Intervention Design

Introduction

The 2008 MRC framework (1) and other work (2, 3), emphasises the need for a strong theoretical foundation to inform both intervention development and evaluation. NPT can play a role in supporting this work.

Firstly, NPT may be adapted to guide the investigation of the feasibility of whole interventions in advance of their implementation, with the aim of providing a systematic and comprehensive mapping of the human, organisational and resource changes that an intervention will require. This is important work because the implementation of some interventions is likely to only be possible with major structural or organisational changes to healthcare delivery. For example, thrombolysis for acute ischemic stroke is likely to be effective in reducing later disability but the treatment needs to be given within a few hours of the first symptoms of the stroke. This might require a major change in the way acute stroke is managed and in the way clinicians treating stroke have to work. As one author writing about this treatment put it 'Some neurologists are now getting out of their beds in the middle of the night to go and see patients - an activity that would have been almost unheard of a short decade ago' (4).

Trials evaluating this sort of intervention might be called 'aspirational' because they assume a health care system that is rather different from the one that currently exists. It would be good to know up front that an intervention is aspirational because trialists could, for example, start working early on with policymakers and others to build support for making big changes if the intervention proves to be effective. NPT could help to identify aspirational interventions and allow trialists and others to better judge whether the required changes are feasible on a wide scale and whether the likely benefit of the intervention justifies making them. The way you would use NPT in this way will be much the same as for looking at intervention components i.e. use the NPT to generate qualitative data with representatives of those affected by the planned trial to consider normalisation issues

If the intervention if feasible, NPT could be used to aid the design of the individual components of a complex intervention. For example, there may be three potentially effective components to a particular complex intervention. A trialist could ask 'Should I include all of these components, or just some?'  The trialist could do a pilot study that empirically tests each component individually and in combination, which would be useful. However, potentially this would be time consuming and costly. This is where NPT comes in.

NPT could be used by the trialist prior to such a pilot study (or perhaps instead of if time and resources are limited) to consider each potential intervention component in light of the theory (see NPT Core Constructs). The theory could be used by the trialist to guide analysis about the likelihood of normalization of each of the three potential components.

Using NPT for this sort of work could be done in a brainstorming meeting, or through a series of focus groups, involving representatives of those who would be affected by the intervention. They would use their knowledge and experiences to consider the likelihood of normalization of the potential components of interest to the trialist. This is, of course, qualitative work and remember, NPT can be used as a framework for the qualitative methods with which you may already be familiar. See 'Qualitative research' for ideas on how to do this.

As a result of the brainstorming meetings or focus groups, representatives may highlight that a particular intervention component is very likely, for example, to fall down because it requires a lot of work to train people to follow the new system of practice. An intervention component that stands little chance of slotting into normal, routine care for such reasons is likely to be less appealing to policymakers considering use of the intervention in their region, or to clinicians who are looking to improve aspects of their own care provision than one that fits in more easily.

As mentioned above, if resources allow, predictions coming from this qualitative work could be tested in pilot work, which may now involve fewer intervention components than originally planned because some were rejected as a result of the NPT work.

Things to consider

  • Describe in detail the intervention that you are considering.
  • Describe in detail who will be delivering the intervention (e.g. the number, training, and experience of surgeons)
  • Map the components of NPT to your intervention. Could the design of the intervention be changed to improve its chances of becoming normalized? How feasible is the intervention given the current health care system, staff, training, resources etc?
  • Is your trial aspirational, or meant to evaluate an intervention that can slot into an existing healthcare system?

References

  1. Craig P, Dieppe P, Macintyre S, Mitchie S, Nazareth I, Petticrew M. Developing and evaluating complex interventions: the new Medical research council guidance. BMJ 2008:337:a1655 doi: 10.1136/bmj.a1655.
  2. Campbell NC, Murray E, Darbyshire J, Emery J, Farmer A, Griffiths F, Guthrie B, Lester H, Wilson P, Kinmouth AL.  Designing and evaluating interventions to improve health care.  BMJ 2007; 334: 455-459.
  3. Hardeman W, Sutton S, Griffin S, Johnston M, White A, Wareham NJ et al. A causal modelling approach to the development of theory-based behaviour change programmes for trial evaluation. Health Educ Res 2005.
  4. Gubitz G. The NINDS trials of thrombolysis in acute ischaemic stroke. Practical Neurology 2002; 2: 45-49.