Despite the high numbers involved, the internal validity is questionable given that, as the authors stated, the extent to which patients complied with the self-test instructions was not known. Moreover, a high proportion of those patients eligible for screening (11.4%) declined which may have affected the results. This was not discussed by the authors. The authors noted that it was also debatable whether the subpopulation finally screened represented a 'high-risk' group. Whilst no dates associated with the study were provided, the authors mentioned a 'current' debate about whether earlier detection of non-insulin dependant diabetes mellitus would imply better health outcomes in terms of reduced risk of complications.