Does citing early SIDS research skew contemporary conclusions?

Research on infants who have allegedly succumbed to sudden infant death syndrome (SIDS) has been of variable quality over the years. Even now peer-reviewed papers are being published on cases termed 'SIDS' without autopsies having been performed, despite this being a requirement of the three major definitions for over five decades. Clearly cases used in earlier research studies could not have complied with the requirements of as-yet unpublished definitions/guidelines. For this reason care must be taken in citing initial papers as their results may have been skewed by the presence of non-SIDS cases. This may have particular relevance for meta analyses. Reviewing the literature on substance P and its relationship to SIDS provides an excellent example of how diametrically opposed conclusions were reached at different time points. Early studies on SIDS, and studies that use cases that were classified before the standard NICHD and San Diego definitions, should, therefore, be approached with a degree of scepticism and not cited in contemporary papers or at meetings as they have the potential to confuse rather than clarify.