by Chuck Ruby, Ph.D.

Investigators from the French Land Transport Accident Investigation Bureau (French acronym BEA), which has been investigating the March 2015 Germanwings crash, have recently recommended that world aviation agencies require mental health workers report pilots whose mental heath condition "could threaten public safety". This is an alarming mistake. See the NBC story here: http://www.nbcnews.com/storyline/german-plane-crash/germanwings-crash-bea-investigators-urge-new-rules-about-reporting-mental-n537416.

One of the most robust findings of research is that mental health professionals are not good at predicting others' future violence or harmful actions. There are clear risk factors that increase a person's likelihood of doing so, but they do not enable us to predict whether the behavior will occur. The best we can do is manage those risk factors in order to reduce the chances. Complying with BEA's recommendations and putting mental health workers in the position of making predictions, will result in a huge false alarm problem, destroying people's lives in the process, and ironically causing those who would benefit the most from help with emotional problems avoiding such help.

Many occupations require employee background evaluations in order to determine continued psychological suitability and readiness for the job. In addition to airline pilots, other examples are law enforcement officers, employees working in national security programs, and nuclear power plant workers. The employers in these settings have a reasonable interest in knowing their employees can be trusted. And employers can understandably err on the side of caution by denying employment when possible problems exist. But requiring a psychological evaluation for a suitability determination and requiring mental health workers to report people to the authorities are two different things.

A mental health worker, most likely some type of counselor in this situation, has a fiduciary obligation to the person being helped. The relationship between the counselor and the client is paramount, and the confidential nature of the relationship is at the core of this trust. Requiring the counselor to report people who "could threaten public safety" places the counselor in an ethical double bind. If the counselor reports, she or he damages the relationship and harms the person. If the counselor fails to report and the person commits some act, the counselor becomes the target of legal and ethical censure.

Attempting to determine who "could threaten public safety" is impossible. This could actually mean anyone. How many times have uncharacteristic acts of violence been committed by people who never showed any warning? It would be easy to predict future harm when a client makes a statement about intent to harm a specific person(s), but how often does that actually happen? Hardly ever. Besides, there are already laws and guidelines in place that require counselors to warn of such clear cases.

But other than these clear cases, where does the counselor draw the line? How much of a risk is needed prior to warning? How do we quantify that? Must we use standardized instruments to assess the level of risk? How valid and reliable are those instruments? The answer is: not very. With behavior that happens infrequently among the population, any assessment of risk will necessarily have very high false positive results or false alarms. This is just a matter of statistics. Violent behavior is no exception. For instance, the FBI reported a .4% base rate of violent crime in 2013, and this includes all types of violence, not just murders and assaults (see https://www.fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2013/crime-in-the-u.s.-2013/tables/1tabledatadecoverviewpdf/table_1_crime_in_the_united_states_by_volume_and_rate_per_100000_inhabitants_1994-2013.xls). With such a low base rate occurrence, even the most accurate assessment instrument is doomed to result in extremely high false alarms. In other words, the great majority of those identified as potentially dangerous and reported by counselors to aviation authorities, will never commit a dangerous act. But being reported as at risk for such behavior can ruin their lives and damage the value of professional help.

The slope is slippery. For what other occupations will this kind of big brother monitoring apply? Should counselors be required to report police officers? Teachers? Bus drivers? Investment bankers? Physicians? Who would report the counselors?