Topics in Confidentiality and Duty to Warn

When HIPAA first came out, the big concern was that it would limit the release of psychiatric information and would place barriers to being able to talk to other members of a patient’s treatment team. But it turns out that HIPAA actually had the opposite effect in many ways. In exchange for a patient consenting to a broad privacy policy, providers can share information with other parties for treatment, payment, and healthcare operations purposes without specific informed consent for every release of information. For example, if a psychiatrist is covered by HIPAA and calls a patient’s primary care physician who is also covered by HIPAA, specific consent is not required for coordination of their treatment or for release of medical information.

What about communication with a therapist?

Many states have laws that are more restrictive than HIPAA in terms of strict psychotherapist/patient confidentiality, so that written informed consent would still be required for release of therapy records. Each state is slightly different, but the important take-home point is that HIPAA does not place additional barriers in most situations for access to psychiatric records. As a matter of practice, it is generally preferable to talk to your patient when you are able to about information that you are going to release and to obtain the patient’s consent to either release or obtain information. But it is also important to be aware that there are situations in which HIPAA allows you to share medical information without the patient’s consent.

What about confidentiality issues when it comes to patients who we judge to be dangerous? Any advice for clinicians?

Clinicians have a duty to protect potential victims of a violent patient, and this is commonly known as a Tarasoff duty after a California legal case. While states vary in their requirements, there are some basics that apply to most clinicians. When you are in a situation where you perceive that your patient is dangerous and has a specific plan to harm an identified victim, in most jurisdictions the psychiatrist has to take some kind of action to prevent something bad from happening.

That’s pretty straightforward. But in practice, patients will typically express vague ideation, perhaps saying something like, “If I ever see that guy again I don’t know what I will do,” as opposed to saying “Here is the name of my boss and I have a gun and I am planning to take the gun and shoot him tomorrow.” What is our duty in the more vague scenarios?

Those situations really require the clinician to be a clinician and to do a careful assessment of the patient for the risk factors the patient has for committing an act of violence and ways of minimizing that risk. Then ultimately, the clinician must make an informed clinical decision about whether it is something that needs to be acted on or not. If we went about breaching patient confidentiality every time a patient had some kind of urge to hurt somebody or made a vaguely homicidal statement during a therapy session, it would be untenable. Some states provide more specific guidance about psychiatrists’ responsibilities to protect third parties. In Massachusetts, a clinician has a duty to warn when either the patient has made an explicit threat towards a “reasonably identified” victim or has a known history of physical violence and the clinician has a reasonable basis to believe that there is a clear and present danger to a potential victim. So if a patient makes a vague threat but the clinician judges that there is no reasonable basis to believe that he or she will follow through, there is no duty to warn, at least in Massachusetts and many other states.

Once we decide that we have a duty to warn, are there laws regarding what precautions we are required to take?

There are a number of different ways that clinicians could discharge the Tarasoff obligation. One way would be to communicate this threat to the potential victim. Another would be to notify a law enforcement agency that is located either where the patient or the potential victim resides. A third would be to arrange for either a voluntary or involuntary hospitalization. As a rule of thumb, the important action is to protect both the patient and the intended victim. So, there is no need necessarily to breach confidentiality if you do something else to avoid the harm from occurring, such as having the patient hospitalized. And as long as the patient is in a secure setting and cannot leave without an evaluation, it is generally not necessary to breach confidentiality by communicating directly with the identified potential victim.

Since you mentioned the topic of involuntary hospitalization, let’s discuss that a bit. Describe some of the laws regarding when we can commit somebody involuntarily.

Again, the laws vary somewhat from state to state, but there are generally three things that allow somebody to be involuntarily detained either for an evaluation or for treatment: first, danger to oneself as evidenced by either threats of suicide or suicidal actions, second, danger of inflicting serious bodily harm on another individual, and third, inability to provide for one’s basic needs as a result of mental illness. In those situations, the psychiatrist is generally on good grounds for having the person detained and evaluated even involuntarily.

I’ve certainly had the experience of incurring a patient’s wrath by going through with a commitment, and have even been fired by a patient. How do we avoid getting sued over this issue?

It’s extremely uncommon to be sued in this situation. In fact, in Massachusetts, nobody has ever been successfully sued for issuing an involuntary commitment. As long as the psychiatrist has thought through and exercised professional judgment in coming to the conclusion that the person needs an evaluation, there is no basis for a suit. There was a relatively recent case in another state involving a suit, but that resulted from a complete misuse of the commitment power of a psychiatrist for personal reasons.

There is clearly a spectrum of cautiousness out there in the community, from psychiatrists who are more likely to commit to others who are reluctant. How can we help patients stay safe short of involuntary commitments?

It is always better to try to engage the patient in a treatment plan rather than to automatically have an ambulance show up and take them to the hospital involuntarily. Options include speaking with the patient and having them sign into a hospital voluntarily to receive treatment, referring to a residential program or a partial hospitalization program, increasing the frequency of outpatient sessions, adding in group treatment, and coming up with a safety plan with other supports that the patient may have within the community.

What is a good “safety plan”? What does that mean?

Well, a good safety plan is one that would acceptably reduce the patient’s risk of harm to himself or to somebody else. This always comes down to making a clinical assessment, sometimes using consultation with colleagues . Overall, it’s best to think about what the least restrictive means to assure the patient’s safety would be.

What is the status of the thinking of the “safety contract” these days?

The most recent American Psychiatric Association practice guidelines for suicide prevention suggest that the safety contract is of limited utility. On the other hand, it is something that clinicians still do, but not in isolation of other measures to provide for the patient’s safety.

What would be an example of a good safety contract? Does it have to be in writing?

It is not necessarily a formal or written contract, but it should be part of an overall safety plan between the clinician, the patient, and oftentimes third parties to help the patient have multiple options short of trying to harm themselves. One example is to come up with a schedule of when a patient who has had transient suicidality might check in with other parties so that people can make sure that they are safe. And if the patient doesn’t call the identified person at the specified time, then the contract should specify what happens next. Or in the case of a clinician, if the patient does not check in or does not come to their appointments, the contract might dictate that the clinician will proceed with involuntarily detaining the patient for an evaluation.