Wednesday, May 9, 2012

Miffed at the Myth of MIVF and nonplussed by nil per os (NPO)

I will have to consult with physicians from a former generation to determine from whence came the concept of "maintenance intravenous fluids" or MIVF.  Early in medical school, especially during the pediatrics and surgery rotations, medical students are taught how to calculate fluid deficits and ongoing losses with the goal of selecting the optimal electrolyte solution and rate with which to replace them.  (Internists generally just select some base-10 friendly number such as 100 or 150 cc/hour and call it a day.)  This is a worthy enough goal, because dehydration is indeed one of the easiest and most worthwhile things to correct in the acutely ill patient.  (If you've ever puked your guts out for a day and then gotten a liter of intravenous fluid, you know what I mean.)

But the whole practice begs the question:  If there is a fluid deficit, why not just correct it once and be done with it?  If the average deficit in diabetic ketoacidosis (DKA) is 5 liters, why not just give 5 liters and stop?  I used to chide the residents during their ICU rotation:  "Bolus is for resus, rates are to maintain" to reinforce the notion that deficits should be repleted quickly.  But now I have come to question the second clause in that aphorism - do we need rates to maintain?

One reason for rates is to solve a problem that we ourselves created:  the NPO order.  Save for patients who need to be NPO in preparation for anesthesia or surgery on the gut, I don't know why NPO is so frequently written on admission.  It creates the fluid deficits that we will later have to address, makes patients uncomfortable, forces more cumbersome routes of drug administration, and exacerbates nutritional deficiencies.  If patients are allowed to take a diet, they won't need MIVF any more than I do - they will ask for a glass of water.  (And if a patient with say, DKA, eats a meal tray and then pukes it up - so what?  If s/he would rather eat and puke than be NPO, so be it.)

Another reason is the [false] belief that we can calculate with some degree of precision the "physiologic" daily fluid intakes and losses and that that calculus can allow us to meaningfully intervene on a patient's fluid balance.  I am not confident that we can rely on this calculus or that, in internal medicine, there is any calculus going on at all - it is just some rate per hour in increments of 25cc/hour, e.g., 75 or 100cc/hour.  And I know of few physicians who are aware of the volumes infused in all of the other intravenous medications they order on admission.  (I just had to go check with one of the RNs to find out that 3.375g of Zosyn comes in 100cc of carrier, so a days worth of Zosyn accounts for 25% of daily fluid needs.)  Oftentimes, the intravenous medications alone are sufficient to replace daily fluid needs.  Nonetheless, we often see patients in the ICU receiving not only massive volume infusions from multiple drips (the necessity of these is a topic for another post) but also the obligatory MIVF at 100cc/hour, perhaps a neglected carryover from the transfer from the floor.

Which segues me to one of the nettlesome drawbacks of MIVF - they are simply forgotten about, and lead to surreptitious status iatrogenicus - fluid overload and congestive heart failure, especially in the elderly.  An 89 year old woman with an ejection fraction of 38% weighs 55 kilos on admission for a fall and receives 100cc/hour of MIVF for 4 days in the hospital.  Pulmonary is consulted when her oxygen requirements escalate.  During this time, she has received TEN LITERS of fluid, and gained 22 pounds FROM MIVF ALONE!  (Where went the calculus during this admission?)  How many doses of furosemide will it take to correct this problem, and how long will it prolong her hospitalization?  I wager that this happens more frequently than dehydration would happen if we banished MIVF altogether and let people drink.

But suppose that we decide that in any particular patient there is a daily deficit that needs to be replaced - why do we replace it with a coutinuous infusion of MIVF?  Why not do what I do daily when I detect a fluid deficit via thirst - replace it with a bolus.  I see no good reason that justifies continuous MIVF infusions.  If the daily deficit is 1600cc, give it as a bolus in the AM or divided into two doses like any other medication.  This would facilitate the combat against another woesome order that promotes status iatrogenicus - Activity:  Bedrest.  This is another topic for another post, but surely attachment of the MIVF line impairs mobility in the hospitalized patient with all of the associated untoward consequences, e.g., deconditioning.  It's hard to get out of bed with a Foley, an MIVF line (for STRICT I's and O's for the calculus that won't be done of course!), a pulse-ox probe and telemetry wires, invariably woven into a confusing and complex tapestry, essentially a form of Posey bodynet over the bed.

If the extent of the deficit is unknown, or there are concerns about "putting the patient into failure" the intermittent bolus is definitely the safest route, because it forces you to reassess at intervals the adequacy or excess of your replacement strategy.  A rate left unattended risks BOTH under-resuscitation and fluid overload.  Bolus with reassessment is the safest and most reasonable way to go.

And of course there is the cost.  A bag of saline costs the hospital about $1, but the charge can be from $10-100.  Ignoring the distinction between costs and charges, imagine that you're in the hospital for 4 days and part of your bill is $100 (or $400 or whatever) for 10 bags of saline.  The soda machine on the golf course suddenly looks cheap.

And that last observation harkens to one of the fatal flaws of a third party payer system in American healthcare - the consumers of the services are shielded from data on cost.  If those bags (and the zofran syringe and the lovenox, and the CT scan, etc.) came with price tags attached to them, you can be assured that patients would be asking a lot more questions about the necessity of each and every item used in their care.  And that you can take to the bank.

For further reading, see:


  1. Enjoy your blog very much and agree with most of your musings. In defense of continuous IVF, if one is attentive enough to be able to give a fluid repletion bolus 1-2x a day one should surely be as attentive with a continuous drip. (The question of intravascular vs total body water can be argued another day.) Regarding mobility, if the pt can get out of bed he/she probably doesn't need IVF and can take PO. Strict I/O's do help with fluid balance. And surely you don't disagree that there is a baseline (if unknown) fluid requirement - pts who are NPO should probably be getting some sort of intake, the content of which can be debated. And many criticall ill patients do, in fact, fall into that category - those in extremis, with bowel ischemia, immediately post-op, etc.

  2. I agree with all of this, particularly the differences between surgical and medical patients, with my focus on the latter.

    To quote Franklin, "A foolish consistency is the hobgoblin of little minds." The status quo is often the status quo for a reason - but on many other occasions, it serves no purpose and should be reconsidered or challenged. That was the main point of this post.

  3. I have always been confused about the NPO thing. Patient is NPO, getting dehydrating nasal O2 and is dry. Like tongue sticking to roof of mouth dry. I have had my hand slapped away if I approached with the relief of an ice chip, hearing "He/she's NPO!!!" Ok, gotcha for food and impending O.R., and certain medical conditions. But he/she weren't so compromised they'd be making saliva, right? And...swallowing it, perhaps? Can't I offer the relief of a cold ice chip to drip backwards into the throat? Do we really have to wait for someone to find a disgusting glycerine swab? I mean, really.
    P.S. : read your blog as a marathon today. A link came to me by way of a suspicious power. And I think I'm in love with you:)


Note: Only a member of this blog may post a comment.