What is 'prescriptive'?

balance shutterstock_1050269185%.jpg

The decision support provided in our Decision Annalisas - DecisionALs for short -is 'prescriptive'. What does this mean and why do we emphasise it?

We do so essentially to make very clear what DecisionALs offer and what they don't, what they are and what they are not.

To distinguish DecisionALs from other types of support on offer, we need to introduce three terms commonly used in the study of how decisions are made and how they might be improved.

The three approaches

Descriptive approaches focus on describing how decisions are made - by individuals or by organisations, communities and other groups.

Normative approaches focus on establishing, without reference to how decisions are made, the fundamental principles and processes that an ideal decision maker would implement. While normative principles can be derived from many sources, the normative principles are conventionally regarded as being the formal axioms and logical processes of Decision Theory.

But neither purely descriptive nor purely normative approaches can offer ways to improve decisions. Descriptive approaches, as such, merely describe what is happening in the real world and normative approaches, in saying how decisions ought ideally to be taken, pay little or no attention to the practicalities of the real world.

The only routes to improved decision making therefore involve prescribing practical changes to how decisions are made or supported. These routes are pursued within Prescriptive approaches which may be thought of as fitting somewhere between the normative and descriptive.

The possible bases

On what bases can such desirable - potentially real-world decision-improving - prescriptions be identified? There are two main possible bases.

One basis is the normative principles of Decision Theory and their reformulation in Decision Analysis. Decision Analysis is essentially the ideal processes of Decision Theory converted into processes that are practical, given the time, resource and cognitive constraints of the real world. We can call a prescription made on this basis analysis-based prescription and this is precisely what we mean when we call our TopicALs 'prescriptive'. They produce an opinion which reflects, as closely as practicable - for many reasons this may not be very close - the logical processes of an idealised decision maker.

Interestingly, the principles and processes underlying DecisionALs are endorsed by most people if they are asked how a decision should be made, even when they don't follow the principles and processes themselves in their day-to-day decision making. It is not unusual, moreover, to judge the decisions of others by this standard. In this respect the opinion emerging from our TopicALs might be thought of as that of another, idealised, you!

The other possible basis of prescriptions for improvement is description of the decision processes of expert decision makers on the one hand, compared with non-expert decision makers on the other. Identifying the differences between them that make the experts experts, leads to what can be called expertise-based prescription.

It should be noted however, that, in the world of decision support, the term 'prescriptive' is almost exclusively associated with our analysis-based approach. Those who favour the expertise-based approach rarely use the term ‘prescriptive’. Some even use it pejoratively. Most prefer to characterise themselves as operating within an essentially descriptive approach, which in many ways is true, even though, by definition, some element of prescription is necessary in order to distinguish experts from non-experts and good from not-good decision outcomes. In terms of the decision support they offer, their prescriptive content relates mainly to the information that should be known in order to make a good decision – a good decision being often equated with an informed decision. They basically assume most people are cognitively competent at information processing and can 'make up their minds' to arrive at a good decision, if they are in possession of good information and encouraged to think carefully about what matters to them.

The case for analysis-based prescription - and DecisionALs

Let’s be very clear. Despite being prescriptive, our DecisionALs in no way assume or imply either that most people are cognitively incompetent, or that support based on expertise-based prescription is useless!  They simply reflect our belief that an opinion reflecting analysis-based prescription can be a very valuable form of decision support.

So why is our approach not better known in healthcare, especially clinical medicine?  Essentially because expertise-based prescription has been virtually the only route to improvement considered professionally acceptable. This is reflected in the curricula of medical schools and in the apprenticeship basis of practice training, where any form of decision analysis is rare to non-existent.

We believe this will need to change – and will indeed change - since the analysis-based prescriptive approach has one compelling advantage in the provision of patient/person- centred care and genuinely shared decision making. In its multi-criteria form decision analysis provides a generic approach to all decisions. It does not require reasoning expertise and knowledge acquisition in the particular condition in order to follow and share expertise-based prescriptions. So long as expertise-based prescription is the basis of the patient-doctor conversation, patient empowerment will be a difficult if not impossible task. The Multi-Criteria Decision Analysis-based prescriptive approach implemented in DecisionALs allows the person/patient to input their preferences as criterion importance weights in a straightforward manner and to have them transparently integrated with the evidence and expertise of the doctor in probabilistic form. To read more on this, see the items 'Person Important Outcomes' and 'MCDA-based decision support'.