Intended for healthcare professionals


Personal knowledge

BMJ 2006; 332 doi: (Published 19 January 2006) Cite this as: BMJ 2006;332:129
  1. Kieran Sweeney (kieran.sweeney{at}, honorary clinical senior lecturer in general practice
  1. Peninsula Medical School, Exeter EX2 5DW

    Doctors are much more than simple conduits for clinical evidence

    Choudry and colleagues' paper in this issue (p 141) is a brave attempt to quantify the under-recognised notion of personal knowledge in clinical practice.12 They carefully analysed the prescribing decisions of hospital doctors caring for patients with atrial fibrillation, before and after their exposure to a patient with an adverse event—either serious haemorrhage when taking warfarin, or a thromboembolic event while not taking warfarin. The researchers wanted to know whether doctors' knowledge of a previous adverse event affected their subsequent prescribing.

    For the group of doctors who were exposed to patients with adverse bleeding events, and who cared for atrial fibrillation patients subsequently, the odds that they would prescribe warfarin were 21% lower for subsequent patients. There was no statistically significant change in warfarin prescribing after a doctor cared for a patient who had had a stroke while not taking warfarin, nor—in support of the specificity of the study's findings—was there any change in prescribing of angiotensin converting enzyme inhibitors by doctors with exposure to patients who had either bleeding events or strokes. Doctors' experiences of bleeding events associated with warfarin can influence warfarin use, the team concluded, and adverse events associated with the underuse of warfarin may not affect subsequent prescribing.

    Doctors are neither passive recipients of, nor simple conduits for, clinical evidence.3 We conduct an “inner consultation” with evidence, analysing it in both a logical and intuitive way.4 In so doing, we are exposed to what Tvensky and Kahneman call the “availability heuristic”—a fancy way of saying that we are more likely to recall events which are more easily recalled—and the “chagrin factor,” whereby doctors tend to avoid actions that cause them hassle.56 Patients conduct similar internal conversations, adding the experience of a consultation to their previous intellectual and emotional understanding of illness.

    Currently two compass points guide these consultations: statistical significance and clinical significance. While necessary, these are not enough to clarify the dynamic interaction between a patient and a doctor. A third dimension is personal significance, a concept that captures the reciprocity of the evaluation and interpretation of a new idea by a doctor and patient together.7

    All of us, including doctors, have our own “view from somewhere.”8 Scientists also develop individual perspectives on what they choose to look at, but are rather more able to reconcile particularities, through the conventions of intersubjective agreement, to create what they call “an objective account.” This has the air, however, of a “view from nowhere.”9 At stake here is something quite profound, and poorly accepted within the medical community: the personal participation of the knower in all acts of understanding. This is the central thesis of Polanyi's great work, echoed in the title to this editorial.1 The philosopher Polanyi, a professor of physical chemistry and then of social studies, argues that comprehension is neither an arbitrary nor passive act and requires tacit skills of judgment. These skills are neither algorithmic nor exhaustively specifiable, but underpin the connection between passion and knowing.

    Intellectual passions have an affirmative content. In science, they help us decide what to explore, selecting some options over others. Heuristic passions help us discover things, urging us at times not to accept at face value what is given or conventionally accepted. And persuasive passion turns both of those into controversial debate, by which scientific communities define themselves. In medical consultations there are two participants, both personally knowing, both passionately participating, but from different perspectives, different “somewheres.” The outcome of their interaction, in the form of the clinical decision, is an emergent property of two ways of knowing: biomedical and biographical. This forces doctors to confront an intersection of the two epistemologies and to ask questions about their respective legitimacy (is one inherently superior?) and inter-relationship (on which occasions should one dominate?).

    Chaudry and colleagues' paper illuminates this murky area for us and provides convincing evidence that, within each doctor, these two ways of knowing compete for influence. When the authors remark, at the end of their paper, that “one would hope that providers who prescribe more frequently, and who are specialists, would be least likely to be influenced by experiences with individual patients” one senses how much we all have to learn about personal knowing.


    Research p 141


    • Competing interests None declared.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    View Abstract