I was treating John W., a psychiatric patient who was an insurance executive. He was angry about his work situation. He complained bitterly about his supervisor whom he felt singled him out for unfair criticism. Over the course of months, John’s complaints escalated to the point where he said, “I swear … if there was a chance I’d get away with it, I’d love to get rid of the guy.”

I questioned him about this statement. While he acknowledged having a fantasy of “shooting” the supervisor (John didn’t own a weapon), he denied any intent to do any harm to his perceived harasser. He insisted his wish to “get rid of the guy” was merely a figure of speech.

As his treating psychiatrist, I had to make a judgment call about John’s state of mind and his possible intentions.

It’s a well-entrenched principle that when a patient presents an imminent danger to someone else, the therapist has a legal duty to report this threat to the proper authorities. The obligation derives from the now-famous 1976 Tarasoff ruling in California. In that case, the court decided the patient’s privacy and right to confidentiality ends where the public good is impacted in a negative way. A psychiatrist must report a potential threat of harm to another person if violence is considered imminent. Most therapists understand this duty and abide by that common sense legal decision.

But some states have passed strict legislation requiring a therapist to report a patient if it’s likely serious harm may befall someone else. Likely is a far weaker criterion than imminent. The term “likely” is equivalent to “probable” in the law, which means a 51 percent or greater chance of something happening. “Imminent”, on the other hand, means something is about to happen.

And, here lies certain danger.

As a psychiatrist, I treat a variety of people, some of whom are quite disturbed. Others are highly functioning, but have emotional problems. I worry about the possibility of over-reporting patients who might be considered likely to cause harm. The notion of something being likely requires a judgment about anticipating someone’s future actions.

Dr. Richard Milone, chair of the American Psychiatric Association’s Ethics Committee, said, “Your primary job is to be the patient’s doctor. Any time you start being something else to a patient, it’s the start of trouble.” In the wake of the Newtown, Connecticut school shootings, Dr. Milone suggested psychiatrists are now being asked to treat their patients, and also protect society. Or, as another APA Ethics Committee member said, “Society expects us to be their protectors and to work as double agents.”

Such legislation undermines patient confidentiality. Under such laws, can patients truly trust their therapists? Can they really speak their minds without worrying they may cross some subjective line drawn in the therapist’s mind whereby a patient is deemed likely to cause harm and be reported? Yes, a severely delusional person, who believes others are conspiring against him and intends to get a rifle, is someone who should be reported. But an angry, disgruntled man like John W. is apparently using a figure of speech. Should I report him to the authorities because it could be likely he’ll take action?

Can a trusting, confidential therapeutic alliance between psychiatrist and patient ever develop given the requirement that therapists evaluate, predict, and then report patients’ likely behavior?

During a session, there’s no going “off the record.” Everything in therapy is grist for the mill. And now, everything is subject to being evaluated as presaging a patient’s likely behavior.

The recent rash of mass shootings brings important issues about mental health to the fore. But, in the process of trying to get a better handle on how to prevent such horrors, do we really want to turn our therapists into double agents?

Mark Rubinstein is a psychiatrist and author of Bedlam’s Door: True Tales of Madness and Hope.

Image credit: Shutterstock.com