Nil Per Os - Where Did It Come From and What Does It Do?


I had an amazing blog prepared. Seriously. The best blog probably ever of all-time. But I decided to forego it for this one today. Why? Because now both Anna McDonald and Maureen have asked me to do today's topic. That's not only two requests, but two people who have read my blog! Two! I'm on service for one more day so I can't deny our incoming Chief AND Attending. But for real, the blog I was going to do in place of this one was going to be really, really good. Really good. 

So here are the questions Arnold so elegantly prefaced for me: 

1. Where did NPO at midnight come from?
2. Do people REALLY have to be NPO at midnight for procedures?

 "NPO at midnight" is just another one of those traditions we have carried onwards throughout medicine. EGD tomorrow? Hang the "NPO at midnight" sign. IVDU septic knee washout for the 4th time this hospitalization? NPO at midnight. Just the possibility of a procedure tomorrow? You get the picture. Now personally, I don't really care about being NPO myself after midnight. I only eat breakfast half the time anyways (Doctor's lounge stale muffins FTW). But in the context of a malnourished geriatric/pediatric patient or even the patient who wants to eat so much that they disregard what every physician had told them that morning and eats anyways (not bitter), it can be a bigger deal. This is especially true if the patient who does eat ends up having to have their procedure postponed for the following day. That's roughly an additional $2,500 in the state of Washington (for-profit hospitals) plus a delay in care and increased exposure to all the fun, resistant bugs we carry in the hospital.

So where did this all come from? Well, after some reading around. I think it came from this guySir Joseph Lister was a surgeon back in the late 1800s who was quite the authority on medicine at the time. Per this correspondence by Dr. Roger Maltby in the Canadian Journal of Surgery in 1999 [1], the good sir recommended in his textbook no liquids 2 hours before surgery and no solids 4-6 hours prior to surgery. The aforementioned rule was apparently the standard of care until the 1960s, according to Maltby, until the Americans bastardized the standard and started recommending NPO at midnight instead. And without evidence he asserted! Sounds like 'Murica. I tend to trust Maltby on this topic as seemingly every other study in the Cochrane trial mentioned below was one of his. 

Ok, we now know where it came from. Now what does it do?

Well for once there really IS an easy answer.  That is due to my good friends over at Cochrane - love you guys. In short, there is NO significant difference between the traditional NPO at midnight and a preoperative fluid/food/shortened/anything fast (I'll explain the details below). Cochrane looked at 38 RCTs (22 clinical trials) of mainly healthy adults across multiple countries and evaluated the following primary outcomes [2]:

  • Rates of adverse events (i.e. aspiration/regurgitation)
  • Volume and/or pH of gastric contents (surrogate)
  • Concentration of marker dye as an indicator of gastric emptying (surrogate)

I was specifically interested in the duration of fasting such as 8-12 hours versus 2-4 hours but they looked at a myriad of scenarios across these trials. It reads more like a list of court cases:

  • Short Fluid Fast (120-180 min pre-op) v. Standard Fast - Verdict: No difference in 20 trials
  •  Short Fluid Fast (120-180 min pre-op) v. Standard Fast (w/ H2 blocker):  Verdict: No difference in 8 trials
  • Short Solid Fast (249 and 199 minute means) vs. Standard Fast: Verdict: No difference in 2 trials
  • Short Solid Fast  v. Short Liquid Fast (~330 minutes): Verdict: No difference in 1 trial
  • *My FavoriteUNLIMITED FLUID INTAKE v. Standard Fast: Verdict: No difference in 3 trials! 
  • Plus many more studies that compared volume of permitted fluids or type of fluids

They also looked at a slew of secondary outcomes (hunger, thirst, nausea, etc) which was surprisingly only really positive for thirst. Oddly enough they didn't find any significant differences in hunger or nausea when comparing patients given pre-operative oral fluids to standard fasting patients. That data isn't really applicable though as our patients don't know that and I have a large enough personal sample size of NPO complaints to bias me.

To further magnify how lopsided this argument is I looked at the outcomes of every single trial in every single variable possible with normal adults in the Cochrane review. That's DOZENS of trials. What I found was there was not a variable that favored a standard fast with the exception of one.


There were two trials with 110 participants that found that there was a significantly lower gastric pH in the standard fasting group (p = 0.04). That study was a group that was using a H2 receptor antagonist (H2RA) while comparing fruit juice versus a standard fast. Unfortunately, that was the only subgroup that was significant as the one trial without an H2RA + fruit juice, and the other trial looking at gastric volumes both found no difference. So if you have a patient with hypoglycemia that HAS to have fruit juice AND you're using a H2RA, you better damn well know that you are possibly decreasing the gastric pH in their stomach. That's on you son.

Oh and for the sake of completeness. This pretty much applies to children as well. [3]

So what did I learn today? I reinforced that I love Cochrane more than ever. I also learned that if I have a patient who "accidentally" drinks some water before their procedure, they are still getting that procedure done. Well, at least I'll argue for it. The same applies to patients that eat prior to a procedure.

And lastly, every headache you come across is not a tooma. Thanks Arnold.