intervention participants. Effectivenessrelates to the range of effects an intervention
might have. For example, even if it changed behaviour – did it enhance overall quality
of life or have any unintended consequences (e.g. did participants find it onerous or
upsetting)? Adoption refers to whether the users (e.g. nurses, teachers, managers,
members of the public) are persuaded of the utility of the intervention and use it.
Adoption depends on how easy the intervention is to implement, whether those who
will deliver it and/or their clients like and value it and whether it is compatible with
their other main goals (Paulussen, Kok and Schaalma, 1994). Since cost is important
to most people, interventions are unlikely to be used if adopters cannot afford them.
Understanding this adoption and diffusion process is critical to the overall impact of
any intervention (Rogers, 2003). Even effective interventions have no impact on public
health if they are not adopted and translated into everyday practice. Implementationrefers
to the ease and feasibility of faithful delivery. If an intervention is complex, expens -
ive or requires specialist training or teams of people to deliver it then it is less likely
to be sustainable in real-world settings so it may not be adopted or, if adopted, may
be delivered with poor fidelity. Maintenancerefers to the longer-term sustainability of
the intervention in real-world settings. For example, if an organization or community
does not have the resources to deliver an intervention then, no matter how effective,
it will be dropped over time. Similarly, if implementation problems are encountered
then even if the intervention is retained it may be changed and adapted to the setting,
which may mean altering or dropping change techniques critical to its initial effect -
iveness, so rendering it ineffective. These practical, real-world considerations are critical
to translating potentially efficacious interventions into sustainable enhance ment of
routine health preventive and health care services. They need to be carefully considered
in stages 1 and 5 of the intervention mapping process.
Contextual factors may include an individual’s characteristics, family, social network
or organization (such as a school or community). Interventions may be differentially
effective for different groups (e.g. men versus women, older or younger people or those
high in conscientiousness versus those low in this trait). Such between-group analyses
of effectiveness are called moderation analyses and can clarify for whom an intervention
is most effective. Such analyses require large samples to provide the statistical power
to identify group differences.
Process evaluations typically use mixed methods to answer multiple research
questions about how an intervention operates. For example, quantitative data may be
used in mediation and moderation analyses. Qualitative methods can be used to explore
processes in detail, such as participant perceptions of interventions and how the
intervention and its context interact with each other. If problems arise such as poor
implementation or low recruitment levels, interviews with intervention staff or
participants could be conducted to investigate the underlying reasons.
A useful distinction can be drawn between the function of process evaluations in
relation to (1) pilot and feasibility studies and (2) definitive trials. In the former, findings
from a process evaluation can be used to improve intervention development and
redesign research methods, including outcome measures. In the latter the process
evaluation should be conducted in a manner that does not alter the intervention
or contaminate the outcome evaluation. Of course, the results of a process evaluation
may necessitate intervention redevelopment, for example when an intervention is
found not to alter targeted change processes. Close collaboration between process and
218 MOTIVATION AND BEHAVIOUR