Ten steps towards improving prognosis research
BMJ 2009; 339 doi: https://doi.org/10.1136/bmj.b4184 (Published 31 December 2009) Cite this as: BMJ 2009;339:b4184All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
“The tide of low quality, low impact prognosis research” (1) is an
adept description which could be applied to many areas of clinical
research. Recently while performing a literature search on a contentious
surgical issue results could only be described as a mile wide and an inch
deep. But what causes such a plethora of poor quality research? The answer
is simple: publications are fundamental for an individual to advance in
medicine. Application forms ask for a list of publications without concern
for quality. Curriculum Vitae are assessed on publications. This attitude
is reinforced throughout medical training. Is it any surprise that there
is a large volume of poor quality studies when most trainees are expected
to publish and present? Inevitably this leads to excellent trainees who
are passionate about treating patients but less passionate about
performing research being forced to publish and “inflate the importance of
[their own] research”(1).
Undoubtedly trainees learn important lessons about critically
appraising research by performing their own. This skill is imperative to
all trainees. But is it vital for trainees to be able to design and
complete a high quality research project? Let us not get confused over the
need for trainees to be able to design and complete good quality audits.
Unhelpfully for trainees audit and research are often grouped together.
Audit is fundamental for all clinicians to continually improve local
practice. However I would suggest that the royal colleges take a new
approach to teaching trainees about research. Instead of learning about
research ad hoc, often relaying bad habits, there should be validated
paths open to all trainees. A well structured course teaching the
principles and practice of research climaxing in a supervised projected
would be an obvious choice. This means that trainees who are not
gifted in producing high quality research can “tick the research box”
without feeling pressure to publish. In addition trainees who are
interested in future research could be taught the importance of integrity
and quality.
Perhaps the recent BMJ title cover “Getting them while their young”
would be a good motto as we try to improve the quality of research.
1. Hemingway H., Riley R.D., Altman D. G.; Ten steps towards
improving prognosis research BMJ 2009; 339: b4184
Competing interests:
None declared
Competing interests: No competing interests
Many thanks to Hemingway et al for this long needed paper although I regret that in the current online version, part
of the Purpose section was apparently truncated.
The authors are sadly to the point when they mention that
prognosis studies are often improvised as the “cherry
on the cake” of a work designed for quite another
purpose. This is true of marker studies that often use
available biological material, literally “in the
freezer”, and of multivariate risk prediction models
which are often a kind of fishing expedition across all
available data items, no matter the nature of their
putative relationship with the outcome.
Another point, that the authors did not raise, is the need
for appropriate statistical methods. Published prognosis
results (I dare not say studies) are too often based on
statistical inference methods, using the “p value” as a measure of predictive value,
although a full range of statistical methods are available
for predictive analysis. This is likely a consequence of a
lack of planning and thus thinking about these analyses.
Last, as somewhat indirectly pointed by the authors, in
medical matters, prognosis is not an aim per se.
The clinical usefulness of a marker or a combination of
thereof should be considered at design time. Should this
marker be used for therapeutic decisions? Might it be used
as a proxy outcome in clinical trials? Such questions are
central to medical prognosis research.
Competing interests:
None declared
Competing interests: No competing interests
Waiting for an accurate prognosis at the end of life may risk missing the boat.
“Prediction is very difficult, especially about the future”
Niels Bohr, Physicist
Hemingway et al conclude that much prognostic research is poor, and
that systematic reviews can reach only limited conclusions because of
various methodological problems.1 But there is another reason:
prognostication in itself may be very difficult if not impossible at
individual patient level. Two systematic reviews have reported that
prognostication in people with advanced non-malignant illness is virtually
impossible at around 12 months before death.2,3 If prognosis is held as a
vital piece of information necessary for action, waiting for an accurate
prognosis may cause inappropriate delay. “Prognostic paralysis” can
prevent care planning.4
UK Governments and indeed the BMJ is advocating that palliative care
should be available not just for cancer, but for all those in the last
year of life. This is progress. But prognostication especially in people
with multi-morbidity is challenging even if statistically valid tools that
predict mortality are available. In prognostic uncertainty in the last
phase of life we must assess patients’ palliative care needs and treat
accordingly.
The “Surprise question”, can be used to identify potential candidates
for a supportive and palliative care approach : “Would I be surprised if
this patient were to die in the next six to 12 months?” 4 But even this
question is asking clinicians to identify patients who are “sick enough to
die”, and hence need support, rather than to make an estimate of when the
patient will die.
Scott A Murray
Kirsty Boyd
1. Hemingway H, Riley RD, Altman DG. Ten steps towards improving
prognosis research. BMJ 2009;339:b4184
2. Coventry PA, Grande GE, Richards DA, Todd CJ. Prediction of
appropriate timing of palliative care for older adults with non-malignant
life-threatening disease: a systematic review. Age Ageing 2005;34:218-27.
3. Glare PA, Sinclair CT. Palliative medicine review:
prognostication. J Palliat Med 2008;11:84-103.
4. Murray SA, Boyd K, and Sheikh A. Palliative care in chronic
illnesses: we need to move from prognostic paralysis to active total care.
BMJ 2005. 330:611-12.
Competing interests:
None declared
Competing interests: No competing interests