Dr. Walkeris an emergency physician at Kaiser San Francisco. He is the developer and co-creator of MDCalc (www.mdcalc.com), a medical calculator for clinical scores, equations, and risk stratifications, and The NNT (www.thennt.com), a number-needed-to-treat tool to communicate benefit and harm. Follow him @grahamwalker for daily thoughts on emergency medicine, and read all of his past columns athttp://bit.ly/WalkerEmergentology.
One of the toxicologists broke down the indication for treating overdoses for me pretty simply when I rotated through Bellevue's Poison Control Center: some clinical signs or symptoms (like ataxia from lithium), blood levels that are getting worse (like a drug level trending up or not trending down),and my favorite, “the chicken factor.” This is, if I can paraphrase it, “when a patient's level is so high that you're too chicken to let it go any higher.”
We typically care about two types of values in medicine: relatives and absolutes. That's essentially what the toxicologists mean. Levels worse than a prior level? Relative. Signs and symptoms? Chicken factor? Absolute. It is almost always easier to pull the trigger on the absolutes than the relatives because the former requires very little further investigation while the latter always takes more time and energy to dig deeper. You can actually think of all data collection this way: at what point will a particular piece of data force an action (giving a medicine, ordering a test, discharging a patient)?
- If an immigrant from Nicaragua tells you he's lost 30 pounds and has hemoptysis and night sweats, you don't really care if he has had them before. He passed your threshold—not compared with anything else—and he's getting worked up for active tuberculosis. Absolute. If a healthy 40-year-old comes to you with some blood-tinged sputum after a week of a cough, however, you're going to ask him if it's happened before and probably some other screening questions. Relative.
- If a patient with fatigue has a hemoglobin of 5, you're probably going to stick a finger up his bottom. Absolute. If he comes in with a hemoglobin of 9, you probably want to know what his last value was. Relative.
You do this all the time with some of your chronic emergency department patients as well. You might ask your colleague, “Do you always admit this guy for his pancreatitis?” or you curbside the nurse, “I heard you took care of this patient last night. Did she have that big hematoma on her scalp yesterday when you discharged her?”
All of this is why the electronic medical record has become so damn useful. It makes the relative values so much easier to find and compare; this is especially true for emergency physicians. We obviously do a lot of treatment, but I would argue that the vast majority of our job is rapid data-gathering and interpretation.
Step one: Meet patient, gather data on his reliability and sources of information.
Step two: Take history and do physical, gather data on general problem and findings specific to that problem.
Step three: Order testing — studies, labs —perhaps to gain some more objective data on the patient and his problem.
Step four: Gather previous data — discharge summary, prior creatinine, list of medications, prior echo, prior chest x-ray, or EKG.
Step five: Throw steps 1–4 into a big pot, stir, simmer for a few minutes, and serve up a treatment anddisposition plan.
Technology doesn't do a whole lot to speed up the first few steps, but it helps immensely with step four, and I thank the database gods who keep it up and running every day.
I'm very fortunate to see mostly patients in my health care system (Kaiser), which allows me to look at every old chest x-ray, EKG, head CT, CD4 count, clinic visit, and medication summary that I can possibly think of. I think it makes my data collection faster and more efficient, and it also helps me formulate a better plan, which ultimately helps me take better care of patients. I can't think of the number of times I've looked at a chest x-ray and might have called a pneumonia when that same spot has shown up on the patient's film for the past five years. In fact, prior EKGs are recognized as so important that when a new EKG is done, the prior one is automatically printed out and handed to me at the same time.
Easily being able to see prior clinic notes and prescriptions is incredibly useful, too. Reassurance is often the appropriate plan of action for a child who has had a fever and cough for a day, but you probably have a much lower threshold for some imaging or further evaluation if he is back for visit four for the same symptoms.
And what if you see that an anesthesiologist decided to write “history of difficult intubation” in the patient's problem list? I can't possibly imagine more critical information when you're considering sedating and paralyzing someone. Similarly, if I see “comfort care,” “lung cancer, stage 4,” “metastasis to bone,” or “metastasis to brain” and the patient appears to be not long for this world, I can prioritize calling family and social work over respiratory therapy, and wear my palliation hat instead of my critical care one when I walk into the room.
This is not to say that technology is a cure-all or that the modern-day EMR is perfect. We still have major problems with usability and user interface in such a critical system like a medical data-gathering program, and we have to deal with “copy/paste creep” — junky data that arepassed from admission to admission because you can just take some other physician's note and make it your own. Now, thereare more billing, administrative, medico-legal, and regulatory requirements to document more often, partially because it's easier to document.
But overall, I'd still say the electronic medical record has improved my care substantially, especially compared with the old paper chart. The chicken factor has always been easy, but electronic medical records are making those loosey-goosey eggs a little bit easier to crack.