Learning from negative findings

Mark I. TaraginEmail author View ORCID ID profile

Israel Journal of Health Policy Research20198:38

https://doi.org/10.1186/s13584-019-0309-5

Received: 18 January 2019
Accepted: 17 April 2019
Published: 1 May 2019

The original article was published in Israel Journal of Health Policy Research 2018 7:74

Abstract

A recent IJHPR article by Azulay et al. found no association between the patient activation measure (PAM) and adherence to colonoscopy after a positive fecal occult blood test result. This commentary will use that article as a jumping-off point to discuss why studies sometimes get negative results and how one should interpret such results. It will explore why the Azulay study had negative findings and describe what can be learnt from this study, despite the negative findings.

It is important to publish studies with negative findings to know which interventions do not have an effect, avoid publication bias, allow robust meta-analyses, and to encourage sub-analyses to generate new hypotheses.

To support these goals authors must submit articles with negative findings with sufficient detail to support the above aims and perform sub-analyses to identify additional relationships that merit study.

The commentary will discuss the importance of publishing articles in which the hypothesis is not proven and demonstrate how such articles should be written to maximize learning from their negative findings.