Concept on the use of "number needed to be exposed" in epidemiology / 中华流行病学杂志

ABSTRACT

<p><b>OBJECTIVE</b>To introduce the concept, methods for calculation and application of "number needs to be exposed (NNE)" in Epidemiological studies.</p><p><b>METHODS</b>Data was analyzed from a study on the association between diaspirin cross-linked hemoglobin (DCLHb) with 28-day mortality in patients with severe traumatic hemorrhagic shock.</p><p><b>RESULTS</b>The crude "number needed to be exposed for one additional person to be harmed" (NNEH) was 3.7 (95% CI: 2.2-11.5) for the exposure to DCLHb. After controlling the confounding bias of the baseline mortality risk, the adjusted NNEH became 2.6 (95% CI: 1.6-8.0) id., on average, among 2.6 patients exposed to DCLHb, one additional case of death would have developed within 28 days after initial hospitalization if the distribution of baseline mortality risk in exposed group had been equal to that in the unexposed group.</p><p><b>CONCLUSION</b>NNE could be expressed as the estimated average number of persons needed to be exposed for contributing (either developing or preventing for) one additional case of disease or death in a prospective study when compared with the unexposed persons. As a new index for measuring the absolute effect of an exposure, NNE presented the results on epidemiological studies in a more intuitive and understandable manner. Consequently, this method could be favorably accepted by clinicians, health policy makers and the public.</p>