If we really want to give our patients information that is clear and useful we need to look outside of the medical and dental guidelines and the Global Intelligence Community has been wrestling with this for a long time 2—4. This challenge with using descriptive words only was first documented by Sherman Kent in for the CIA in an attempt to improve intelligence briefings following the Bay of Pigs disaster in He proposed that numerical odds were added to the descriptive words to add clarity between analysts and the decision makers.
The concern was that the numerical probability would be taken as a fact rather than as a probability and the forecaster could be accused of being wrong if the event did-not occur. Hence the preference to use words that can be interpreted elastically. The question then is, can using numeric probabilities help in communicating risk to our patients? The first test was to see how patients interpreted the words without any numbers or probabilities to anchor off.
We asked sixty consecutive general practice patients without asking the same patient twice what chance of success excellent, good, fair, guarded or poor meant. An example of the proforma given to the patient is indicated below. The patients were asked not to over-think the question but go with their first instinct and no additional guidance was given.
The result where charted below using a Box plot. Once the numbers were included much better resolution was achieved with the median figures being closer to the expected values. One must note however that there were still huge outliers in the interpretation as marked by the red asterisk on the second chart.
To gain valid consent clinicians need to understand that what they are saying does not necessarily align clearly with what the patient is understanding and there is a general trend to over-optimism.
This can lead to an exaggerated sense of disappointment should a treatment fail and a sense of frustration from the clinician who feels they explained the risks prior to treatment. Reprints Share. Related Articles Choose your words carefully. Choose your words carefully with patients to reduce liability, improve safety. Choose your temporary staffing agency carefully. Related Products Choose your words carefully. Choose your words carefully. Choose your new case managers carefully Single Article.
The accuracy of our predictions is probably overestimated because the prognosis in patients for whom predictions were recorded was probably more straightforward than that in patients for whom predictions were not recorded.
However, the group's median survival of 12 months is almost identical to that of complete cohorts of our patients with incurable cancer from to Milsted et al , ; Chye et al , Our results are probably best considered to reflect a group of newly referred patients with incurable cancer for whom oncologists were willing to record estimates of prognosis.
Most previous studies of prognostication in incurable cancer have been in people with far-advanced disease being referred for end-of-life care, not in people recently diagnosed and being referred to medical oncologists for consideration of anticancer treatment. These studies have shown that doctors' predictions were inaccurate, with a tendency to overestimate life expectancy Vigano et al , , a , b ; Christakis and Lamont, People in these previous studies were being admitted to hospices or hospice programmes and most died within a few weeks or months.
The lower accuracy and tendency to overestimate life expectancy in Christakis and Lamont's study probably reflects their population's shorter survival, and the large number of generalist physicians less familiar with advanced cancer. The distribution of our group's actual survival times was skewed to the right towards longer times , as are most survival distributions.
This is because the minimum survival time can be no shorter than 0, whereas the maximum survival time can be many years. The same constraints should apply to estimates for an individual. Someone with a predicted survival of 6 months can die no sooner than immediately, but may live for several years.
This suggests that if ranges are to be estimated around a predicted survival, then they should also be asymmetrical — the interval above the predicted survival should be larger than the interval below it. The good fit of an exponential model was fortuitous the first and only fit we tried and surprising because our population included a mixture of types and extents of advanced cancer with different expected survival durations.
We are not suggesting that the survival distributions of all groups of cancer patients are exactly exponential. More homogeneous groups should have survival curves that are sigmoidal steeper in the middle, flatter at the beginning and end ; more heterogeneous groups should have survival curves better approximated by a declining exponential steeper at the beginning and flatter at the end.
However, keeping the exponential shape in mind is helpful in thinking and talking about predictions of life expectancy, even if it does not provide an exact fit. The median survival is the time taken for a group to be halved half still alive, half already dead , and in an exponential distribution, this time is constant along the whole curve and analogous to the half-life of radioactive decay. These observations have important implications for how we might think and talk about predicted life expectancy.
Firstly, the predictions were well-calibrated, so predicting the median survival of a group of similar patients seems a reasonable starting point. Secondly, predictions were imprecise and probabilistic, so it is probably better to think and talk about ranges e.
Thirdly, survival times are skewed to the right towards longer times , so ranges around any point estimate for example the predicted median survival should be asymmetrical with wider intervals above than below. Fourthly, it is helpful to think of median survivals as half-lives and to use simple multiples of the predicted median survival e. We suggest it may be better to think and talk about ranges based on an exponential model.
Box 1 outlines the suggested steps for predicting life expectancy in people with advanced cancer using this approach. We recommend deliberately leaving estimates rough to accurately convey their inherent imprecision.
Before discussing estimates of life expectancy with an individual, it is important to determine what kind of information they want. Do they want any information at all, and if so, would they prefer orders of magnitude e. Box 2 gives examples of how estimates of life expectancy might be discussed and explained, depending on the patient's information preferences. Our data do not indicate how to improve the accuracy of individual predictions. There was a strong correlation between oncologists' predictions and their patients' actual survival times.
However, over half the variation in patients' survival times remained unexplained. The prognostic importance of performance status and quality of life are well documented in advanced cancer Stockler et al , ; Chow et al , Symptoms and signs of advanced cancer, nutritional status, and laboratory tests have also been identified as important Maltoni and Amadori, At the time of initial referral to a medical oncologist, other factors may also be important, such as disease tempo, response to previous treatments, co-morbidities, and planned future treatments.
Better understanding of these factors and their significance should help doctors refine and improve the accuracy of their predictions. However, it may be that life expectancy, like many other complex phenomena, is inherently unpredictable and the best we can do is improve our appreciation and communication of this uncertainty. The chances for improvement are small. A doctor may use this term when he or she is unsure if the patient will be able to recover.
Many people think of a guarded prognosis as another way to say that the patient is in poor or serious condition. Meanwhile, an individual with a guarded prognosis may have stable vital signs right now, but their condition could worsen or improve. Other Types of Prognoses. A good prognosis implies the patient will improve successfully.
They will encounter few setbacks along the way. A serious prognosis is less serious than a guarded prognosis. The individual may recover, but there is still a chance for other problems. They could face an infection that could become worse, for instance.
A patient with a critical prognosis is more serious than a patient with a guarded prognosis.
0コメント