Prognostication – Do Physicians Have The Magic Ball?

As per dictionary.com the definition of “to prognosticate” is as follows:

Do physicians have a magic ball?

prog·nos·ti·cate 

1. to forecast or predict (something future) from present indications or signs; prophesy

2. to foretoken; presage: birds prognosticating spring

3. to make a forecast; prophesy

Is it only me, or does this sound very sobering? In other words, to prognosticate means to foretell the future. And who can really do that?

As a physician who deals with a lot of seriously ill people I have faced this problem many times: how can I estimate someone’s prognosis?

Why is this important? Does it really matter?

Time is running out.

I think it does. I believe that prognostication gives a framework to patients, their families and physicians. It makes a big difference whether someone expects to go on living for several more years or whether he has a high likelihood of dying within the next few weeks. There may be things to take care of, amends to be made, trips to be had. Making peace with the idea of the impeding departure from life on this planet may lead to appreciation of the time left and the chance to resolve open issues.

As we already discovered, prognostication is not an easy task, if not borderline impossible. Who can predict the future? None of us, really! 

How do we address prognostication scientifically then?

There are different tools that attempt to aide physicians with this problem.

How to predict somebody's life expectancy?

Ultimately, it comes down to looking at the patient as a whole:

• How old are they?
• How many illnesses do they have? 
• How well controlled are their illnesses?
• Do they have any terminal disease?
• At what stage is that disease?
• What is their functional status? (see previous blog post on that topic: click here)
• How rapidly did they decline? (e.g. compared to how they were doing 6 months ago)

Statistics provide framework

In June 2001 Dr. Walter and Dr. Covinsky published an article in JAMA titled: Cancer Screening in Elderly Patients in which they pointed out that more than just age alone a person’s overall life expectancy should be crucial to whether or not certain screening tests should be performed (more about these issues of cancer screening in future blog posts, click here for link to their article). They used statistics published by the CDC called “United States Life Tables” to categorize life expectancies (click here for an example of such a report).

They put these statistics into tables divided by genders and stratified by age groups. Then they looked at the life expectancy of the average person (50^th percentile) of an age group as well as the life exptectancies of people healthier (top 25^th percentile) or sicker (lowest 25^th percentile) than the average group.

Therefore, when a physician is trying to estimate someone’s general life expectancy he/ she is looking at whether the person is about average in health status or sicker versus healthier than average.

In oncology there is a commonly used term “median survival”. Dr. Eldridge explains this well on about.com (click here for link):

“Median survival is defined as the time after which 50% of people with a particular condition are still living, and 50% have died. For example, a median survival of 6 months would indicate that after 6 months, 50% of people with that condition would be alive, and 50% would have passed away.”

This concept may be difficult to grasp when it affects you. Ask your doctor how to apply this to your life.

Some tools frequently used are called ECOG Performance Scale (click here) and “Karnofsky Performance Status Scale” (click here for link) which help physicians estimate how functionally impaired a person is from the disease burden. The ECOG is often used in palliative care or oncology. As a matter of fact, the ECOG is one of the most important factors that oncologists take into consideration when assessing whether someone is a candidate for treatment or not.

I want you to keep something in mind: physicians are normal humans, not prophets that know the future. We are striving to learn more and more about the subject of prognostication and we are trying to do a good job. In the end of the day it remains an estimation based on a multitude of factors. We also don’t like to give “bad news” and often shy away from facing how ill a person is. Many times this does not get communicated to patients and their families.

Also, consider that physicians consistently overestimate prognosis. A very simplistic tool is the "surprise question": Moss et al. published an article in the Journal of Palliative Medicine in July 2010 (abstract here). Oncologists were posed with 853 patients and had to ask themselves: "Would I be surprised if this patient died in the next year?" There was a correlation between this question and the 1 year mortality of the patients. In a way this "surprise question" evokes a gut answer by the physician and as the study showed, sometimes a gut answer is not so far off.

Time is fleeting.

Don’t accept concrete numbers, though. Nobody can say “you will die in 2.5 months”, the only thing that can be said is “there is a high likelihood of her dying within the next few weeks to few months” (or hours to days, days to weeks, months to years… I think you see what I mean). 

The search for the perfect tools to prognosticate goes on. There are many potential benefits in knowing and accepting the limitation of time.

In the end of the day, nobody knows the future. It is wise to plan ahead with the right documentation (see my previous blog posts on Health Care Proxy and MOLST form) and the discussion about the topic with your loved ones.

No matter how you look at it. Life is short and full of surprises.

Make the best of today. Be prepared for the future but don’t allow anxiety and fear to dominate your present.

Make today count.

Make peace with others.

Forgive who has done you wrong and don’t hold grudges. 

Make today count.

I wish you well,

Dr. B