We thank you for your reflections and comments on our paper. We understand your concerns but nevertheless believe that although the issues you highlight are important, they do not alter our overall conclusions.
The aim of the safety planning-type interventions as mentioned in this meta-analysis is to prevent suicide attempts (suicidal-self-harm; SSH) and not non-suicidal-self-harm (NSSH) specifically. However, we are aware that NSSH is an important risk factor for suicide attempts.
Nonetheless, you have identified an important challenge in the field, namely how best to operationalise suicidal behaviour as an outcome measure. It is true that not all studies included in our meta-analysis provide a clear and standardised definition of suicidal behaviour or attempted suicide, and it is possible that the exact definition of suicidal behaviour may differ between studies in terms of suicidal intent. However, each study's definition of suicidal behaviour has been applied to its treatment group and its control group and, therefore, will not have influenced the effect size. Of course, across the studies some instances of suicidal behaviour may have been missed. Therefore, we would urge the field to make further progress in agreeing standard nomenclature of suicidal behaviours to be used in such trials.
Notwithstanding the caveats noted above, therefore, we believe that suicidal behaviour, defined in this meta-analysis as the combined rate of suicide attempts and fatal suicides, is an appropriate outcome in order to draw conclusions about the effects of safety planning-type interventions, which were designed to reduce suicide attempts and suicides.
Conflict of interest
None declared
eLetters
No eLetters have been published for this article.