How to do it?
Guidance from the United Kingdom Medical Research Council on developing and evaluating complex interventions provides a suitable framework for deprescribing intervention development, from selecting intervention components and translating these into an intervention package, through to feasibility testing and evaluation.(4) Many of the behavioural strategies, sometimes described as behaviour change techniques (active components of the intervention) used in deprescribing are relevant for broader medicines optimisation approaches, and so components of interventions may be interchangeable. However when it comes to operationalisation of either process in practice, there may be differences.(5)
Considering how the intended target and scope reflect the reality of clinical practice is important for intervention development. Often patients may have more than one medication issue that requires intervention, and an accumulation of single interventions targeting various medication issues may increase workload for providers. This may also add to the substantial treatment burden among people with multimorbidity (i.e. the work required to manage their conditions), associated with delivery of care centred around many single conditions and disease guidelines. A suite of components as part of a complex intervention targeting multiple medicines may provide a more cohesive, and sustainable, approach than multiple single-medicine focused interventions. The acceptability of these approaches should also be considered; clinicians and patients may find efforts to optimise all medicines at the outset challenging, whereas starting with a single medicine may provide early success encouraging an incremental approach to optimise multiple medicines.
Targeting multiple medicines presents additional complexity to an intervention, such as in assessing what should be deprescribed (or optimised), and how to do it. An intervention may have to involve multiple criteria to identify medications to deprescribe or optimise, and a set of medication-specific deprescribing guidance. Alternatively, a more general approach relying on implicit judgement of the healthcare professional could be used, which will increase heterogeneity in intervention delivery. In contrast, a single medication target may allow a greater degree of specification of the approach to deprescribing or medicines optimisation, likely yielding greater intervention fidelity.
Where changes to multiple medications are required, the need to decide the priority, timing, and order of these changes will also increase the potential for variation in intervention delivery.(6) This may have implications for deciding when to measure outcomes in a study. The time to implement multiple medication changes may be highly variable between patients. For example, medications may require varying lengths of time to taper while deprescribing, and some medications may be deprescribed sequentially rather than all at once. Therefore, the study length of follow-up will need to be sufficient to allow all changes to be implemented, and for any effect on outcomes to occur.