Are you looking at some kind of joint study with veterans to be able to understand that bigger continuum? It's quite possible, I would presume, that some of the conditions that might have led to the suicide might have actually occurred while they were in uniform. Even if the outcome wasn't recognized while they remained in uniform, it might still inform and would also perhaps give us a different perspective on the statistical relevance of the sample size. Would that be something you have been considering?
On February 23rd, 2017. See this statement in context.