Algorithm is going to get you (Kappitel 1)

“So on a scale of 0-10 with 10 being worst pain imaginable and 0 being no pain where would you score your pain………….. just roughly…………. to give me an idea………. so that I’ve got a baseline…… maybe for the last week or so…… an average….. out of 10….. JUST GIVE ME A NUMBER!”

The latter part, I hope obviously, is soliloquy. But it conveys frustration felt by physios the world over. All I want is a number to write down so I can carry on with my assessment algorithm. The next step in the chain. But are we collecting data for its own sake? Do we have a purpose other than routine? How much do we gain from pseudo-quantification of a subjective phenomena? And how much do we lose?

On reflection I’ve always felt awkward trying to coerce someone to score their pain. Have you ever noticed how frustrated they can become? Pain is subjective. By its nature difficult to score. Which is why some won’t score no matter how hard you push, others clam up, many insist on 0.5 decimals, others score 11 out of 10! All indicators of discomfort with the process. And whilst I see its intended target (grasping severity) equally it can be dismissive of their experience.

It’s an odd concept. Consider asking people to score other subjective experiences out of 10. When your partner says they love you. (Or hate you!) When your dad says he is proud of you. When your friend tells you they hope to get a new job. When your child is sad because they are being bullied at school. How would each react to being asked to score these experiences? They’re describing a feeling, a subjective experience. By asking them to score this we may invalidate their experience. Open it to comparison. Some experiences become more valid than others. Intensity and quality are appropriate to measure but can be gained without numbers.

algorithm

The subjective risks becoming a quantitative data collection process. That can be measured. That act as markers. There is a deep rooted attraction to objective, mathematical data. It is neat, concise and ‘accurate’. I always remember a lad I played footy with at Uni doing a Maths degree. He’d often get a mark of 100% in some of his modules! (I wonder how many University essays are marked at 100%) The hope with algorithm is that it’s mathematical, methodical and if you get things just right it’s 100% guaranteed. All you need is comprehensive data and a robot. Magic!

algorithm2

So can health be reduced to data? Can science be reduced to numbers? Can pain be reduced to a score? Of course some people can score pain without inner turmoil. But when we force a subject to be an object we lose something. In the same way if we force objective data into purely subjective. Imagine the difference between describing your favourite song, painting, game compared to actually seeing, hearing, experiencing or observing this in person. Hence “a picture paints a thousand words” (although even this is subject to interpretation). The danger of algorithm is that we force their story into numbers and measurements and miss powerful words and narrative. Allow them to paint a picture, not colour by numbers.

flanagan quote

No calls should be made to abandon objective data. Rather gather it appropriately. It does ask questions of how we collect meaningful data. Can we capture qualitative outcome measures? We need in this age to show our impact. This is easier in number. We are also keen to distance ourselves from alternative therapies. This has pushed us away from non-numeric outcomes in order to be viewed “real” science. Our fear is that alternative therapies may produce positive data. Rightly so we should aim higher than this. But should we throw it out?

What about in science? The VAS (Visual Analogue Score) and NRS (Numeric Rating Score) are commonly used in research and has been shown reliable but no way of measuring pain accurately exists (construct validity). The pain experience then becomes ordinal data (or even interval/ratio). As a surrogate measure VAS may still has a role in research but clinically it is more questionable. Does using VAS make a better clinician?

Matt Dancigers blog challenges the purpose of outcomes nicely. Are outcomes patient centred or are they focused on the outcomes of the clinician or research funders? Are longterm health outcomes clinically more useful than immediate pain or function? Not to dismiss pain and function. These are often the reason for seeking intervention. But I wonder by stopping short of health (at pain and functional improvement) do our long term effects suffer?

Health is a vast and complex arena. Sometimes pain is easy to change. Immediate behaviour change often prompted by consciousness of pain plus an appropriate strategy. But overall health change is harder. It requires lifestyle change. When the prodding of the conscious dulls through a narrative of time or helplessness. Sometimes pain is difficult to change, should we focus on health instead? How much should we be looking for pain management strategies and how much for health management strategies?

So let us think. Think where an algorithm is going to get you. Are we looking for quick answers to robotic questions? Do we deal in singular thought processing or binary outcome (yes/no)? Unfortunately the subjective history is nuanced which doesn’t suit an algorithmic approach. Sorry robots.

Be more human. Be less robot

Thanks for getting this far.

Neil

Further Reading

I recognise there were more questions than answers with this blog. Would be great if you have a spare minute to head over to Facebook and like our page and engage in further discussion around VAS, outcome measures and algorithm. http://bit.ly/bemorehumanFB

@MattDancigers blog https://ptbraintrust.wordpress.com/2015/02/11/on-a-scale-of-0-to-10/

Some helpful ideas from IRISS on qualitative outcomes http://bit.ly/qualoutcome

A nice piece in NEJM on why we use pain intensity score + why this hasn’t helped http://bit.ly/VASintensity

Content

Clarke et al 2012 http://nrl.northumbria.ac.uk/10947/1/A_Clarke_-_PROVISIONAL_PDF_-_I_feel_so_stupid_I_cant_give_a_proper_answer.pdf

Dancigers 2015 https://ptbraintrust.wordpress.com/2015/11/18/who-is-satisfied/

Estefan 1987 http://bit.ly/estefanrythm

A nice discussion and replies to article on limitation of VAS:

Kersten et al 2012. Discuss limitation of VAS http://bit.ly/VASinterval

Price et al 2012. Defend use of VAS as ratio data http://bit.ly/VASration

Franchignoni et al 2012. Discuss ontology and limit of VAS. http://bit.ly/VASontology

10 thoughts on “Algorithm is going to get you (Kappitel 1)”

  1. Really good read. As an osteopath curious about being “keen to distance ourselves from alternative therapies” and “our fear is alternative therapies will show positive result” Why? Are osteopaths regarded by physios as alternative?

    Liked by 1 person

    1. Nope. I was thinking more maybe homeopathy, reiki, energy medicine etc. My eyes have been open to really good osteopaths on Twitter!

      I think we worry in ‘science’ that subjective/qualitative can lead to “quackery” but we can swing the pendulum too far the other way!

      Thanks for reading. Glad you enjoyed it!

      Like

      1. Thanks for taking the time to clarify. We osteopaths feel a bit under attack at the moment. We are used to the alternative label. It never used to feel like a negative thing, though.

        Liked by 1 person

      2. I must say I went to osteo college in 1992 and we were never taught anything to do with “subluxation”, I’ve never heard the word used in any osteo clinic, and I’ve never really understood what it means. I think we thought it was more an old-school or chiropractic term.

        Liked by 1 person

  2. Nicely done Malts, I have to say that I’ve almost completely stopped asking the VAS in my assessments. I question how much value it has, with everyone’s pain perception being different. I think observation of the patient can give far more information regarding their pain severity.

    Nige the Physio

    Liked by 1 person

  3. Morning Neil,

    Excellent stuff once again.

    We are in the odd position as physiotherapists in the fact that we are sort of science led but then have to objectivise the non-objective. Pain, ROM, movement pattern, muscle power………
    I sometimes just ask ‘are you able to scale or grade your pain’ and leave it at that.
    The classic response can be ‘it is only a five out of ten but feels like the worse pain I have had because it is nagging’. This then begs the question is a 5/10 nagging worse than a 9/10 intermittent stabbing.
    I can remember algorithms from my engineering studies. We usually left about 7 decimal places at least for each number to give the most accurate answer. Just as you said in your blog, you get a right or wrong answer. Easy life.
    I think algorithms in physiotherapy can be useful in establishing the salient features of something and putting order to them. They can also be too rigid or even potentially kind of dangerous and stop us thinking. For example, checking red flags, cranial nerves or neurological examination etc.. If the algorithm does not have a feedback loop and it is followed literally, it can feel like that issue is over, and we can move on now in safety. This is obviously not true as we must always try to remain vigilant, at times to issues that the patient has not come with.
    Keep up the good work, but don’t get blog burnout. ( I would put a face here but do not know how to).

    Kind regards

    Liked by 1 person

    1. Great comments. Thanks for reading, reflecting and taking time to feedback!

      My next blog will look at algorithm v method. Which is along the lines you outlined above. I suppose the algorithm is a metaphor for cut + dry reasoning and simplistic question + answer.

      I won’t go too much more into it as I’ll probably just repeat my next blog!

      Thanks again

      Neil

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s