Wednesday 25 June 2014

When is a Parachute Just a Parachute?

There is a great joke about the use of randomly controlled trials to test the efficacy of parachutes. Gordon Smith and Jill Pell, an obstetrician and an epidemiologist respectively, published a "systematic review" in the BMJ in 2003 examining the use of parachutes to prevent "major trauma" resulting from "gravitational challenge". They conduct a literature search, find no relevant studies and satirically conclude "the basis for parachute use is purely observational, and its apparent efficacy could potentially be explained by a “healthy cohort” effect." 

It's a sharp paper; funny and immediately clear about what it is mocking. Parachutes are not "evidence-based" and we cannot point at statistics to validate their use. Should we really go on using them ? Of course! It is in the nature of parachutes that we can see quite clearly that they work and why. Try asking anyone to jump out of a plane without one and see what they tell you.

Coldplay's Parachutes: Not Evidence Based,
Not Effective, just Dismal.

Smith and Pell's paper is an explicit call for "common sense" in the evaluation of interventions. Rather than mindlessly rely on the absence or presence of "data" or "evidence" when making an intervention, we need to be open to the idea that sometimes our understanding of the world, and the use of observation will allow us to gauge whether an intervention "works".

What is a Parachute?

In clinical psychology the parachute-RCT example is sometimes used to defend the idea that not everything we do can or even should be subjected to research-validation. Variations on the argument occur frequently, even when the paper itself is not cited. Smith and Pell are surely basically correct about the fact that an "evidence-base" is not always essential in order to know what is the right thing to do, but that is the easy part. The question which inevitably arises next is which sorts of interventions are "parachutes" and which are not.

We meet a great literary example of parachute failure in Will Self's 2000 novel "How The Dead Live". Mr Khan is a drab seeming data-obsessed Clinical Psychologist who approaches the novel's narrator Lily Bloom, an elderly lady who is at that moment dying of lung cancer on a hospital ward:

Excerpt from How The Dead Live. More available here.

Khan appalls us with his complete failure of common sense, common decency and basic human empathy. Whatever it is that people need when they are dying on a hospital ward, this psychologist is failing to provide it. Rather than hold himself open in some way to the despair and loneliness faced by the dying Lily, he cravenly avoids it and goes about his seemingly trivial data collection. We who read this (especially those of us who fancy ourselves to be caring clinicians) feel we could do something, anything, that would be more helpful. We can plausibly be correct about this; you don't need an RCT to learn kindness.

The fear of being a "Mr Khan" may well play a role in a more generalised skepticism about evidence-based practice in Clinical Psychology. Despite the importance of the "scientist-practitioner" paradigm, an ongoing uncertainty about when we do and when we don't need statistical evidence rumbles in the profession. Some have gone so far as to suggest that non-clinicians shouldn't do therapy research, as they just don't get that intangible common-sense-something that makes therapy helpful. 

Perhaps it's no Wonder New York's Needs Adapted
Treatment Service is Called "Parachute"

To be sure, there are parachutes in mental health. Some fairly influential ideas from psychotherapy seem to fit the bill: Carl Rogers' "unconditional positive regard"; Winnicott's "holding", Sullivan's "evenly hovering attention". What binds these ideas together is their view of an almost ethical stance clinician takes towards client, listening to them, taking them seriously as a person. This post by Gordon Milson on time in mental health services is another example; perhaps, he argues, the EBP movement runs the risk of forgetting that different people need different amounts of time before they can bring themselves to form a relationship with their therapist. Waiting for someone, being patiently there for them, might well be a parachute.

When is a Parachute Not a Parachute?

Alas, not everything is a parachute. While some of the things that psychologists and psychiatrists do are simply ways of breaking a fall, other interventions are intended to be active flying machines. These cases cannot be allowed to slip in under the radar, but should be empirically assessed to see which of them flies and which is are little more than hot air.

Sometimes a Parachute is not a Parachute.

Most of what psychotherapy aspires to is surely not a parachute, though this is sometimes skirted around by describing it as simply "talking to people". If psychotherapy were no more than this then it wouldn't be necessary to test it. Talking to people (and, more importantly, listening to them) is often an inherently good thing, but we should not conflate psychotherapy as an opportunity to do good with psychotherapy as itself inherently good. Psychotherapy as an activity is designed to be a form of talking that changes people's minds and behaviours. Where the mechanisms that make parachutes work are clear and obvious (you don't need a PhD in aerodynamics to understand them well enough) the putative mechanisms that make psychotherapy effective are hotly contested. Being kind and empathic might be quite simple, but effectively helping someone overcome a psychological problem is assuredly not. Calls for "common sense" in our discipline are limited by the fact that there is rarely much that is "common" to how different people see the mind.

If psychotherapy can do good then it can presumably do harm as well. What is more, psychotherapy is expensive, meaning that if you are going to provide it you need to show there is more to it than the proverbial chat with someone lacking those crucial qualifications. By introducing the notion of a "parachute" into the vernacular, Smith and Pell found a way to call out Evidence-Based-Practice enthusiasts when they reach levels of absurdity. If we are to continue to find the idea helpful we need to establish some sense of when their argument does and does not apply.