I sympathize with her as I see others here do. We are limited with what we could do.
Please bear in mind that my comments are not based on a real and thorough knowledge of what she testified and the Holmes case. I am trying to speak in generalities.
Resnick in forensic conferences emphasizes that "a building crescendo of paranoia" in and of itself is grounds to have someone held in a psych facility for further evaluation and cites several cases where people were new in their first-break psychosis, had a building crescendo of paranoia, and the clinician discharged the patient only to have that patient kill someone within the next few days.
He used the teaching case of a man whose records he evaluated from Youngstown, Ohio who had no prior mental illness, apparently was in first-break, thought people were out to get him, showed enough sx of psychosis for a confident dx but the clinician discharged him because his wife vouched for his safety and the clinician prescribed Stellazine.
Within the next day, the couple made love, he woke up and now thought the wife was part of the conspiracy against him and he killed her. He also mentioned several other cases where a clear identifiable crescendo of paranoia had built up even with someone without a history of violence and they committed a severe act of violence while psychotic.
Based on what Resnick stated I have placed people in the hospital when detecting this building crescendo of paranoia and have often times quoted him in court when these cases graduated from the hold phase to the involuntary commitment phase.
Studies show that paranoia is a high predictor for future violence. A problem here is that even good, non-antisocial, people will attack others if they believe it is under certain conditions such as self-defense and in the case of paranoia the person believes they are being threatened by someone else.
But this said, forensic training is a very specialized segment within our field. I too have had cases where there were bad outcomes and like this doctor, could not predict the future. We discharge patients all the time that made a suicidal or homicidal comment that later don't commit a dangerous act.
IMHO suspected psychosis and homicidal thinking is enough for a hold and from there it should be considered. A building crescendo of paranoia too should be in and of itself grounds for at least a hold. The risk is hurting the therapeutic alliance.
Now all of this said, am I attacking Fenton? No. 1-As mentioned I haven't read enough of this case for real and the media rarely gives balanced presentations. 2-In her defense, if she had another psychiatrist review the case she is at a level where she is above the standard of care, so in a malpractice sense, unless someone could point out an egregious error I doubt she committed that.
and 3-There is that risk of hurting the therapeutic alliance and I'm sure all of us (maybe not you first years) discharged people where you felt there was some risk though not significant.
In a case where I was stuck and didn't know what to do I would either have another doctor evaluate the patient for a second opinion and/or perform an HCR-20. The latter is a test for evaluating the risk of future violence and using it would put someone above the standard of care.