When patients have a terminal disease, candid but compassionate conversations about life expectancy prove vitally important. These conversations help patients and family plan care that is more appropriate, make life plans, and start coping emotionally. Conversations about life expectancy are associated with improved quality of life, greater satisfaction with healthcare, patient survival improved by days to months, and other benefits. Nevertheless, the importance of these conversations begs the question, “Do doctors really know how long a patient has to live?” How accurate are their predictions?

What Do You Call It When A Doctor Predicts How Long Someone Has to Live?

In medical terminology, this is referred to as issuing a prognosis. A prognosis refers to the clinician’s prediction of how long a patient has to live. This is also sometimes referred to as a survival estimate.

What Affects the Accuracy of Survival Estimates?

When the question is “How accurate are survival estimates?”, the answer may be a moving target. The healthcare profession has been scrutinizing its ability to do this for decades. Because new methods have been introduced and implemented, accuracy may be changing. Additionally, the accuracy of survival estimates may be affected by variables such as the disease in question, the type of professional making the prediction, the experience of the clinician, and the length of life remaining. Some diseases follow a more predictable course than others, and there is a popular theory that prognoses become more accurate as the survival time shortens.

How Accurate Are Survival Estimates?

Across various studies, prognostic accuracy spans from 23% to 78%. Put another way, survival estimates tend to range from three months short to three months long. This data was gathered by Nicola White and colleagues.¹ Their systematic review included 42 studies covering 30 years of research and over 12,000 prognostic estimates. However, the studies were often different from each other in important ways, and there are variables affecting prognostic accuracy. For these reasons, it may be helpful to drill into the data a little deeper.

Doctors Tend to Be Overly Optimistic When Estimating Survival

Inaccurate prognoses may be short or they may be long, but they are usually too long. When lumping together all the studies, White et al. found that clinicians tend to overestimate survival by a factor of two. In their collection of data, that was a median prediction of 44 days compared to a median survival of 25 days.

Which Healthcare Providers Make the Most Accurate Prognoses?

Overall, White’s research team concluded that no one type of healthcare provider clearly stood out as better at estimating survival. They state this, but in their two studies that compared oncologists to other types of clinicians, the oncologists gave more accurate prognoses. Various studies compared hospital doctors, general practitioners, oncologists, nurses, radiologists, and allied health professionals. Interestingly, nurses tended to be better at recognizing patients who were imminently dying, and nurses who had seen their patients within the last 24 hours were more accurate than those who had not. There was some evidence that general practitioners were better prognosticators than hospitalists, and that radiologists were more accurate than allied health professionals. Despite these differences found by individual studies, the overarching conclusion is that most clinicians are roughly equal at estimating survival.

On the other hand, the evidence does seem to endorse a team approach. Overall, doctors and nurses together tend to issue more accurate survival estimates than either alone. Multi-disciplinary teams tend to offer the most accurate survival estimates. 

What Other Factors Affect Prognostic Accuracy?

Some papers reported that experience level made predictions more accurate, but others did not find this. Timeframe did affect accuracy. Compared to patients who had a year to live, survival estimates became more accurate closer to the end of life. 

Conclusion

Predicting survival is a sensitive subject. Survival estimates are often news that people don’t want to hear but do want to know. In terminal disease management, this can make candid conversations about prognosis challenging. Under one year, prognoses tend to be accurate within three months. That may mean three months short or three months long, but prognoses are more often overly optimistic. Doctors of various types, nurses, and radiologists can all make survival estimates with relatively equal accuracy. The most accurate approach to survival estimates is probably the multi-disciplinary team.

References:

1. White N, Reid F, Harris A, Harries P, Stone P. A systematic review of predictions of survival in palliative care: how accurate are clinicians and who are the experts?. PloS One. 2016 Aug 25;11(8):e0161407.