Intended for healthcare professionals

Letters

Secondary prevention in acute myocardial infarction

BMJ 1998; 317 doi: https://doi.org/10.1136/bmj.317.7166.1152a (Published 24 October 1998) Cite this as: BMJ 1998;317:1152

Data cited from two studies were inaccurate

  1. Niall A Herity (nherity@dial.pipex.com), Senior registrar in cardiology.
  1. Antrim Hospital, Antrim BT41 2RL
  2. 1500 E Medical Center Drive, B1F245, University Hospital, Ann Arbor, MI 48109, USA
  3. Department of Internal Medicine, 1500 E Medical Center Drive, TC 3119 N, Ann Arbor
  4. Wigan and Bolton Health Authority, Wigan WN1 1AH
  5. University of Wales College of Medicine, Centre for Applied Public Health Medicine, Cardiff CF1 3NW
  6. South West Cardiothoracic Centre, Department of Cardiology, Derriford Hospital, Plymouth PL6 8DH
  7. Cardiovascular Medicine Y, Bispebjerg Hospital, DK-2400 Copenhagen NV, Denmark
  8. Fish Foundation, PO Box 24, Tiverton, Devon EX16 4QQ

    EDITOR —Mehta and Eagle's review of secondary prevention in acute myocardial infarction contains some important errors.1 Firstly, in table 3 the relative risks of death cited for references 17, 18, and 20 are in fact percentage reductions in mortality divided by 100. These figures are so out of keeping with the rest of the table that it is surprising they were not spotted earlier. Even as percentage mortality reductions they are inaccurate. In the original paper by Pfeffer et al (ref 17) the 21% reduction observed referred to cardiovascular rather than total mortality, and the 95% confidence interval was 5% to 35% rather than to 25%. Pfeffer et al calculated a 19% reduction in total mortality (95% confidence interval 3% to 32%, P=0.019),2 although a crude calculation gives 17%.

    Secondly, the data cited from the paper by Ambrosioni et al (ref 20) refer to a combined end point (death or severe congestive heart failure) rather than to total mortality. The mortality statistics are 51/784 (6.5%) in the placebo group and 38/772 (4.9%) in the treatment group, so that the risk reduction is 25% −11% to 60%; P=0.19)3 and the relative risk 0.75.

    These are the only two original articles that I have checked. The errors are of fact rather than typography and raise questions about the remainder of the article.

    References

    Authors' reply

    1. Rajendra H Mehta, Cardiology fellow.,
    2. Kim A Eagle, Albion Walter Hewlett professor.
    1. Antrim Hospital, Antrim BT41 2RL
    2. 1500 E Medical Center Drive, B1F245, University Hospital, Ann Arbor, MI 48109, USA
    3. Department of Internal Medicine, 1500 E Medical Center Drive, TC 3119 N, Ann Arbor
    4. Wigan and Bolton Health Authority, Wigan WN1 1AH
    5. University of Wales College of Medicine, Centre for Applied Public Health Medicine, Cardiff CF1 3NW
    6. South West Cardiothoracic Centre, Department of Cardiology, Derriford Hospital, Plymouth PL6 8DH
    7. Cardiovascular Medicine Y, Bispebjerg Hospital, DK-2400 Copenhagen NV, Denmark
    8. Fish Foundation, PO Box 24, Tiverton, Devon EX16 4QQ

      EDITOR —As Herity points out, the relative risks of death cited in table 3 are the percentage reduction in mortality divided by 100. In a footnote to the table we stated that these were “cited as percentages in original paper.” A more appropriate footnote might have been “cited as percentage reduction in mortality.” This might have helped avoid the confusion created by our effort to simplify the tables. The same is true for table 4: the risk of death cited is the percentage reduction …

      View Full Text

      Log in

      Log in through your institution

      Subscribe

      * For online subscription