Article Text

PDF

Controls not matched
  1. BARRY J MARSHALL,
  2. GRACE Y HO
  1. H pylori Research Laboratory
  2. Department of Microbiology
  3. University of Western Australia
  4. Australia
  5. bmarshall{at}hpylori.com.au

    Statistics from Altmetric.com

    Editor,—The paper by Kerr et al reported an association between H pylori and sudden infant death syndrome (SIDS). We have reviewed their data and believe that the methods used may have led to incorrect conclusions.

    Kerr et al examined retrospective material from 32 cases of SIDS infants and 8 non-SIDS controls. They used nested PCR followed by an ELISA detection step which would have made their method exquisitely sensitive. Consistent with this, no other method was able to confirm that H pylori was actually present. Instead, Kerr et al used a relative increase of “H pylori signal” above that of the mean +2SD for a control group, as an indicator ofH pylori presence. This prompted us to more carefully consider the appropriateness of their control and patient groups.

    Since ethnicity and socioeconomic details of the SIDS infants were not given, we could not confirm that these matched the control infants. We also noted important clinical details of the controls which could make them inappropriate. It appears that most of the controls would have had very little bacterial contamination of the PCR specimens because they died in hospital while on antibiotic therapy for sepsis, or were deceased very soon after premature birth. In addition, they might have been transferred to refrigeration very soon after death. SIDS infants however, probably died at home, many hours before being refrigerated.

    Finally, as H pylori is a gastric organism, it was surprising to find the bacterium in lung or trachea of eight patients (ureC gene) or six patients (cagA gene) in whom gastric specimens were negative.

    Since Kerr's paper was widely reported in the media, we believe that it needs to be stated that the case for H pylori as a cause of SIDS is certainly unproven and is in quite considerable doubt.

    View Abstract

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.