When all Possible Diagnoses are Rare and Doctors’ Predictions Differ Widely

“An infection of that size of that bacteria in that part of the lungs is very unlikely. It’s probably a fungal infection.” “A fungal infection producing so few symptoms is very unlikely. It’s probably cancer.” “Cancer after negative biopsies in two different places is very unlikely. It’s probably all due to the bacterial infection.” “Negative biopsies are a dime a dozen. Cancer is still a real possibility.”

Doctors’ predictions were all over the map during my husband’s recent medical episode. So were their reactions to new test results. Intellectually I understood why: once all the remaining logical possibilities were rare, most of their clinical experience didn’t help in assessing probabilities. At the time, the diverging probabilities and their explanation provided health-economist-me and my biostatistician-husband with a bit of distraction from our emotions. Now that things look virtually certain to be fine, I see some lessons for patients, their families and physicians.

Throughout the episode we got the opinions of lots of doctors. There were the official ones: our primary care physician (PCP), the hospitalist, the pulmonologist and the infectious disease doctor. We also solicited feedback from friends and family members who are physicians, particularly my hospitalist brother, who was always willing to talk the whole thing through.

I’ll the skip the clinical twists and turns. Here are the essentials: Mid-May Howard got his two-month-old cough and week-old nighttime breathing noises examined. A chest X-ray revealed a big mass. Before the cultures and biopsy, lung cancer looked reasonably likely. (Howard had never smoked, but many non-smokers get lung cancer.) Once the culture for haemophilus bacteria (usually an upper respiratory bug) came back positive and the biopsy came back negative, though, predictions varied a lot. As new results came in, reactions to them continued to vary for some time.

All the doctors did agree on the reason Howard’s case was unusual: He had been taking an immunosuppressive drug for a skin condition. That made possible a variety of unusual infections—including what turned out to be the only condition he had.

I was very aware of the diverse predictions, because I made a point of always asking two questions. The first is the one Jerome Groopman recommends: “What are all the things it could it be?” The second was my personal obsession: “What do you think the (rough) probability of each of those options is?” All the doctors stressed—correctly and importantly—“We don’t know.” But I wanted to know as much as possible and so I kept asking.


Happily, all looks well now. I could react by being happy to think about other things—and I am. But being overly intellectualized, I hope to rescue a few insights from all the time and worry. And I see some.

My first takeaway is that Groopman is so right: Always ask for all possible explanations of the symptoms. Groopman argues this question stops doctors from putting on blinders. In our case, it helped this patient’s family member from getting into an unreasonably negative—or positive—frame of mind.

I also see lessons about how to deal with uncertainty and accept that action is not always a good choice. (And I mean not a good choice even ignoring the financial costs to society.) I need these lessons, because I, like many others, have a lot of trouble not acting. The part of my brain that deals with emotions is shouting at me: “Do something! Do everything!” And it shouts that despite all my training about uncertainty and the limited value of tests in various circumstances.

The first lesson is to think the way I was trained. At one point our superb PCP, knowing our professions, was reminding us of Bayes’ rule and emphasizing that another CT scan at that moment was not going to tell us much. This lesson may not sound helpful for patients and family members without such training. But actually it is. Patients, their families and doctors talking to them should just ask: “what are we going to learn from this test and how will it affect what we do?”

The second lesson was to think through the consequences of waiting under each of the possible options. My first instinct had been to want Howard to have a needle biopsy quickly, even though it’s fairly invasive. But questions like, “what if it was cancer, would waiting another few weeks matter?” proved helpful. There were none of the signs a fast growing cancer would produce. Waiting a few weeks was extremely unlikely to matter.

The third lesson is to do those things in advance that are worth doing. That way one feels prepared. Our PCP had looked into surgeons in case it was cancer. I knew who we would go to and what tests and documents would be needed if it did turn out to be cancer. And obviously, there are a lot of tests and treatments that are worth doing sooner, rather than later.

You may be thinking: okay it’s good that you figured out what works for you, but that doesn’t mean these approaches are right for others. To some extent that is right, people need to know their own personalities and what works for them. And ideally their primary care doctors need to know them too—and be involved. But everyone—particularly us “do everything” personalities—needs to learn to live with uncertainty and to wait when waiting is the best choice.



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s