Balancing the benefits and harms of care

BMJ 2004; 329 doi: (Published 01 July 2004) Cite this as: BMJ 2004;329:59
  1. Anna Donald, honorary fellow in epidemiology
  1. University College London

    Can patients have too much information about the benefits and harms of treatment? Until a year ago, I wasn't sure of the answer to this. Why worry people needlessly about outcomes that are rarer than hen's teeth? And why burden junior staff with lengthy consent forms?

    Having just spent the past 12 months being treated for breast cancer, I am now convinced that people need as much good quality information as they can get. Why?

    Firstly, I wanted to optimise my prognosis by making good decisions. Less obviously, I wanted to be able to comfort myself post hoc that I had made the best decisions possible at the time. (It would be cold comfort were my family and I later to discover reliable research favouring different decisions than was available last year.)

    Secondly, I needed to recognise the effects of treatment, both good and bad, when they occurred. For example, being aware that neutropenia below about 0.5x109/l carried a small but real risk of sepsis made it easier for my partner and me to get to an emergency department when it happened and to reconcile ourselves to curtailing our holiday in Scotland. By contrast, without good information, I spent about six weeks needlessly worrying about hip pain that I feared might be metastatic, before discovering it to be a lesser known effect of one of the drugs I was taking.

    I spent about six weeks needlessly worrying about hip pain

    Thirdly, I needed reliable treatment information in order to come to terms with the reality of my condition. Knowing the likely effects of chemotherapy and radiotherapy helped manage practical matters, such as work arrangements and family visits from Australia. In addition, being able to grasp that I really did have cancer (“Me? A 30-something doctor, and an evidence based one at that!”) became something of a milestone of the disease process—one achievement, at least, from a topsy-turvy year.

    Finally, good information helped me forge more open relationships with health professionals, family, and friends. For example, having reliable prognostic data helped us speak openly about the (happily small) chance of dying. This in turn made speaking about everything else easier.

    So what kind of treatment information seemed “good”? Firstly, it had to be in “bite sized” pieces, repeated many times and available on the days when I found myself able to cope with it. This rarely coincided with doctors' appointments, so online information and printed brochures were good. Even better were the specialist nurses, whom I could telephone in moments of terror such as when I was convinced my headache was a brain metastasis.

    Information also had to be psychologically appropriate. Clinical texts stated all my worst fears just a bit too baldly. Some consumer texts were written too much from the author's experience to be of much use generally, or infused with too much moralistic outrage at doctors to be of much use to me. Some, though, like the DIPEx website (, provided valuable insights into people's experience with treatments, which helped flesh out statistical evidence.

    On that note, above all else, I needed treatment information that was specific and properly evidence based. My job had accustomed me to having reliable evidence for most health questions. But when the crunch came I surprised even myself with my need for it. By evidence based, I mean information derived from a rigorous method of screening all available research for validity to my question. I found that positive answers were helpful (such as antioestrogen drugs are likely to work), but so too were negative ones (we don't yet know the effects of chemotherapy on cognition) as they enabled me to make honest decisions and to take into account other, more certain considerations.

    I found the widespread, magical thinking that favoured an exclusive diet of crystals, celery juice, and the like to be verging on immoral. By contrast, my own NHS breast care nurse was outstanding in her ability to discuss the pros and cons of treatment in a psychologically sound way.

    Perhaps my need for evidence—including certainty about uncertainty—partly reflected a need for something trustworthy when so much that I had taken for granted was being turned upside down. I found reliable evidence in four places, beyond which I didn't need anything else. The first three—Clinical Evidence, Best Treatments, and the Cochrane Library—are all widely available (and, no, I didn't get paid extra to write this).

    The fourth evidence source was my colleague's summary of the evidence, tailored for me. His “Bottom Lines for Early Breast Cancer (Premenopausal)” saved me a lot of time and made me trust my doctors faster. It enabled me to undergo chemotherapy without too much fuss and to begin the long process of coming to terms with the possible consequences, such as fertility ablation. But then, as my oncologist helpfully pointed out, death is terrible for fertility.


    • Competing interests I am a director of Bazian, a company specialising in evidence based health care.

    View Abstract

    Sign in

    Log in through your institution