Testing hypotheses

BMJ 2004; 329 doi: (Published 30 September 2004) Cite this as: BMJ 2004;329:773
  1. Steve Iliffe, senior partner (s.iliffe{at}
  1. Lonsdale Medical Centre, London

    Medicine is the natural home of the untested hypothesis, says Hugh Pennington while wondering why doctors are so unscientific.1 When the pain started to go down my left arm, one day some four years ago, I considered my family history of heart disease and came to the obvious conclusion. So the next day I cycled from University College London to the Royal Free Hospital, up Hampstead Hill, to my relief without a twinge. With the angina hypothesis disproved, I could safely ignore the pain, which was conveniently intermittent if unpredictable, and get on with a busy job.

    Digging up a tree root changed everything. The pain worsened, but subsided with the cocktail of paracetamol and ibuprofen that I recommend to so many patients, so I could carry on hacking and heaving. Carrying heavy bags on a holiday journey further tested my second hypothesis—muscle sprain—and I was able to experience the peculiar apprehension that comes with being ill away from home. Then the escalating pain, interrupted sleep, and diminished power in the left arm demolished the muscle hypothesis and replaced it with a neurological one. A colleague in neurology diagnosed a cervical disc prolapse, confirmed by magnetic resonance imaging.

    I was soon back on my bike, although I gave tree roots a wide berth, with just the occasional dart of pain and, more often, odd tingles in the hand or forearm. Thus, when the arm throbbed at the start of a tennis match just before Easter I put it down to the neck problem and carried on. The feeling disappeared, only to recur the next day as I was cycling up a relatively gentle incline, but it stopped when I reached level ground and did not recur when I cycled home. The dodgy disc hypothesis held, and I dug out some analgesia in case I needed it.

    That night I could not sleep for the pain in my arm. Paracetamol did not touch the pain, and, as it spread across the chest and I began to feel nauseated, I finally realised that the disc was innocent. The emergency services moved quickly, and the hospital departments dealt with the incident with great skill. Undergoing an emergency coronary artery bypass graft is an education, particularly if you have no major risk factors for heart disease (the “family history” now being attributed to smoking rather than genetics).

    Back on my bike again and even eyeing a tree root in need of extraction, I understand that medicine is the natural home of hypothesis testing, but that we do not test hypotheses under circumstances of our own choosing but according to our hopes, fears, and competing demands.


    1. 1.
    View Abstract