'A priori' external contextual factors and relationships with process indicators: a mixed methods study of the...

It is widely recognised that complex public health interventions roll out in distinct phases, within which external contextual factors influence implementation. Less is known about relationships with external contextual factors identified a priori in the pre-implementation phase. We investigated which external contextual factors, prior to the implementation of a community-centred approach to reducing alcohol harm called 'Communities in Charge of Alcohol' (CICA), were related to one of the process indicators: numbers of Alcohol Health Champions (AHCs) trained.

A mixed methods design was used in the pre-implementation phase of CICA. We studied ten geographic communities experiencing both high levels of deprivation and alcohol-related harm in the North West of England. Qualitative secondary data were extracted from pre-implementation meeting notes, recorded two to three months before roll-out. Items were coded into 12 content categories using content analysis. To create a baseline 'infrastructure score', the number of external contextual factors documented was counted per area to a maximum score of 12. Descriptive data were collected from training registers detailing training numbers in the first 12 months. The relationship between the baseline infrastructure score, external contextual factors, and the number of AHCs trained was assessed using non-parametric univariable statistics.

There was a positive correlation between baseline infrastructure score and total numbers of AHCs trained (R_s = 0.77, p = 0.01). Four external contextual factors were associated with significantly higher numbers of lay people recruited and trained: having a health care provider to coordinate the intervention (p = 0.02); a pool of other volunteers to recruit from (p = 0.02); a contract in place with a commissioned service (p = 0.02), and; formal volunteer arrangements (p = 0.03).

Data suggest that there were four key components that significantly influenced establishing an Alcohol Health Champion programme in areas experiencing both high levels of deprivation and alcohol-related harm. There is added value of capturing external contextual factors a priori and then testing relationships with process indicators to inform the effective roll-out of complex interventions. Future research could explore a wider range of process indicators and outcomes, incorporating methods to rate individual factors to derive a mean score.

ISRCTN81942890, date of registration 12/09/2017.