The great advantage of fighting a world war is that, if you win, you quickly find out a great deal about the deposed tyrant. We knew more, and more quickly, about the deposed tyrant Hitler than his erstwhile confederate tyrant Stalin, whose secrets were opened up a decade later, and only partially, by his successor Krushchev, in a speech which remained secret for some years after.
When a tyrant falls it is perfectly natural to make a “never again” list, and if that list is too inclusive then it seems a small price to pay to avoid the rise of another beguiling tyranny. A wide list of “never again” resolutions in Germany included the steps taken to deal with the actions of Nazi doctors, who participated in what were originally presented as mercy killings, and then went on to become a model for for widespread killings as a matter of covert state policy. Robert Lifton chronicled these chilling matters in “The Nazi doctors” in 1986.
I reviewed it for the British Medical Journal in 1987, up against a very tight word limit despite the size of the book. Digging up that review from the BMJ archives 28 years later (in itself a mildly emotional event, because I had forgotten it) I see that I concluded: “Lifton's book represents a considerable achievement. I believe that it should be in every medical school the world over, and that it would be fruitless to teach medical ethics without it.”
What conclusions did post-war German government draw about medical matters? Medical confidentiality was given an understandably highly protected status. The ministrations of the Gestapo in Nazi Germany and later the Stazi in East Germany were vivid reminders of the dangers of giving the authorities carte blanche to poke about in citizens’ private matters. Medical records can be very intimate. Doctors already have the authority to break the confidentiality vows if they feel their patient puts others at risk, said Frank Ulrich Montgomery, head of the Federal Chamber of Physicians recently. However, having the authority is different from having the habit of doing so when the public interest seems to demand it. As this debate gets started in Germany there have been the usual calls for the need to avoid hasty moves, with less discussion of the dangers of slow immobility.
One case, however terrible, should not overturn a general principle. Esteemed psychiatrists of my acquaintance assure me that the suicidal co-pilot would most probably have been picked up in the UK, and would not be anywhere near a cockpit, meaning that UK psychiatrists would have at least threatened to break confidentiality in similar cases, at least to the extent of saying to employers, if the person was unwilling to disclose his problems himself: “This person must not pilot an aircraft in his current state of health”. There are regulations and guidelines on these matters. However, these same UK psychiatrists think that, for entirely understandable reasons, Germany has taken medical confidentiality too far.
Here is a quick snapshot of some of the problems in monitoring pilots for mental health problems:
1) Detecting low frequency/high impact events is notoriously difficult. The false positive rate is very high, and it is difficult to maintain vigilance throughout the system because most of the time nothing happens.
2) Pilots are motivated to make light of their problems, or they will lose flying hours, income, and promotion. That puts the screening physician in the position of having to detect mental state and also to detect lying. This is not impossible, but rarely part of medical training. Clinical Assessment of Malingering and Deception Third Edition Edited by Richard Rogers is an essential reference.
3) Humans are mutative. If they are closely monitored they may become even more devious. As a consequence, airlines would have to ensure that there were very high payments for honesty, to encourage disclosure of all personal problems.
4) Crew members should inform management when a colleague seems mentally unwell, but that will put them under equal scrutiny, and is not conducive to crew teamwork, apart from the difficulties of spotting it anyway.
5) More intrusive monitoring might lead to more stress related errors.
James Reason of “Human Errors” used to say that error reduction was like a bathtub: errors fall right down and then, just as you feel safe, they start rising at the other end. Call it The Godel Limit: not every risk can be specified, and at some point the safety cost of protecting against a small risk becomes bigger than the benefit obtained. For example, having two persons in the cockpit at the same time will reduce single agent risks but may increase social distraction risks.
Juvenal was there first: Quis custodiet ipsos custodes?
Nonetheless, even if mental illness is assessed as occurring in 25% of the population, there is a strong case for recruiting pilots from the more stable 75% without a history of such disorders. Passenger safety is more important than pilot careers.
Another approach is to get rid of the pilots altogether. Planes can take off and land automatically at most airports, and level flight is pretty boring and routine anyway. Once we can get automatic systems to cope with frozen Pitot tubes, volcanic ash clouds, engine flame out, major storms and other such rare events, it could be autopilot all the way. Pity about the uniforms. The software would need to be good, and every programmer entrusted with upgrading the system would have to be subject to psychological screening………. It’s Juvenal all over again.
Finally, if you have read this far, please acknowledge the power of a headline.