When is Shared Decision Making actually Shared?

By Lorenzo 19th June, 2016

Annalisa, Brook, Elicia, FPA, Health, King's College London, LSHTM, MCDA, Multi Criteria Decision Analysis, My Contraception Tool, NHS, Self-Harm, Shared Decision Making

There is a lot of publicity about Shared Decision Making. For example the term comes-up a lot in regards to healthcare, when a patient is included in the dialogue with the healthcare professional to decide on the most appropriate course of treatment. Sounds great, except that the issues around implementation and barriers to entry for daily use can result in at best an inconsistent approach to how Shared Decision Making is perceived by both professionals and the general public. Given that Maldaba’s been involved in this area since before the term had become common parlance, I thought it was time to write about our perspective of design, development and implementation.

Maldaba produced version 1 of Annalisa, our Shared Decision Making tool based on Multi Criteria Decision Analysis (MCDA) theory in 2006.  Annalisa was designed and developed in collaboration with Professor Jack Dowie, Emeritus Professor of Health impact Analysis at the London School of Hygiene & Tropical Medicine.  Put crudely, MCDA is the process we go through when we choose one option from several, taking into account a number of factors (attributes).

For example when I choose an ice-cream flavour to a hot summer’s day, there are a number of flavours I can choose from (the options).  These flavours will fit certain factors (attributes) differently.  Some will be cheaper than others, some will be more thirst-quenching, and some will be sweeter.

Those attributes can be externally and objectively verified.  Price, sugar-levels and water content can all be measured and quantified.  Where ‘thirst-quenching’ is a more complex attribute because it combines water content levels with the absence of chemicals that make you thirsty (salt), it’s still possible to quantify.  If experts disagree with one another they can discuss the equation for ‘thirst-quenching’ and as the attribute’s definition evolves over time, so the measure of each flavour against that attribute may be objectively updated.

That’s all very fine, but nothing so far helps me choose the best flavour for me.  How thirsty am I?  How much money am I willing to spend?  What about flavours I don’t like?  None of these are objective.  No-one else can say how my current state best resolves to Vanilla or Chocolate (Mango sorbet, anyone?).  I need to address each attribute, and measure their importance for me right now.  Once I’ve done that, Annalisa can help me choose the right flavour.

How?  Well the objective markers I discussed earlier are Ratings.  Each option is rated objectively as to how well it meets each attribute.  My personal preferences are Weightings.  I can subjectively give each attribute a weight, indicating how important that attribute is to me right now.  Some simple maths then multiples my subjective weighting by each objective rating, adds them together, and scores the options.  Annalisa then shows me both the best flavour for me right now, and how close (or far) the other flavours scored to the winner.  Here’s how that looks in Annalisa:

And for those who want to understand the constituent parts, here’s how each attribute contributes to the final score:

Why isn’t everyone doing this?  Traditionally MCDA software has been complicated, designed for professionals, with lots of tables where numbers are entered.  Our approach was to create something for mass consumption.  The interface is graphical; you just drag those blue sliders from left to right to say how important an attribute is to you.

It’s crucial to remember that there are no ‘right’ answers or outcomes to this process.  There’s only what’s right for you at a given time.  Your preferences will change over time and therefore so will your weightings.  Ultimately that may or may not result in different ‘best’ outcomes for you.

The principles on which we created Annalisa were of empowering individual choice, and of transparency in that process.  Annalisa has no opinion; it doesn’t care which option you choose.  Annalisa just wants you to make sure you understand what your motivations are in making those choices.  When we launched version 1 of Annalisa I showed it to a close friend, who at that time was selecting a nursery for his eldest child.  He had to choose between two.  One was closer and cheaper but didn’t fit his idea of the sort of place he wanted to leave his daughter.  There wasn’t anything wrong with it, but the nursery didn’t appeal to him.  The other was further away, more expensive, and felt much more closely aligned with his expectations.

We set-up an Annalisa Topic, measuring the distances, comparing the prices, plus other factors (we entered around four attributes in total).  Once the ratings were in, my friend weighted those attributes.  I hid the scores in the interface whilst he did this to keep the process as honest as possible.   To his great surprise, the closer/cheaper nursery won.  Was that the wrong outcome?  Not at all, based on how he’d declared the importance of travel time and cost, against ‘good fit’ (this was one of our attributes).  Did he change his mind and send his daughter there?  Of course not, he went back and adjusted his weightings to ensure that the more expensive nursery won.  The great moment was when it challenged him for a few minutes, and he was forced to acknowledge there was a gap between his emotions and how he’d rationalised picking the nursery.  Annalisa closed that gap for him.

The nursery story is an unmitigated success.  The point is not to force people into making decisions they’re not comfortable with.  The point is to force them to be open, honest and transparent about making that decision.  My friend had to acknowledge either that driving the extra distance and paying the extra fees wasn’t so important to him, or else he had to change his decision about which nursery to send his daughter to.  As long as he does one of these things, he’s using Annalisa properly and he’s making a sound decision, in full knowledge of why he’s making it (there’s nothing wrong with an attribute like ‘good fit’, though you have to be careful about how you rate this).

This is what makes Annalisa so powerful: you’re making the right choice for you, right now.  Nobody else can do that for you.  When it comes to healthcare (or any other scenario that involves professionals) then it’s truly shared decision making when the professional has supported the individual to make a personal choice, individualised to their circumstances, and that individual is empowered to say what matters to them.

Beyond ice cream and nurseries, Annalisa has been used by Brook and the FPA here in the UK to create My Contraception Tool, a free-to-use decision-making app to help people choose the best contraception for them.  It’s also been used by King’s College London to develop DASH, a tool for young people to seek help and avoid self-harming.  Researchers around the world are using Annalisa to make decisions about medical treatments, including cancer treatments.  For more information, visit the Annalisa website, or contact us.