn’t ac­tu­ally ig­nor­ing me; he would listen pa­tiently to all my con­cerns. Nev­er­the­less, he wasn’t the one watch­ing the guy writhe around in bed, un­com­fortable and deliri­ous, for twelve hours ev­ery day, and I felt ig­nored, and I was pretty frus­trated.
On day three or four, I was listen­ing to Ray’s Sols­tice album on my break, and the song ‘Some­thing Im­pos­si­ble’ came up.
Bold at­tempts aren’t enough, roads can’t be paved with in­ten­tions...
You prob­a­bly don’t even got what it takes,
But you bet­ter try any­way, for ev­ery­one’s sake
And you won’t find the an­swer un­til you es­cape from the
Labyrinth of your con­ven­tions.
Its time to just shut up, and do the im­pos­si­ble.
Can’t walk away...
Gotta break off those shack­les, and shake off those chains
Gotta make some­thing im­pos­si­ble hap­pen to­day…
It hit me like a load of bricks–this whole thing was stupid and ra­tio­nal­ists should win. So I spent my en­tire break talk­ing on Gchat with one of my CFAR friends, try­ing to see if he could help me come up with a sug­ges­tion that the doc­tor would agree was good. This wasn’t some­thing ei­ther of us were trained in, and hav­ing some­thing to pro­tect doesn’t ac­tu­ally give you su­per­pow­ers, and the one cre­ative solu­tion I came up with was worse than the sta­tus quo for sev­eral ob­vi­ous rea­sons.
I went home on day four feel­ing to­tally drained and hav­ing asked to please have a differ­ent pa­tient in the morn­ing. I came in to find that the pa­tient had nearly died in the mid­dle of the night. (He was now in­tu­bated and se­dated, which wasn’t great for him but made my life a hell of a lot eas­ier.) We even­tu­ally trans­ferred him to an­other hos­pi­tal, and I spent a while feel­ing like I’d per­son­ally failed.
I’m not sure whether or not this was a no-win sce­nario even in the­ory. But I don’t think I, per­son­ally, could have done any­thing with greater pos­i­tive ex­pected value. There’s a good rea­son why a doc­tor with 10 years of school and 20 years of ICU ex­pe­rience can over­ride a newly grad­u­ated nurse’s opinion. In most of the pos­si­ble wor­lds, the doc­tor is right and I’m wrong. Pretty much the only thing that I could have done bet­ter would have been to care less–and thus be less frus­trated and more emo­tion­ally available to com­fort a guy who was hav­ing the worst week of his life.
In short, I fulfilled my re­spon­si­bil­ities to my pa­tient. Nurses have a lot of re­spon­si­bil­ities to their pa­tients, well speci­fied in my years of school­ing and in var­i­ous doc­u­ments pub­lished by the Col­lege of Nurses of On­tario. But nurses aren’t ex­pected or sup­posed to take heroic re­spon­si­bil­ity for these things.
I think that over­all, given a sys­tem that runs on hu­mans, that’s a good thing.

The Well-Func­tion­ing Gear

I feel like maybe the hos­pi­tal is an emer­gent sys­tem that has the prop­erty of pa­tient-heal­ing, but I’d be sur­prised if any one part of it does.
Sup­pose I see an un­usual re­sult on my pa­tient. I don’t know what it means, so I men­tion it to a spe­cial­ist. The spe­cial­ist, who doesn’t know any­thing about the pa­tient be­yond what I’ve told him, says to or­der a tech­netium scan. He has no idea what a tech­netium scan is or how it is performed, ex­cept that it’s the proper thing to do in this situ­a­tion. A nurse is called to bring the pa­tient to the scan­ner, but has no idea why. The scan­ning tech­ni­cian, who has only a vague idea why the scan is be­ing done, does the scan and spits out a num­ber, which ends up with me. I bring it to the spe­cial­ist, who gives me a di­ag­no­sis and tells me to ask an­other spe­cial­ist what the right medicine for that is. I ask the other spe­cial­ist – who has only the sketchiest idea of the events lead­ing up to the di­ag­no­sis – about the cor­rect medicine, and she gives me a name and tells me to ask the phar­ma­cist how to dose it. The phar­ma­cist – who has only the vague out­line of an idea who the pa­tient is, what test he got, or what the di­ag­no­sis is – doses the med­i­ca­tion. Then a nurse, who has no idea about any of this, gives the med­i­ca­tion to the pa­tient. Some­how, the sys­tem works and the pa­tient im­proves.
Part of be­ing an in­tern is ad­just­ing to all of this, los­ing some of your delu­sions of hero­ism, get­ting used to the fact that you’re not go­ing to be Dr. House, that you are at best go­ing to be a very well-func­tion­ing gear in a vast ma­chine that does of­ten te­dious but always valuable work. –Scott Alexander
The med­i­cal sys­tem does a hard thing, and it might not do it well, but it does it. There is too much com­plex­ity for any one per­son to have a grasp on it. There are dozens of mu­tu­ally in­com­pre­hen­si­ble spe­cialties. And the fact that [in­sert generic nurse here] doesn’t have the fain­test idea how to mea­sure elec­trolytes in blood, or build an MRI ma­chine, or even what’s go­ing on with the pa­tient next door, is a fea­ture, not a bug.
The med­i­cal sys­tem doesn’t run on ex­cep­tional peo­ple–it runs on av­er­age peo­ple, with pre­dictably av­er­age lev­els of skill, slots in work­ing mem­ory, abil­ity to no­tice things, abil­ity to not be dis­tracted think­ing about their kid’s prob­lems at school, etc. And it doesn’t run un­der op­ti­mal con­di­tions; it runs un­der av­er­age con­di­tions. Which means work­ing over­time at four am, short staffing, three pa­tients in the ER wait­ing for ICU beds, etc.
Sure, there are prob­lems with the ma­chine. The ma­chine is in­effi­cient. The ma­chine doesn’t have all the cor­rect in­cen­tives lined up. The ma­chine does need fix­ing–but I would ar­gue that from within the ma­chine, as one of its parts, tak­ing heroic re­spon­si­bil­ity for your own sphere of con­trol isn’t the way to go about fix­ing the sys­tem.
As an [in­sert generic nurse here], my sphere of con­trol is the four walls of my pa­tient’s room. Heroic re­spon­si­bil­ity for my pa­tient would mean...well, op­ti­miz­ing for them. In the most ex­treme case, it might mean kil­ling the itin­er­ant stranger to ob­tain a com­pat­i­ble kid­ney. In the less ex­treme case, I spend all my time giv­ing my pa­tient great care, in­stead of helping the nurse in the room over, whose pa­tient is much sicker. And then some­times my pa­tient will die, and there will be liter­ally noth­ing I can do about it, their death was causally set in stone twenty-four hours be­fore they came to the hos­pi­tal.
I kind of pre­dict that the re­sults of in­stal­ling heroic re­spon­si­bil­ity as a virtue, among av­er­age hu­mans un­der av­er­age con­di­tions, would be a) ev­ery­one step­ping on ev­ery­one else’s toes, and b) 99% of them quit­ting a year later.

Re­cur­sive Heroic Responsibility

If you’re a gear in a ma­chine, and you no­tice that the ma­chine is bro­ken, your op­tions are a) be a re­ally good gear, or b) take heroic re­spon­si­bil­ity for your sphere of con­trol, and prob­a­bly break some­thing...but that’s a false di­chotomy. Hu­mans are very flex­ible tools, and there are also in­finite other op­tions, in­clud­ing “step out of the ma­chine, figure out who’s in charge of this shit, and get it fixed.”
You can’t take re­spon­si­bil­ity for the in­di­vi­d­ual case, but you can for the sys­tem-level prob­lem, the long view, the one where peo­ple eat badly and don’t ex­er­cise and at age fifty, mor­bidly obese with a page-long med­i­cal his­tory, they end up as a slow-mo­tion train wreck in an ICU some­where. Like in poker, you play to win money–pos­i­tive EV–not to win hands. Some­one’s go­ing to be the Minister of Health for Canada, and they’re likely to be in a po­si­tion where tak­ing heroic re­spon­si­bil­ity for the Cana­dian health care sys­tem makes things bet­ter. And prob­a­bly the cur­rent Minister of Health isn’t be­ing strate­gic, isn’t tak­ing the level of re­spon­si­bil­ity that they could, and the con­cept of heroic re­spon­si­bil­ity would be the best thing for them to en­counter.
So as an [in­sert generic nurse here], work­ing in a small un­der­staffed ICU, watch­ing the end­less slow-mo­tion train wreck roll by...maybe the ac­tual meta-level right thing to do is to leave, and be­come the freak­ing Minister of Health, or befriend the cur­rent one and in­tro­duce them to the con­cept of be­ing strate­gic.
But it’s fairly ob­vi­ous that that isn’t the right ac­tion for all the nurses in that situ­a­tion. I’m wary of ad­vice that doesn’t gen­er­al­ize. What’s differ­ence be­tween the nurse who should leave in or­der to take meta-level re­spon­si­bil­ity, and the nurse who should stay be­cause she’s needed as a gear?

Heroic re­spon­si­bil­ity for av­er­age hu­mans un­der av­er­age conditions

I can pre­dict at least one thing that peo­ple will say in the com­ments, be­cause I’ve heard it hun­dreds of times–that Swim­mer963 is a clear ex­am­ple of some­one who should leave nurs­ing, take the meta-level re­spon­si­bil­ity, and do some­thing higher im­pact for the usual. Be­cause she’s smart. Be­cause she’s ra­tio­nal. What­ever.

Fine. This post isn’t about me. Whether I like it or not, the con­cept of heroic re­spon­si­bil­ity is now a part of my value sys­tem, and I prob­a­bly am go­ing to leave nurs­ing.

But what about the other nurses on my unit, the ones who are com­pe­tent and mo­ti­vated and cu­ri­ous and re­ally care? Would fa­mil­iar­ity with the con­cept of heroic re­spon­si­bil­ity help or hin­der them in their work? Hon­estly, I pre­dict that they would feel alienated, that they would as­sume I held a low opinion of them (which I don’t, and I re­ally don’t want them to think that I do), and that they would flinch away and go back to the things that they were do­ing any­way, the role where they were com­fortable–or that, if they did ac­cept it, it would cause them to burn out. So as a con­se­quen­tial­ist, I’m not go­ing to tell them.

And yeah, that both­ers me. Be­cause I’m not a spe­cial snowflake. Be­cause I want to live in a world where ra­tio­nal­ity helps ev­ery­one. Be­cause I feel like the rea­son they would re­act that was isn’t be­cause of any­thing about them as peo­ple, or be­cause heroic re­spon­si­bil­ity is a bad thing, but be­cause I’m not able to com­mu­ni­cate to them what I mean. Maybe stupid rea­sons. Still both­ers me.