Although I don't have much personal experience prescribing/working with prazosin myself, I have always considered it to be a first line treatment for individuals with PTSD. In my clinical rotations so far I've seen prazosin used to treat nightmares and symptoms of PTSD, and it seems as though it is effective. Like many, therefore, I was surprised to read this article and see that in this particular study, prazosin was not an effective treatment for nightmares associated with PTSD.
There are a few flaws of the study that make me questions the generalizability of the results. The exclusion criteria seem to have been quite strict (no psychotic or cognitive disorders, no substance dependence within the last 3 months, no cocaine or stimulant use, no active SI/HI, to psychosocial instability). It is always challenging to balance the need to limit potential confounding variables while maintaining a realistic eligibility criteria. It seems, though, that the exclusion criteria, especially the vague "psychosocial instability" exclusion criterion may explain why the results of this study contradict what we all seemed to know about prazosin and PTSD.
At this point there is still a thorough discussion of the benefits of using prazosin for PTSD nightmares on UpToDate that cites multiple studies showing its efficacy (https://www.uptodate.com/contents/nightmares-and-nightmare-disorder-in-adults?sectionName=Prazosin&anchor=H4261530858&source=see_link#H4261530858).
I think that given clinical experience (both mine and my colleagues), as well as the current recommendation on UpToDate, I will continue to view prazosin as a tool to use when treating individuals with PTSD and nightmares. I plan to keep this article in the back of my mind for times when prazosin does not work, however at this point I don't see myself changing my views/practice based on this article.