A third of the way through medic school, I had to show I could handle a simulated megacode--a cardiac arrest that morphs from one lethal arrhythmia to another. My instructor sat patiently as I verbalized every bit of personal protection not requiring a carry permit, performed a "quick look" on a chronically unresponsive manikin, defibrillated a disorganized rhythm I prayed wasn't pulseless electrical artifact, delegated CPR to imaginary probies, started an IV on a disembodied arm, then pushed a milligram of epinephrine because ineffectiveness wasn't a contraindication.

   "Mike," my examiner interrupted as I contemplated my next lifesaving intervention, "do you know what a mulligan is?"

   "Uh, no." I hoped "mulligan" wasn't the correct answer to a quiz question I'd missed earlier.

   "Not a golfer, eh? A mulligan is a second chance, and you just got one. Don't waste it."

   In a nanosecond I realized I'd committed the most egregious of scenario-based sins: failure to control the airway. I should have oxygenated and intubated my synthetic patient immediately after directing my phantom crew to commence chest compressions. With my preceptor's permission I rewound resuscitation to that frame, then demonstrated my ability to ventilate through the correct hole.

   Today I'm wishing I'd had a few more mulligans, because patients may have died whom I could have saved if only I'd forgotten to manage their airways promptly. That omission--serious enough 16 years ago to almost cost me a passing grade--is now standard of care. It says so in the American Heart Association's '05 version of Advanced Cardiac Life Support. Not only do we delay invasive airway management, but also pulse checks, rhythm reinterpretation, medication administration and patient movement, so chest compressions aren't interrupted for more than 10 seconds at a time. I don't know how many cardiac arrests I ran before 2005--a hundred, perhaps--but I'm quite sure we stopped CPR for much longer than 10 seconds at least once during every one of them.

   Soon we'll be introduced to another iteration of ACLS. It happens every five years--longer than the average bear market, shorter than the life cycle of North American cicadas. Is there a reason, other than reliance on a decimal system, that we commit to new algorithms for half a decade per revision? That's a long time to experiment with critical care. Would we be saving more lives if our founding fathers had favored $3 bills, and counted to six instead of 10 when they were angry?