Welcome to the century of the patient

I listened to this podcast with great interest. I have been
practicing and teaching medicine for many years and always tried to make
my diagnoses and decisions 'transparent' to allow patients, students,
nurses and other clinical colleagues to understand what I am thinking and
to join in. By 'transparent' I mean that all actions are linked
explicitly to a possible diagnosis and that each possible diagnosis is
linked explicitly to the findings used to support it. I have written a
textbook called the Oxford Handbook of Clinical Diagnosis with the help of
some of my past junior staff that describes how to reason 'transparently'
in this way. The problem is that such transparency is not fashionable and
students are not expected to become transparent in their reasoning. This
is a serious because patients often participate in decisions by secretly
rejecting those that they do not understand e.g. by not taking the
prescribed medication and hiding this from the doctor so as not to cause
awkwardness.

The reasoning used to combine evidence from the literature with
evidence obtained from the patient in 'evidence based medicine' is often
far from transparent. This will be an impediment for those patients who
do wish to participate in 'evidence-based decisions'. Furthermore, no
evidence is available on many important aspects of medical decisions.
This includes the evidence for most appropriate diagnostic and treatment
selection criteria, evidence for the best diagnostic leads with the
shortest differential diagnoses or evidence for the best discriminators
between such differential diagnoses. Instead we are offered likelihood
ratios and specificities designed for assessing screening tests, but which
are inappropriate for differential diagnosis, diagnostic confirmation and
treatment selection. We also have to cope with the results of therapeutic
trials on patients that differ considerably from those we see in day to
day care.

When patients are unable to participate in decisions about their care
it is not usually because doctors and nurses are preventing them from
doing so, but because the basic facts, opinions and decisions and how
these three things are linked in a transparent way is inaccessible to
them. These are not the 'general' facts in the published literature, but
the 'particular' facts and related opinions expressed about a 'particular'
patient. These 'particular' facts (symptoms, signs and test results),
opinions (new, old and existing diagnoses) and decisions (recommendations
to start, stop or continue treatments, do tests, or agree to follow up
appointments) are not linked transparently but usually scattered
throughout voluminous paper or electronic records or locked up in a
clinician's head.

Before patients, students, nurses, doctors and other health
professionals can participate in medical decisions properly they must be
aware of all the current decisions (or the options), the diagnosis or
differential diagnosis on which each of these decisions or options is
based and the findings that support each diagnosis. They cannot get this
information from the internet but only from the clinician who used the
patient's own findings to arrive at the diagnosis that suggested the
options. It is only when patients have this basic information that they
can go to the books or the internet to find background information to see
if they agree with the decision options and which decision might suit them
best (which includes doing nothing). Without this information they will
simply flounder around. It is not enough to focus on one decision; many
adverse effects of any one decision may well be influenced by other
coincidental diagnoses, causes, complications, and treatments and their
urgencies and probabilities.

I am not convinced that doctors and nurses do not understand
uncertainties. The problem is that those onlookers of non-transparent
decisions may be confusing 'decisiveness' with certainty about an outcome.
One can be decisive about what has to be done and still have a low
expectation of success. In other words, one has to be decisive (after
suitable reflection) but have a sensible degree of certainty in the
outcome. Unfortunately I think that the mathematics of probability as
currently taught is unable to explain or model adequately many of the
thought processes used in day to day clinical practice. It is like trying
to use plane geometry based on a 'flat earth model' in discussions with
expert intuitive circumnavigators of the globe. The inevitable result is
mutual incomprehension. There is a great need for a meeting of minds
between clinicians and mathematicians to sort this out. It is not only
evidence based decisions that would benefit but proper assessment and more
efficient use of expensive technology.

I was very moved by a discussion in the third part of the podcast.
Here was a doctor describing that, as a patient, how difficult it was to
navigate her way though a health care system by having to deal with a
number of doctors who were each focussing on issues related to their own
speciality alone. A GP sympathised and suggested that solving this
problem was a huge challenge. Something similar happens day in day out
when a patient is sent from home to A&E, from A&E to EAU, from EAU to a
medical ward and then from a medical ward to an outlying surgical ward and
then back home again.

This problem might be solved by creating a 'transparent' summarising
written rationale. This can be done by writing a 'Management plan' that
lists all the current decision options, current treatments, test results
planned and awaited and the next review (e.g. ward rounds and clinic
appointments). For each of these items in the management list, the
diagnosis (or differential diagnosis) is written down in another list and
numbered, with a space between each diagnosis. Under each diagnosis, its
outcome or progress marker(s) is written down and if available, how it
presented and how it was confirmed or shown to be probable. Finally the
number of the explaining diagnosis is placed in brackets after each item
in the management plan. This transparent rationale makes it clear what
the evidence is for each diagnosis and the management of each diagnosis.
Such a basic, simple transparent rationale would empower patients to
participate in decisions about their care (i.e. if they wish) by providing
them with the basic information to get started. At least, no one seems to
have objected to this aspect of the proposed NHS reforms.

This simple, transparent rationale could be hand written in pencil or
ink, or typed into a Word document and stored on a USB memory stick in a
suitably protected way. It can also be stored on a hospital or GP
computer and emailed securely to another doctor at the patient's request.
This can be done using a single computer or a NHS wide network. The
important thing is that a transparent explanation is always made available
to help the patient and any doctor or nurse to make sense of what is
happening and to 'sing from the same song-sheet'. The same information
can be copied into a ward sister's handover spreadsheet or to a house
officer's 'to do' list. This 'transparent' summary is no more than what
every competent clinician should hold in his or her head and pass on. The
only difference is that it is also written down.

With such basic transparent information, patients will have the
option of participating in decisions about their care and when
appropriate, be supported by decision aids. Without such a basic
understanding, they will not be able to do so and the Salzburg declaration
(and this aspect of the NHS reforms) will get forgotten.

Reference

1. Llewelyn H, Ang HA, Lewis k, Al-Abdullah A. The Oxford Handbook of
Clinical Diagnosis, 2nd ed., Oxford University Press, Oxford, 2009