Healthcare, medicine & pharmaceuticals

The robot must not injure the patient

People who have a disability that demands intensive care, like autism, Altzheimer’s or stroke patients, may benefit from the care of a trained robot. At least, that is the belief of a professor at University of Southern California, who is working with a team of experts in different fields to design a robot with the personality and temperament most suited to its patients. While a robot can be used to make someone move in a certain way, it is much more challenging to motivate a patient and develop trust so the robot and the patient can work together.

One of the challenges inherent in patient/robot interactions is the so-called “uncanny valley” effect, where people become uneasy if the robot seems too human. But experiments with patients show that they do respond to encouragement from the robot, and it can influence them much more in the room than on a computer screen. They also uncovered the difficulties in working with an introverted or extroverted patient if the robot did not also assume the same nature. A robot’s sociability can be measured by its distance from the patient, speed of its movement, pitch of voice and type of communication – all this must be painstakingly programmed.

Researchers were able to program the robot so it gradually responded to the patient’s style, dubbed “machine learning”. People are very good at masking their feelings, so other experiments used sensors to measure physiological changes so that the robot could more accurately gauge the patient’s true response.

Of course, not everyone believes that using robots for this kind of therapy is healthy or desirable. Sherry Turkle, a professor at MIT, believes it is dangerous to foster relationships between people and machines, particularly the vulnerable, such as the elderly and children. She believes there is “no upside to being socialised by a robot” and that patients will bond with the robot as if it were human, creating attachment and unreasonable expectations.

Researchers are grappling with the ethical implications of social robots, as well as the complexity of their programming. There is always the question – why not have a human being care for someone? In some ways, creating a caring robot is more of an intellectual or scientific challenge than a necessity. But if there are people who cannot relate to others, who benefit from sharing life with a robot, perhaps there is some benefit in that.

A spoonful of sugar makes you better

A rather ironic problem has surfaced in medicine, which has the pharmaceutical manufacturers scrambling for a solution. Their drugs are not standing up well to the placebo effect. In too many cases – drug trials for schizophrenia, depression, Crohn’s disease - patients taking the placebo are doing as well as the ones taking the real drug. Even well established drugs like Prozac are starting to falter in follow-up trials. But the manufacturers are worried they can’t help ill people – and their own profits will sicken too.

The pharma response has been uncharacteristic – to start sharing information with one another to get to the bottom of it. The discovery of the placebo effect prompted a change in US testing procedure in 1962, to control for it. Manufacturers are now wondering if there needs to be a change in testing to more accurately demonstrate what changes are caused by the drug and what is placebo. This is especially so, given that many drugs today work on the central nervous system, which is responsible for the placebo effect in the first place.

Some researchers found that drugs behaved differently depending on where they trialled, for example, Valium beat placebo in France and Belgium, but not in the USA. They also discovered that observers in one trial gave different ratings for the same results in another trial. In Italy, one researcher has spent 15 years mapping the biochemical reactions that cause the placebo effect and found placebo-activated opioids can relieve pain but they also modulate heart rate and respiration. He also found Alzheimer’s patients, for example, were not able to anticipate reduced pain so continued to experience pain. So placebo aids recovery by using the mind’s ability to predict the future.

Even the colour of a drug has a placebo effect – blue pills are more tranquillising than red. So is it all in the marketing? Marketing lifts expectations, so people buy a pill and look forward to a certain result. The placebo effect may be strengthening simply because people are more jaded in their expectations of what is going to help them. Or it may be time to redefine the boundaries of mental illness.

You might ask – so what, at least they are getting better? Whatever the problem, pharma wants to introduce drugs that “route around the brain’s own centralised network for healing”. This seems to be an extraordinary mission – preventing the body from doing its own healing so the pharma companies can do it instead! We may have misread their intentions. But it may be time to question whether sometimes it wouldn’t be better just to take a spoonful of sugar – without the medicine.

We can read your lies

There may be times when you have wished that someone would read your mind but there are already cases of people who have had their minds read and ended up in jail. Thanks to neuroimaging techniques, which purport to measure brain response and interpret it – people can be indicted for crimes, at least in India.

Researchers in Mumbai use the Brain Electrical Oscillations Signature (BEOS) test, where an electro-encephalograph (EEG) machine is hooked up to someone’s brain via tiny wires and measures the frequencies of electrical signals from the surface of their scalp. It tests for eleven psychological variables. A computer then calculates the response, which, in guilty cases, is “experiential knowledge”. In other words, you could not have known it unless you had done it.

US and British courts do not accept this kind of evidence but private enterprise is keen to make sure they eventually do. There is growing academic enthusiasm for researching functional magnetic resonance imaging (fMRI), with only four papers published in 1992 but over 19,000 of them in 2007. Marketers may recall the famous Coke v Pepsi test.

But like all popular discoveries, there are detractors who believe these tests stretch the limits of brain imaging beyond statistical probability. It is challenging to measure an emotion in numbers, so the correlation can never be any more accurate than the original figure used for the emotion. Standard psychological testing can never be more accurate than 70-80%, which makes more than 80% accuracy from brain scans look implausible.

It is perhaps no surprise that the inventor of the BEOS test describes himself as “an electronics man, not a psychology person”. But using electronics to send people to jail seems shocking. The concept that physiological traits are localised to certain parts of the brain seems to be well accepted, but this kind of testing seems to claim ultimate understanding of the brain! Few scientists believe we are that close. Until then, some poor victims will be unfairly accused.

Mobile phones are safe, perhaps

People who have been using mobile phones for less than ten years appear to be safe from any likelihood of cancer. But the international Interphone study, into whether mobile phones increase cancer rates, has been unable to assess longer-term use simply because people cannot recall that far back. Not only that, phones more than ten years ago were considerably different from what they are today.

Some cancer experts have publicly warned that there is enough evidence to suggest an increase in head tumours, but the International Commission for Non-Ionising Radiation Protection (an independent body) says the longer term results are not likely to be available for some time. This leaves the public once again in the position of guinea pig.

In some ways, it may be like the numerous studies into tobacco use. They constantly warned that tobacco was addictive and dangerous but people continued to smoke and the tobacco companies continued to sell cigarettes. Many people are now addicted to using their mobile phones and it seems unlikely they will change their habits without good evidence. Mobile phone companies can breathe easy for now.

Mindfulness goes mainstream

Scientists need to prove what many monks and secular practitioners have known all along – meditation can be useful for people with depression. The Mood Disorders Centre in the UK has compared a technique called mindfulness-based cognitive therapy (MBCT) with drug treatment and found MBCT is as effective as maintenance antidepressants in enhancing quality of life.

As in our story above, neuroimaging techniques are being used to see what the brain is doing while meditating. It appears that experienced meditators are able to shut off the “chatter” of the mind and return to a calm state more quickly than others. Regular meditation even alters the brain structure, causing the cortices to be thicker, especially in areas involved with attention and sensory processing. It is hoped this technique can be used against other psychological conditions like chronic anxiety and eating disorders (– and skepticism?!)

Are you an orchid or a dandelion?

Here is a fascinating idea, which may reassure depressed or anxious people who can’t get on with mindfulness meditation (above). Scientists are exploring the idea that we are either orchids or dandelions, depending on our genetic make-up. Orchids are people who are fragile and changeable, and need more care than other people. Dandelions are hardy and resilient, able to grow in almost any soil or climate. You might think it is better to be a dandelion than an orchid, but this thinking is flawed.

It appears that the very same genes that make us self-destructive or unable to get on in life are the same genes that ensure progress of the species, adaptability, and creativity. So people who are depressed or anxious, given the right environment or parenting, can be more successful than dandelions, who do not have the same innate drives. While someone may be genetically vulnerable to depression or anxiety, that does doom them to a lifetime of depression. It means they have a heightened sensitivity to all experiences, not just negative ones.

The problem with work on behavioural genetics is it tends to focus on vulnerability and studies do not look for the positive aspects of “bad” genes. Yet families with both dandelion and orchid children are more likely to succeed over time because they are “diversified” and supply a stronger basis for the species. It also helps to answer the question why orchids did not die out when their environment became harsh, because maladaptive genes ought to die out.

The orchid (or sensitivity) theory is still relatively new but researchers are fascinated by its implications for people with “bad” genes. Leiden University found that children with genetic vulnerability who are put in the right setting do better than in a poor setting, and even do better than the dandelions from a good background. More interesting still, only human beings and rhesus monkeys have this ability to adapt to new, changing circumstances and live anywhere. So we are born to last! Nothing to get depressed about, after all.