Thaler’s pop science hit book Nudge1 has certainly created a stir; bringing behavioural science to the centre stage in the policymaking, private, and public arenas. The knock-on effect has been a growing excitement about the behavioural science buzz and the society-wide innovations such a ‘liberal paternalistic’ approach can offer.

In our day to day lives we make countless decisions. Many without paying much thought; and many despite often knowing they are not in our overall best long-term interests. In this sense, we are not making decisions to optimise our health and happiness in the way a completely rational being should. The concept of a ‘nudge’ acknowledges this and builds factors into our environment that makes it easier for us to make decisions that are better aligned with our long-term interests.

A good example of this is ‘opt-in’ automatic enrolment forms. Decisions, such as setting up a pension can seem cognitively draining, confusing and involve complex decisions about options we do not fully understand. Without a highly urgent motivation, the lack of ease is a barrier to action. However, automatic enrolment makes it cognitively and physically easier to make a decision for the best outcome – and increases saving.

However, much of our understanding of nudges comes from studies using simple laboratory-based tasks and we must always be cautious when extrapolating these to more complex real life situations; not all behaviours have been subject to simple effective modification by nudges alone. For example, it would be ludicrous to expect a nudge technique to persuade a patient with a peanut allergy to consume a pack of nuts. Medicines taking behaviours are another prime example of this. Many approaches to increase adherence in recent years have focused on removing practical barriers to medicines taking and providing simple nudges to make it easier to adhere. These engineer an environment for the patient whereby taking the medicine becomes easier – it involves less physical and cognitive strain. An example of this could be a pill reminder, informing the patient when to take their medicine. However, recent reports have suggested such approaches may have limited impact and do not increase adherence in all cases.2

Decisions patients make about medicines taking, like all decisions come down to two reasons – they can’t and don’t want to. Can’t: the action is not easy to make and don’t want to: their beliefs are leading them to choose not to. (Referring back to the patient with a peanut allergy, they may be perfectly capable of assembling and eating a peanut butter and jam sandwich, but regardless, do not want to.) This is outlined in the Perceptions and Practicalities Approach3 (PAPA™), which informs the NICE guidelines on adherence.4 This explains decisions as an interplay between intentional and non-intentional factors. Nudges typically remove practical barriers and making the optimal decision ‘easier’, when motivation to overcome practical barriers would otherwise not be sufficiently high. This is demonstrated in Figure 1a.

However, in cases where the non-adherence is intentional, there is motivation driving against adherence due to patients’ perceptions. Nudges will not address these perceptual factors and thus are less likely to be effective. The motivational drivers of the patients’ intention against adherence must also be addressed. (See Figure 1c.)

So how can we address the motivational drivers of intentional non-adherence? Studies analysing the patients’ perceptions of their disease and treatment are highly informative here. Patient perceptions, including those that they may not be extrinsically aware of, are accurate at predicting medicines taking behaviours. These perceptions can be understood as a trade-off between necessity beliefs and concerns. This has been shown, through a meta-analysis of over 25,000 patients covering 24 conditions in 18 countries, to be the most significant predictor of adherence.5

Patients can be individually mapped according to their level of necessity beliefs and concerns, as shown in Figure 2. Patients who have high concerns, but low necessity beliefs are sceptical, so have motivation not to adhere, whereas those with high necessity beliefs and lower concerns are accepting of their medicine. These patients have a greater positive motivation and are more likely to adhere.6 Increasing motivation for adherence requires either increasing patients’ necessity beliefs, or decreasing their concerns. That is, shifting patients towards the bottom right quadrant of the map.

An optimal approach to increase adherence must address the underlying perceptions that underpin a patients’ motivations and intentions to adhere, as well as making it easy to do. Therefore, in order for a nudge to be effective, perceptual barriers that relate to intentional must first be addressed.

Figure 1. Impact of nudges on adherence. (a) When patients have low motivation to adhere and not adhere, adherence may be low if not made easy to achieve. A nudge can increase adherence by making it easier. (b) When patients have high necessity beliefs they are more likely to be adherent, so the impact of a nudge on increasing adherence may be small. (c) Patients have greater concerns than necessity beliefs and are likely to be non-adherent. Nudges that make adherence seem cognitively and practically easier are unlikely to be effective if the concerns are not addressed.

Figure 2. Necessity beliefs and concerns perceptual map. Patients with greater necessity and lower concerns are most likely to be adherent.

References

1. Thaler RH, Sunstein CR. Nudge: Improving Decisions about Health, Wealth, and Happiness. Yale University Press; 2008.

2. Slomski A. Pill Reminders Don’t Improve Adherence. JAMA. 2017;317(24):2476. doi:10.1001/jama.2017.7588

3. Chapman SCE, Horne R, Eade R, Balestrini S, Rush J, Sisodiya SM. Applying a perceptions and practicalities approach to understanding nonadherence to antiepileptic drugs. Epilepsia. 2015;56(9):1398-1407. doi:10.1111/epi.13097

4. Medicines adherence: involving patients in decisions about prescribed medicines and supporting adherence | Guidance and guidelines | NICE. https://www.nice.org.uk/guidance/cg76. Accessed June 27, 2017.

5. Horne R, Chapman SCE, Parham R, Freemantle N, Forbes A, Cooper V. Understanding Patients’ Adherence-Related Beliefs about Medicines Prescribed for Long-Term Conditions: A Meta-Analytic Review of the Necessity-Concerns Framework. PLOS ONE. 2013;8(12):e80633. doi:10.1371/journal.pone.0080633

6. Mann DM, Ponieman D, Leventhal H, Halm EA. Predictors of adherence to diabetes medications: the role of disease and medication beliefs. J Behav Med. 2009;32(3):278-284. doi:10.1007/s10865-009-9202-y

Artificial intelligence, robotics, and nano-devices, among other rapidly expanding technologies have an overwhelming potential to shape many aspects of our not too distance future, not least our medical care.

With somewhat ease, we can now visualize a future with automated referrals, prescribing, monitoring and discharge – streamlining and synchronizing the care we receive.

However, it will be critical that such advances do not undermine what behavioural science and psychology have cemented over the last half century: that, contrary to standard economic theories, humans are not always rational. How we make decisions, such as whether to take a medicine, are complex and multi-factorial.

Consultations can often be key to ensure the optimal decision is reached. However, with automated services such as robotic drug dispensing, such human-human interactions will be lost.

Devices do not always have to be ‘unhuman’ though; many have been designed to be behaviourally smart. Such devices consider the irrational facets and individual complexities in human behaviour to improve outcomes and help people make better decisions for healthier, happier lives.

With the expanding accessibility of big data there is also a new opportunity to take advantage of, as quantitative behavioural insights and analysis can facilitate increasingly tailored and personalized support through medical devices.

Incorporating an understanding of human behaviour into artificial intelligence and other innovative medical devices will be vital in future healthcare, but done correctly could offer great potential to improve health outcomes.

How we perceive the implications of an illness and potential treatments in the present compared to the future can shape our attitudes towards treatments and thus influence adherence and outcomes.

We have a tendency to choose small short-term gains over long-term larger ones, which is described as temporal discounting. The value of an item today appears to be worth more than in the future.

This tendency to make choices which bias the present, often to our long-term detriment, is particularly prominent in smokers1 and can also be applied to medical adherence. Discounting in the value of future health risks, has been found to be correlated with adherence and treatment outcomes in both diabetes2,3 and multiple sclerosis.4,5

One potential reason why we favour the present could be linked to future-self continuity, how we perceive ourselves now, in comparison to ourselves in the future. This also links to illness perception. If the person does not feel ill, the benefit is not obvious, so it is difficult for patients to even discount.

In experiments, when participants are asked to assess how similar they perceived themselves to their future self, using the sets of circles (see below) and then undertake a temporal discounting task to see how likely they were to choose delayed monetary rewards (e.g. £15 today or £50 in three months), future-self similarity (assessed by the circle task) correlate with their likelihood of choosing delayed reward. The more similar people perceive themselves as more similar to their future self the more they save.7

When shown avatars of either in their current state or looking elderly as a retired version of themselves8 and asked questions such as: How much of your current income would you like to allocate for your retirement fund? Seeing the older avatar of themselves people increase the average percentage of their income they would choose to save.8

This has strong implications as it demonstrates that making a connection with our future self can help us make better, more forward-thinking plans, and patients make treatment choices more aligned with their long-term needs.

1. Bickel W, Odum A, Madden G. Impulsivity and cigarette smoking: Delay discounting in current, never, and ex-smokers. Psychopharmacology. 1999;146 (4): 447-454.

2. Lansing A, Stanger C, Crochiere R, et al. Delay Discounting and Parental Monitoring in Adolescents with Poorly Controlled Type 1 Diabetes. Journal of Behavioral Medicine. 2017

3. Lebeau G, Consoli M, Le Bouc R, et al. Delay Discounting of Gains and Losses, Glycemic Control and Therapeutic Adherence in Type 2 Diabetes. Behavioural Processes. 016;132: 42–48.

4. Bruce J, Bruce A, Catley D, et al. Being Kind to Your Future Self: Probability Discounting of Health Decision-Making. Annals of Behavioral Medicine: A Publication of the Society of Behavioral Medicine. 2016;50 (2): 297–309.

5. Jarmolowicz D, Reed D, Bruce A, et al. Using EP50 to Forecast Treatment Adherence in Individuals with Multiple Sclerosis. Behavioural Processes. 2016;132: 94–99.

6. Hershfield H, Wimmer G, Knutson B. Saving for the future self: Neural measures of future self-continuity predict temporal discounting. Soc Cogn Affect Neurosci. 2009;4 (1): 85-92.

7. Hershfield H, Garton M, Ballard K, et al. Don’t stop thinking about tomorrow: Individual differences in future self-continuity account for saving. Judgment and Decision Making. 2009;4(4): 280–286

8. Hershfield H, Goldstein D, Sharpe F, et al. Increasing Saving Behavior Through Age-Progressed Renderings of the Future Self. Journal of Marketing Research. 2011;48: S23–S37