Normalization Process Theory
A recent approach, Normalization Process Theory (NPT) (Murray et al., 2010), can be helpful in developing an intervention and purposively involving clinical or community teams or when using a pragmatic or embedded trial design. NPT identifies four factors to evaluate and those which indicate whether an intervention has potential to ultimately be implemented and normalized in a practice setting. Its first factor “coherence” refers to “sense-making” or whether an intervention is easy to describe, distinguishable from other interventions, fits with organization goals, and its benefits clearly understood. Evaluating coherence in the developmental phase of an intervention can provide immediate feedback to the investigative team as to how the intervention is perceived, how best to describe the intervention, and/or how to modify its delivery characteristics to enhance its acceptability.
The second NPT indicator is “cognitive participation,” which refers to whether users of a program consider the intervention as a good idea and has the potential of making a positive difference. This serves as an indicator of whether stakeholders and interventionists will be invested in the program and motivated to make it work. If the intervention is perceived as too time-consuming and not particularly beneficial or not providing added value to a practice, then commitment to it will be low and implementation and sustainability threatened.
The third NPT factor is “collective work,” or how the program will affect the work of agency or clinical staff, whether the program promotes or impedes their work, whether extensive training is required, and whether the program is compatible with existing work flow. As most interventions will require a practice setting to change its practices and/or work flow, understanding the demands of an intervention would be important to identify and articulate upfront in developmental and testing phases. For example, changes might need to include the establishment of screening, referral, and follow-up mechanisms, supervision, and adherence to fidelity of the intervention.
The fourth consideration is “reflexive monitoring” or how users of the program will perceive it: what kind of supports will be needed for its integration into a context and/or what booster trainings may be needed.
Evaluating an intervention along these dimensions early on when developing the delivery characteristics or evaluating its outcomes can help inform how best to tweak the design of the intervention, and the type of training and supportive structures that will be necessary for the successful integration of the program in a practice setting.
Table 19.2 lists only a few of the many available frameworks in implementation science that can provide guidance in developing behavioral intervention studies that have as the goal to facilitate and maximize translation into practice. It is essential to have some familiarity with the elements of these models since one model may not encompass the entire process and impact of implementation or be relevant across all interventions. For example, the PARiHS model does not specify how implementation should be planned, organized, or scheduled (Helfrich et al., 2010). Rogers’s (1962) Diffusions of Innovations model emphasizes collaborative learning across an organization, but does not specify the process for accomplishing this aim. Therefore, it may be necessary to draw upon different but complementary theories or models or to select elements from relevant models to guide an implementation project.