Impacts on clinical decision making - changing hormone therapy management after the WHI.

Background: Medical news often receives intense, but distorted, media coverage, which can lead to high levels of insecurity in both patients and doctors.

Objective: To elicit general practitioners' self reported behaviour regarding hormone therapy (HT) advice and prescribing, before, immediately after, and 2 years following the release of the first results of the combined oestrogen and progesterone arm of the Women's Health Initiative (WHI) study; to elicit GPs' understanding of statistical risk terminology; and to explore their personal preferences relative to the trade offs between quality and length of life in medical treatment.

Method: In October 2004, a questionnaire was sent to all 169 GPs working on the central coast of New South Wales.

Results: The response rate was 67.5%. Before the release of the WHI study, 43.8% of GPs recommended HT; 5.9% did so immediately after, and 1.8% 2 years later. When expressed as number needed to treat (NNT), 20.8% of GPs stated that they were unable to interpret the absolute risk of HT use. Half of the 84 GPs who stated that they understood the concept of NNT were not going to reconsider the advice to give HT. General practitioners with a personal preference toward length of life over quality of life proved to be significantly more likely to advise against HT use (p=0.008 in a group comparison).

Conclusion: The sensationalising of the disease specific mortality differences in HT users had a dramatic and lasting effect on GPs' attitudes to, and prescribing of, HT. General practitioners acknowledged their poor understanding of basic statistical risk terminology. Providing absolute risk terms did not alter clinical decision making in 50% of GPs, clinical decision making may well be more powerfully influenced by a doctor's personal preferences relative to the trade off between quality and quantity of life.