Thursday, April 30, 2015
by Amy Clarkson
When we turn on the news or listen to the radio to hear what the meteorologist is predicting the weather will be, we don’t usually call this ‘the day’s weather prognosis’. We use the term forecast, and yet a forecast is essentially the same concept of what a prognosis is.
Predicting the weather and predicting the outcome of a life limiting disease may carry the same foundation, but they differ vastly on their accuracy. Weather forecasting has an overall accuracy for temperature determination (give or take a few degrees) of 85%. A medical prognosis, however, regarding how long someone has to live (give or take a few days or weeks) is usually only accurate 20% of the time, and actually 65% of the time doctors overestimate prognosis by a factor of 3-5.
Yet so often patients who have a terminal disease leave the doctor having latched onto a prognosis as if it’s as guaranteed as an expiration date on grocery goods.
Many factors go into misinterpretations of prognosis. I’ve mentioned overestimation, but research also shows that the longer a doctor has taken care of a patient or the more they know that patient, the more they overestimate. Besides overestimation, another challenge is how doctors even approach prognosis. Often doctors will use something called the median survival rate when discussing prognosis. Patients get this confused with average. Median, however, is just the middle number of a range. If a doctor says, the median survival is 6 months, that means half of the people with that disease die before 6 months. What median doesn’t tell us is, of those people who died before 6 months, did most of them die in 1 month or 5 months? That’s a big difference!
One thing doctors don’t do well is to discuss the worst-case scenario along with the best-case scenario. They assume it’s too depressing to talk about bad outcomes with the patient. While sometimes during prognosis talks patients are shocked, retreat into their brains, and can’t listen to the prognosis, often times patients are waiting for doctors to initiate these tough discussions. One suggestion that can help with both these problems is to have the doctor actually write out the worst-case prognosis, the best-case prognosis and the most likely prognosis.
What can be very challenging is that many prognoses have a very wide range of possibilities. Some cancers carry the possibility of dying in 6 months or living out the rest of your life essentially cured. It’s the same as if a meteorologist announced that for tomorrow’s forecast the possibility was for snow, or maybe rain, but also could be hot and sunny. If you were traveling to a location like that, you’d pack for the extremes, wouldn’t you? This is how we need to treat prognosis as well. Prepare for the extremes.
A good prognostic conversation should do this; explore the extremes so that you can plan. The next time you hear a prognosis think of it as a forecast, but one with only 20% accuracy. Also, be sure to ask for the worst-case and best-case extremes, it will help you prepare for the journey.
Dr. Clarkson is a hospice physician for Southwind Hospice in Pratt, KS. This post was originally published in Dr. Clarkson's End Notes column for the Pratt Tribune. It is re-published here with the author's permission under a Creative Commons license.