Three Shots at Immortality

The Cifu Sign, The Cifu-Index, and The Cifu-Eight

As the years advance and my days in medicine become numbered, I think about how quickly I will be forgotten. Not a day goes by that I don’t sit in my office, smoke a Camel, and consider how soon they will empty and deodorize my little room.

Sure, people tell me that my patients and students will remember me. But most of my patients are over 80, and my students, well, I am not sure I want to be remembered that way.

I’ve made a few shameless pitches at immortality in the past, which, shockingly, have not caught on.

First, there was the Cifu-Sign.

This is a diagnostic test. If a patient says the words “shopping cart” without provocation, he or she has spinal stenosis. The sensitivity of the Cifu Sign is only 50%, but the specificity is 98%. Ok, I don’t actually know the test characteristics, because I never bothered to study them, but 50% and 98% seem about right to me.

Next, there was the Cifu-Index.

The Cifu-Index is, in a way, a negative version of the h-Index.¹ The Cifu-Index is calculated as follows:

Number of articles published but never cited/Total number of articles published²

If an academic has a Cifu-Index of 1, he or she has only published articles that have been ignored. A Cifu-Index of 0 means that someone is paying attention to all of your work. (It’s kinda like golf, the lower the better. Mine stands at 0.19.)³

There could be a further derivative called the CifuPrime Index:

(Papers not cited + Papers only self-cited)/Total articles published

If your CifuPrime is twice your Cifu-Index, you are a self-important self-citer.

Today, I make the pitch for the Cifu-Eight

This is a bit of a ripoff of the Stanford 25. This list is derived from a card that I have been, halfheartedly, handing to the interns I work with in our urgent care. I think the rules guide better care in any setting where new patients are seen for new complaints – emergency rooms, urgent cares, and even in the office when you are seeing a patient once for a new concern. The rules are meant to guide students and interns, but they also serve as a good reminder for more seasoned clinicians.

1. Don’t suggest an evaluation without a differential diagnosis.

   Suggesting an evaluation without developing a differential diagnosis is all too common. It originates from, and leads to, cognitive sloth. You see it when a trainee goes from reporting a patient’s complaint to suggesting tests without pausing to suggest possible etiologies. There is no doubt in my mind that poorly reasoned testing leads to worse care and excess cost.

2. Consider three possible diagnoses for the primary problem: the most likely, the most common, and a must-not-miss diagnosis.

   This is one of the foundations of strong clinical reasoning. The three-diagnosis differential guards against early closure and protects the patient from a life-threatening diagnosis being missed. If you want to improve your own differentials, add a fourth diagnosis category: the zebra.

3. Present an argument against your favored diagnosis.

   We often devalue or ignore data that does not align with the diagnosis we favor. If you become your own gadfly, you recognize the flaws in your reasoning.

4. Don’t exclude a diagnosis based on the absence of a sign or symptom

   Scott Stern and I did a lot of work to show that while historical and physical findings are sometimes specific, they are seldom sensitive. If you ever hear someone say, “this person can’t have diagnosis X because they don’t have symptom/sign Y”, know they are making a mistake.

5. Obey the law of parsimony

   Yes, there is something to Hickams Dictum, but Occam’s razor still holds the day.⁴

6. Note at least one physical finding.

   The physical exam is a dying art. This is in part because our 21st-century technology has made it less important. That said, the physical exam has a role, and we all need to teach it and practice it. Demonstrating a warranted physical exam maneuver to a trainee on every patient goes a long way toward teaching the importance, utility, and performance of the physical exam.

7. Discover something interesting about the patient.

   You take better care of people you know. Investing the time to make a discovery, even with someone you may never see again, pays dividends. It makes the patient feel cared for. It makes the visit more interesting to you. It may even reveal critical information about the diagnosis.

8. Reflect on something the case reminds you of.

   This serves a lot of purposes. It helps me get to know my trainees better. It probably makes me a better teacher, getting me to talk about related cases. It helps everybody remember the patient and, in doing so, all that there is to learn from them. We all know we learn medicine better when we link facts to a patient. (Sometimes, I even learn about a great mini-series I would have missed).

In reality, I am pretty comfortable just being forgotten. I think I’ll just kick up my feet and have another cigarette, maybe a Lucky Strike this time.

1. Unlike the H-Index, the Cifu-Index should never be shortened to the C-Index. Although I am terribly jealous of Jorge E. Hirsch, I also feel for him, as his name is slipping away. The Wikipedia page for the h-index reads:

The index was suggested in 2005 by Jorge E. Hirsch, a physicist at UC San Diego, as a tool for determining theoretical physicists’ relative quality and is sometimes called (emphasis added) the Hirsch index or Hirsch number.

2. If you can get a Wikipedia page published on the Cifu-Index, there is some Sensible Medicine merch headed your way.

3. I have no idea if 0.19 is good or bad. For comparison, Vinay Prasad’s is 0.08… better.

4. Yup, Hickam and Occam, two other immortals.