Cook et al. BMC Public Health (2018) 18:522
https://doi.org/10.1186/s12889-018-5410-0
STUDY PROTOCOL
Open Access
Communities in charge of alcohol (CICA): a
protocol for a stepped-wedge randomised
control trial of an alcohol health champions
programme
Penny A. Cook1* , Suzy C. Hargreaves1, Elizabeth J. Burns2, Frank de Vocht3, Steve Parrott4, Margaret Coffey1,
Suzanne Audrey3, Cathy Ure1, Paul Duffy5, David Ottiwell6, Kiran Kenth7, Susan Hare8 and Kate Ardern9
Abstract
Background: Communities In Charge of Alcohol (CICA) takes an Asset Based Community Development (ABCD)
approach to reducing alcohol harm. Through a cascade training model, supported by a designated local
co-ordinator, local volunteers are trained to become accredited ‘Alcohol Health Champions’ to provide brief
opportunistic advice at an individual level and mobilise action on alcohol availability at a community level.
The CICA programme is the first time that a devolved UK region has attempted to coordinate an approach
to building health champion capacity, presenting an opportunity to investigate its implementation and impact
at scale. This paper describes the protocol for a stepped wedge randomised controlled trial of an Alcohol
Health Champions programme in Greater Manchester which aims to strengthen the evidence base of ABCD
approaches for health improvement and reducing alcohol-related harm.
Methods: A natural experiment that will examine the effect of CICA on area level alcohol-related hospital
admissions, Accident and Emergency attendances, ambulance call outs, street-level crime and anti-social
behaviour data. Using a stepped wedged randomised design (whereby the intervention is rolled out sequentially
in a randomly assigned order), potential changes in health and criminal justice primary outcomes are analysed using
mixed-effects log-rate models, differences-in-differences models and Bayesian structured time series models. An
economic evaluation identifies the set-up and running costs of CICA using HM Treasury approved standardised
methods and resolves cost-consequences by sector. A process evaluation explores the context, implementation and
response to the intervention. Qualitative analyses utilise the Framework method to identify underlying themes.
Discussion: We will investigate: whether training lay people to offer brief advice and take action on licensing decisions
has an impact on alcohol-related harm in local areas; the cost-consequences for health and criminal justice sectors,
and; mechanisms that influence intervention outcomes. As well as providing evidence for the effectiveness of this
intervention to reduce the harm from alcohol, this evaluation will contribute to broader understanding of asset based
approaches to improve public health.
Trial registration: ISRCTN 81942890, date of registration 12/09/2017.
Keywords: Alcohol, Public health, Asset based community development, Brief intervention, Licensing, Dark logic,
Community-based prevention
* Correspondence: P.A.Cook@salford.ac.uk
1
School of Health Sciences, University of Salford, Manchester, UK
Full list of author information is available at the end of the article
© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Cook et al. BMC Public Health (2018) 18:522
Background
Reducing alcohol related harm continues to be a global
public health priority, with 5.9% of all global deaths and
5.1% of all disease and injuries being attributable to alcohol [1]. Excessive alcohol consumption harms an individual’s health and social relationships [2]. It also harms
society more generally, as urban areas can become less
pleasant and less safe to visit [3] and crime may increase
[4]. Moreover, the consumption of alcohol contributes
significantly to health inequalities. Those living in deprived areas drink the same average quantity of alcohol
as those from more advantaged groups. However, a socalled ‘alcohol harm paradox’ exists whereby, for a given
level of alcohol consumption, alcohol harm is higher
amongst those living in more deprived areas [5]. Possible
reasons for this include patterning of consumption (e.g.
consuming the same quantity on fewer occasions [6])
and combinations of other health risks (e.g. smoking,
obesity) in individuals living in deprived areas [7]. Interventions that are effective at reducing alcohol harm have
been shown to operate at the individual level (i.e. brief advice about drinking [8]), the community level (e.g. licensing policies [9, 10] and measures that control access to
alcohol [8]), and national level (e.g. alcohol pricing policy
[2]). This protocol describes an evaluation of an intervention, ‘Communities In Charge of Alcohol’ (CICA), which
aims to target alcohol harm at two levels; by influencing
individuals (through brief intervention) and communities
(through reducing the availability of alcohol).
Brief interventions and brief advice have been shown
in systematic reviews and meta-analyses to be effective
in a variety of settings including emergency departments
[11] and primary care [8]. There is relatively little evidence about training lay persons for this role, although
pilot work with ex-offenders giving advice to offenders
in community settings seems promising [12]. Accessibility of alcohol is a key determinant of harm [5, 13, 14].
Internationally, systematic review evidence shows that
high alcohol outlet density is linked to higher levels of
crime and poor health [15]. A systematic review (rated
high quality [16]) found that higher outlet density and
greater exposure to advertising tends to be associated
with higher levels of alcohol use [17]. Interventions that
change the alcohol environment thus have the strongest
evidence for effectiveness [2, 16, 18].
In England, local authorities can address public health
through licensing policies. However, because ‘public
health’ is not currently one of the licensing objectives,
the extent to which this is carried out varies across the
country [9]. Local people have the ability to influence
the availability of alcohol via the licensing process, but
do not tend to do so, due to low awareness and lack of
confidence that local views will be valued [19]. Recent
longitudinal, area-level analysis of UK datasets shows
Page 2 of 11
that, at borough level (i.e. lower tier local authorities in
England) both alcohol-related hospital admissions [9]
and crime [10] have reduced faster in areas where more
restrictive licensing policies are in place. Using small
area level data (lower super output areas [LSOA1] with a
mean population of 1500 persons), alcohol outlet density
in Wales is similarly associated with alcohol-related hospital admissions and crime data [5]. The possibility of
utilising community advocacy to reduce the availability
of alcohol has not been explored but could represent an
untapped resource. This aligns to the UK Government’s
attempts to make it easier for residents and other local
agencies interested in licensing to take action [20].
The CICA programme takes an Asset Based Community Development (ABCD) approach, where a health
asset is any factor which enhances the ability to create
or sustain health and wellbeing [21]. This is in line with
the National Institute for Health and Care Excellence
(NICE) guidance on behaviour change, which advocates
building on existing community resources and skills
[22]. The principles of the approach are to: allow time
for communities to realise and acknowledge their individual and collective assets and to rebuild their confidence and networks; enable local people to take the lead,
and; build trust with communities by demonstrating that
involvement leads to change [23]. The approach seeks to
build community networks, which are health promoting
[24]. The programme aims to enable local volunteers
working with the community to identify alcohol harm in
their community, and facilitate them to intervene to reduce these harms.
Methods/design
The overarching aim of this research is to evaluate the
effectiveness of a community Alcohol Health Champions
programme (Communities in Charge of Alcohol: CICA)
to reduce alcohol-related harm. Secondary aims include
determining the cost consequences of CICA and exploring the context, barriers and facilitators to its implementation. As a complex community intervention, it is not
amenable to conventional randomisation (as recognised
in the complex interventions guidance [25]) and will
be evaluated as a natural experiment [26]. This fits
on the ‘continuum of evaluation’ [27], which recognises the need for multiple methods/variants on experimental design [28].
Through CICA, individuals embedded within areas
identified as having high levels of alcohol-related harm
are recruited by local co-ordinators to take part in a 2
day ‘Alcohol Health Champion’ (AHC) course. As a
community asset connected to the area either through
their residency or their work role, an Alcohol Health
Champion’s existing strengths, motivations and skills are
strengthened to enable them to (i) give alcohol-related
Cook et al. BMC Public Health (2018) 18:522
brief advice to others and (ii) tackle the availability of alcohol in the local environment through the licensing
process. Using a cascade training model, an accredited
and standardised training course is delivered initially by
the Royal Society for Public Health (RSPH) and then
cascaded by the Alcohol Health Champion community
with the support of professionals. This community hub
approach means that AHCs have an infrastructure of
support provided by local co-ordinators and local licensing officers. The AHC role description specifies that
champions should only do whatever they are comfortable with and that the programme does not dictate
where, when or how much activity should take place.
This is largely based on the existing health champions
model which utilises lay health workers from a variety of
backgrounds (e.g. voluntary sector, housing, local residents) to work in a voluntary capacity to offer brief advice and brief interventions alongside their other daily
activities [29]. However, for the first time, through
CICA, champions will be trained to focus on alcohol
and will receive additional knowledge and skills to enable them to get involved in local licensing decisions.
A logic model was created as part of the planning of
this evaluation (Fig. 1). This shows the intervention’s
mechanisms of action and the interplay between its core
components. At the heart of the CICA programme,
based on ABCD principles, is the assumption that individuals and communities have strengths, motivations
and skills that benefit everyone. Further, there is an assumption that the AHC training programme and infrastructure of support can help build the strengths,
motivations and skills of these individuals to develop
confidence to put their skills into practice. As communities take action by offering brief advice or getting involved in licensing decisions, a feedback loop illustrates
how such success positively reinforces the strengths,
motivations, and skills of the community. Influencing access and availability to alcohol and building a groundswell of brief advice about alcohol should, as a result,
directly impact on alcohol related outcomes.
Additionally, we also considered the possible unintended consequences of the intervention (see Fig. 2).
Few public health interventions and evaluations explicitly look at unintended harms and, although logic
models considering positive consequences and outcomes
are common, the consideration of the potential negative
outcomes (and their mechanisms of action), or a ‘dark’
logic model, are less common [30]. According to Bonell
et al. [31], not only is it important to produce a dark
logic model ahead of the evaluation/intervention, or during it, but also the framework could be useful to evaluate
the project retrospectively to see how it might have been
strengthened. To develop the dark logic model we first
created a matrix to hypothesise a priori the potential
Page 3 of 11
unintended consequences of CICA using a simple
framework, adapted from Lorenc and Oliver’s five categories of harm (see Table 1) [30]. Each potential harm
was reflected upon using one of three approaches recommended by Bonell et al. [31]: how agencies and structures may interact in unintended ways; comparative
understanding across similar interventions; and consultation with individuals/groups with insights into local
contexts and how interventions might operate within
them. Evidence that supports or refutes the hypothesised
unintended consequences of the intervention will benefit
future design and minimise the risk of future harm [31].
Setting
All Greater Manchester (GM) boroughs within the study
setting have higher than England averages for alcoholrelated mortality, ranging from 46.7 in Trafford to 71.9
per 100,000 in Manchester [32]. GM is heterogeneous in
terms of its application of licensing policy: in a recent
national study, only one local authority was classified as
having high licensing policy intensity (two local authorities were medium, five low and two passive in terms
making use of cumulative impact areas and/or declining
licences [9]).
CICA will be rolled out sequentially across specifically
targeted areas in all 10 local authorities, with the sequence randomly assigned by the research team, so that
it will eventually (within the timespan of a year) be delivered in all areas. The intervention areas themselves
within each local authority will be formed around preexisting communities in LSOA locations. For data analysis purposes, the chosen LSOAs that represent each
community combine to make the ‘intervention area’ unit
of analysis. For example, the smallest intervention area
encompasses one LSOA (mid-year population estimate
of 1648) and the largest contains three LSOAs (mid-year
population estimate combined of 5586).
Characteristics of study population
In order to ensure there is consistency for the evaluation, each local authority used the following guiding
principles as inclusion criteria for selecting an intervention area dependent on local available data:
An area of high alcohol-related harm (defined as
high within the local authority, rather than in
comparison to regional or national average rates)
Alcohol harm considered in terms of a combination
of indicators:
– Alcohol-related crime2 and anti-social behaviour
(ASB)3
– Alcohol-related hospital admissions
Cook et al. BMC Public Health (2018) 18:522
Page 4 of 11
Fig. 1 CICA Evaluation Logic Model
– Weekend evening Accident and Emergency (A&E)
attendances
– Users of local treatment services
– Hospital recording of location of violent incidents (if
available)
– Density of licensed premises in the area or adjoining
areas (if available)
The study population of AHCs will include adults aged
18 years and over, who are already embedded in the
intervention area either through their residency or their
work role (i.e. they must spend the majority of their time
in the intervention area).
Outcome evaluation
The effectiveness of CICA will be measured through the
following data and objectives:
a) Routinely collected quantitative data to determine
the effect on key health performance indicators
(narrow indicator of alcohol-related hospital
admissions,4 A&E attendances and ambulance
call-outs);
b) Routinely collected street-level crime data to
determine the effect on key crime indicators;
c) Routinely collected anti-social behaviour data to
determine the effect on key ASB indicators.
Cook et al. BMC Public Health (2018) 18:522
Page 5 of 11
Fig. 2 CICA “Dark” Unintended Consequences Logic Model
The aims and objectives of the outcome evaluation will be
reached through outcome analysis at the level of the small
intervention area (the equivalent of one to three LSOAs)
comparing areas where CICA has not yet been introduced.
The outcome analysis involves two distinct approaches:
a) ‘Internal’ evaluation: Trends in area-time
intervention areas will be compared before and
after the intervention using a stepped-wedge
randomised trial design [33]. Sensitivity analyses
using different lagging periods (6–24 months)
between introduction of the intervention and
expected effects will allow for a delayed effect on
output measures. If differences in the slopes of the
longitudinal models are observed, the population
impact will be estimated from deviation of the
post-intervention slope compared to the preintervention slope.
b) ‘External’ evaluation: Secondly, the impact of the
intervention will be assessed using two
Page 6 of 11
Cook et al. BMC Public Health (2018) 18:522
Table 1 CICA Matrix of potential unintended consequences
Potential unintended
consequences
How agencies and structures may
interact in unintended ways
Comparative understanding across
similar interventions
Consultation with individuals/groups
with insights into local contexts and
how interventions might operate within
them (CICA Project Advisory Group)
Direct harms
None identified
Lack of depth of knowledge by lay
health advisors could result in time
delays or inconsistent advice for
‘in-need’ populations [48]
Concerns that volunteers recruited from
recovering communities could be at
increased risk of relapse of alcohol, drug
or mental health problems
Psychological harms
None identified
Volunteers embedded within
communities find it hard to ‘switch
off’ [48]
Intervening in licensing could lead to
negative reactions from local retailers
Dissatisfaction and disillusionment
of volunteers [49]
Equity harms
Communities most in need are
probably the least able to form
a strong community group [50, 51]
Motivated individuals becoming
health champions are likely to benefit
from being a champion more so than
those less motivated (who need the
potential positive benefits more) [48]
Individual assets within communities
excluded from participation due to
barriers to recruitment/participation e.g.
literacy, criminal record checks, worry
about impact on benefits
Group and social harms
‘Communities’ chosen to be in
charge of alcohol set by experts
(normative needs) vs. self-identified
communities (felt needs) [49]
Becoming a community champion
could result in lack of acceptance by
own community resulting in
marginalising “do gooders” [48]
None identified
Current recovery dominated culture
within alcohol service provision in UK
influences the selection of volunteers
from ‘recovery’ communities [52]
Opportunity cost harms
Commissioners may miss
opportunities to invest in alternative
public health interventions [53]
Missed opportunities to identify “at-risk”
populations [54] due to stereotyping
those ‘in need’ as only the most severe
drinking patterns [55]
complementary methods: (i) matching intervention
and control areas inside the GM area by area-level
deprivation, population size, age distribution and
baseline alcohol-related burden by calculating
propensity scores [34]. Temporal trends in each of
the outcomes will be plotted graphically and
analysed using hierarchical growth models (similar
to de Vocht et al. [9, 10]); and (ii) using time series
data from GM LSOAs to construct weighted
‘synthetic control time series’ [35] that mimic the
‘intervention area’ as closely as possible prior to the
introduction of CICA. Modelled and measured
post-introduction time series will then be compared
directly to quantitatively estimate the impact of
CICA. Both methodologies (i and ii) are
complementary, and while the latter approach
has the distinct advantage that it provides a
direct comparison to the counterfactual timeseries, it can be considered as less insightful
than the former method because it does not
compare actual areas on the ground.
The external evaluation may be affected by ‘spill-over’.
In other words, if, as a result of the introduction of
CICA, it becomes more difficult for new premises to obtain a premises licence in the specific intervention areas,
None identified
these may decide to establish themselves in neighbouring areas (close to the border). Conversely, AHCs could
get involved in licensing decisions in these neighbouring
areas directly. These are known ‘spill-over’ issues [36],
but difficult to tackle, and therefore directly neighbouring LSOAs will not be incorporated into control areas.
Instead, only LSOAs that are further away are combined
and matched, as outlined above, using propensity score
matching or incorporated in the synthetic control time
series.
Statistical analysis and power calculation
Statistical power calculations were based on the methodology for stepped-wedge randomised trials outlined in
Hussey and Hughes [33]. These were calculated specifically for the primary outcome ‘alcohol-related hospital
admission rates (narrow)’ obtained from the Local
Alcohol Profiles for England (LAPE)4 extracted for all 10
GM local authorities. Statistical power analyses were
conducted at local authority level rather than at the level
of the intervention area (the equivalent of one to three
LSOAs) because the exact areas and comparisons had
not been determined, but this aggregated level provided
indications within the stepped wedge context. The mean
standardised alcohol-related hospital admission rate in
these local authorities for the year 2014 was 207 (per
Cook et al. BMC Public Health (2018) 18:522
100,000 people) with a maximum temporal standard deviation per site of 17.2 (range within sites 5–17) and a
coefficient of variation across sites of 4.35. With 10 areas
and 12 month follow-up (i.e. when all areas have received the intervention and a minimum of 1-month
post-intervention follow-up), and a statistical significance level of 5% and statistical power of 90%, the proposed study will be able to detect a 10% average
difference in rates compared to baseline. For an intervention to be effective and cost-effective a minimal reduction in key indicators of 10% seems reasonable.
Statistical analyses will use standard mixed-effects
models with an indicator of when the intervention was
introduced in each area and a time component to account for the repeated measures nature of the data.
With respect to the comparison with propensity
matched controls and synthetic controls in which time
trends will be compared within the larger area of between different areas, no additional clustering occurs.
Assuming a standard comparison of independent means,
1-sided test, and significance level of 5%, changes in
alcohol-related hospital admission rate in the intervention LSOA, relative to the selected comparison area,
yield an 84% statistical power to detect a similar 10% decrease. Trends in matched areas are evaluated using
mixed-effects log-rate models as previously used at
lower-tier local authority area level for alcohol-related
hospital admissions [9] and alcohol-related crime rates
[10] in England.
To create the synthetic controls Bayesian structural
time series methods will be used [37, 38]. These ‘synthetic areas’ will be based on weighted averages of other
GM local areas, where the weights are chosen so the
synthetic GM area most closely resembles the actual
GM area before the intervention started [39]. Trends between the measured and the modelled, counterfactual,
outcomes in the synthetic controls (generated using
Bayesian structural time series) will be interpreted as the
intervention effect.
In these statistical power calculations we have not
taken into account any potential ‘spill-over’ effects, such
as described for a community action programme in
Sweden [40], and which implies that the above may be
an underestimation of the true statistical power (or, conversely, of the minimal detectable effect). It is unclear
how these should be modelled, and therefore, as outlined above, no LSOAs immediately adjacent will be
matched.
Page 7 of 11
a) Identifying set-up and running costs using a
standardised costing exercise (examination of
commissioning documents and contracts);
b) Resolving costs by sector (health, ambulance and
police) before, during and after the CICA set up;
c) Quantifying benefits due to reduced hospital
admissions, ambulance call outs, emergency
department use, crime and anti-social behaviour.
The costs of training, delivery and support elements of
the intervention will be estimated. The economic evaluation will build on the outcome evaluation by attributing
costs to the health performance indicators collected on
primary outcomes. UK Treasury approved methods
published by New Economy will underpin the costconsequences analysis (CCA) and unit costs are based
on the New Economy Unit Cost Database [41].
A standard costing exercise will use documents and
contracts to identify resources and costs required to deliver the CICA intervention in each local authority. Standardised methods allow comparability of costs. The
economic evaluation follows a cost-consequences analysis which is an approach that is favoured when costs
and outcomes fall on a range of budget-holders and government agencies enabling cost and consequence domains to be presented in a disaggregated form [42, 43].
This will enable decision makers to assess results using
different relevant perspectives.
The key cost categories identified include the set up
cost for the intervention area, comprising staff costs,
consumables and overheads (premises). In terms of consequences, this comprises an analysis of health benefits,
changes in health care resource utilisation as a consequence of alcohol use (A&E attendances, hospital inpatient stays, ambulance service costs), and changes in
contacts with the criminal justice system.
The outcome analysis uses mixed-effects log-rate
models and differences-in-differences models to evaluate
changes relative to propensity-score-matched controls
and will use Bayesian structural time series to model the
synthetic control areas; both to assess and compare potential changes in the health care and criminal justice resources before and after CICA interventions. The
economic component of the study will follow the same
statistical methods used in the outcome analysis and applies unit costs to the resource indicators to derive costs
for each domain. These costs will be presented in a
CCA framework, disaggregated in terms of costs to individual stakeholders and different cost domains.
Economic evaluation
Process evaluation
The economic evaluation aims to conduct a cost consequences analysis of the CICA programme. It will do this
through:
The design of the process evaluation was informed by
MRC guidance for conducting process evaluation of
complex public health interventions [28]. The aim is to
Cook et al. BMC Public Health (2018) 18:522
explore the factors that enable or hinder the implementation of the intervention. This includes establishing,
operationalising and sustaining the CICA intervention.
The objectives of the process evaluation are as follows:
a) Exploring the policy context and variation in licensing
practice, including any impact of devolution in Greater
Manchester;
b) Explore barriers/facilitators at key stages of the
intervention (recruitment of AHCs to initial
training and cascade training, delivery of initial
training and cascade training, using skills beyond
the training in AHC activity; retention of AHCs);
c) Explore response to AHC training, modelling of
health behaviours, perceptions of community
cohesion and development;
Fig. 3 Charted summary of process evaluation methods
Page 8 of 11
d) Determine the numbers of trainees, brief
interventions applied and community
awareness events organised/participated in;
e) Examine and quantify the amount and success of
community involvement in licensing issues;
f ) Determine whether there is a change in composite
measures of alcohol availability.
The aims and objectives of the process evaluation will
be achieved using mixed methods to examine the context, acceptability, facilitators and barriers to the intervention (see Fig. 3). Appropriate analysis techniques
have been selected for each method. Documents will be
analysed using content analysis, and quantitative data
from documents (e.g. on numbers of licence applications, reflective diaries completed by AHCs) will be
Cook et al. BMC Public Health (2018) 18:522
extracted and described. Interviews and focus groups
will be digitally recorded, fully transcribed and anonymised. Analysis will utilise the Framework method [44]:
textual data will be ‘charted’ in themes relating to key research questions and scrutinised for differences and
similarities within themes, keeping in mind the context
in which these arise. We will ‘sense check’ our emerging
qualitative findings with stakeholders, who will help us
with the interpretation of our themes.
Questionnaire data with AHCs (at baseline and 1 year
follow-up) and data on numbers of licences challenged
will be analysed using descriptive statistics (SPSS v23).
Data will be analysed to construct the most robust and
plausible explanation of observed outcomes. The logic
model, programme theory and ‘dark logic’ model may be
modified in the light of study findings.
Reflective diaries will be completed by AHCs who consent, although the detail will be as much or as little as the
AHC is willing to provide. The benefits of completing a reflective diary include supporting the AHC to learn from
their experiences by looking back on facilitators and barriers to carrying out their roles beyond the initial CICA
training [45] and for the evaluation team in understanding
the experiences of the AHCs over the intervention period.
Reflective diaries have been used successfully in other public health interventions, for example, helping peer supporters in a school-based stop smoking intervention:
however, it was acknowledged that the information provided was not a full and accurate indication of the peersupporters’ conversations and interactions [46]. The
research team evaluating the CICA project recognise that
reflective diaries cannot be used for monitoring the actions
of the AHCs and rather aim to give a flavour of the activities
and the reflections of “being an Alcohol Health Champion”
over the intervention period. This will then be followed up
with a sample of more in-depth interviews where experiences can be explored in greater depth, whether or not
these experiences are recorded in the diaries.
Discussion
The ABCD approach is currently being promoted widely
(e.g. in new NICE guidance [29]) and is attractive in terms
of current fiscal challenges and cuts to services, but there
is relatively little evidence for its effectiveness [47]. Therefore, we anticipate the findings will be widely relevant
across a range of topics, not just interventions to reduce
alcohol harm. The results will be of interest to policy
makers, commissioners and public health practitioners responsible for reducing alcohol harm at both a population
and individual level. Evidence on the context within which
community participation programmes for health improvement are implemented will help increase knowledge about
their mechanisms of action and potential inaction. Evidence on the cost-consequences of CICA will quantify its
Page 9 of 11
set up and running costs and whether there are any benefits to reduced hospital admissions, ambulance call outs,
emergency department use, and crime.
As a pragmatic study evaluating a community response to a training and support programme, there is
likely to be variation in the degree to which the intervention is applied, and the exact nature of the activities
in each area. However, the scale of the intervention and
the methodology, which includes an element of randomisation and an evaluation of promising new methods for
analysing natural experiments, alongside the health economic analysis, are significant strengths of this study.
The process evaluation and analysis of the logic and dark
logic models will help us to evaluate the specific context,
the particular actions or responses involving a given set
of actors, and how these are responsible for generating
given outcomes. These strengths should enable a comprehensive assessment of the CICA programme.
Endnotes
1
Lower super output area (LSOA) is a unit of
measurement generated for small area statistics; they are
consistent in size of population, reflect the characteristics of census data where possible and are stable
boundaries https://www.ons.gov.uk/methodology/geography/ukgeographies/censusgeography#output-area-oa,
accessed November 2017. http://www.datadictionary.
nhs.uk/data_dictionary/nhs_business_definitions/l/lowe
r_layer_super_output_area_de.asp?shownav=1, accessed
November 2017.
2
Crime data relates to those events where a crime (in
law) has occurred, and where this is included within the
Home Office Counting Rules for Recorded Crime (i.e. a
‘notifiable’ crime as defined within the National Crime
Recording Standard), Home Office (2017) Counting
Rules for Recorded Crime, https://www.gov.uk/government/publications/counting-rules-for-recorded-crime,
accessed January 2018.
3
Anti-social behaviour data relates to those events reported to the police where a crime has not occurred, but
where the police receive a call for service and the recorded incident is classified by the call handler as involving anti-social behaviour, in accordance with national
guidance, Home Office (2011) National Standard for
Incident Recording (NSIR), https://www.gov.uk/government/publications/the-national-standard-for-incident-recording-nsir-counting-rules, accessed January 2018
4
The narrow indicator of alcohol-related hospital admissions measures the number of individuals admitted to hospital due to a primary diagnosis with an alcoholattributable code or any secondary diagnosis with an external alcohol-attributable code, Public Health England (2015)
Local Alcohol Profiles for England user guide 2015, http://
Cook et al. BMC Public Health (2018) 18:522
webarchive.nationalarchives.gov.uk/20171107173500/
http://www.lape.org.uk/downloads/LAPE%20User%20Guide_Final.pdf, accessed December 2017
Abbreviations
A&E: Accident and emergency; ABCD: Asset based community development;
AHC: Alcohol health champion; ASB: Anti-social behaviour; CCA: Cost
consequences analysis; CICA: Communities in charge of alcohol; GM: Greater
Manchester; GMCA: Greater Manchester combined authority; HES: Hospital
episode statistics; LAPE: Local alcohol profiles for England; LSOA: Lower super
output area; NICE: National Institute of Health and Care Excellence;
NIHR: National Institute for Health Research
Acknowledgements
We would like to thank Dr. Jan Hopkins (Greater Manchester Combined
Authority) for project management of the CICA intervention and the
independent Study Steering Committee for useful input into the protocol.
Funding
This project was funded by the NIHR Public Health Research (NIHR-PHR-15129-03). The views expressed are those of the authors and not necessarily
those of the NIHR or the Department of Health.
Authors’ contributions
PAC, MC, SA, FdV, SP, DO and PD made substantial contributions to the study’s
conception and design. KA, KK, and PD made substantial contributions to the
design of the intervention. SH made substantial contributions to the
development of the ethical considerations relevant to the study by
sharing experiential knowledge. PAC wrote the first draft of the original
study protocol and SCH, EJB and CU converted this into a paper. All
authors contributed to successive drafts. All authors read and approved
the final manuscript and agreed to be accountable for all aspects of the
work in ensuring that questions related to the accuracy or integrity of
any part of the work are appropriately investigated and resolved.
Ethics approval and consent to participate
Ethical approval for the study was obtained from the University of Salford
Research Ethics Committee on 17/05/17 (reference number: HSR1617–135).
The outcome evaluation relies on analysis of secondary data from LAPE and/
or HES and the police obtained initially at LSOA level. Police data are publically
available at street level. HES data are sensitive at LSOA level, although once
alcohol attributable fractions are applied to the data, they are deemed to
represent a low risk of disclosure. Nevertheless, appropriate measures are being
taken to ensure the security of potentially sensitive datasets, including their
storage only on secure university file servers. These data will be aggregated to
compile intervention areas (composed of one to three LSOAs) and their
matched controls.
As part of the process evaluation, potential participants (AHCs, key informants,
stakeholders and people who have had contact with AHCs) invited to take part
in the study are provided with full information about the study. Written
consent is obtained from participants prior to the completion of preand post-training questionnaires and reflective diaries (AHCs only); all
AHCs are given at least 24 h to decide whether or not to take part. Potential
participants for interviews and focus groups are given a minimum of 1 week to
decide whether or not to take part and written informed consent is obtained
prior to the start of the interview or focus group. Data obtained during the
process evaluation are anonymised and each participant is given a unique
code, stored separately to the main data file. Consent forms are stored in a separate location to the main data files. Transcripts use pseudonyms in place of real
names. Data are stored on secure University file servers, accessible only to the
research team.
In line with current policy on open access to data, we will retain all suitably
anonymised research data for 20 years after the end of the study to allow
secondary analyses to take place, and to allow any verification of findings to
take place. Data will be saved as .csv files, which can always be opened by
any program. The model scripts will be provided as.txt files to accompany
the data, so that results can be replicated if required.
Competing interests
KK is responsible for overseeing the delivery of the CICA training programme
on behalf of the Royal Society of Public Health. KA is a council member of
Page 10 of 11
the Royal Society of Public Health. SA is a member of the NIHR public health
research board. All other authors declare that they have no competing interests.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published
maps and institutional affiliations.
Author details
School of Health Sciences, University of Salford, Manchester, UK. 2School of
Health & Society, University of Salford, Manchester, UK. 3Population Health
Sciences, Bristol Medical School, University of Bristol, Bristol, UK. 4School of
Health Sciences, University of York, York, UK. 5Public Health England North
West, Manchester, UK. 6Greater Manchester Combined Authority, Manchester,
UK. 7Royal Society of Public Health, London, UK. 8Fallowfield Community
Guardians c/o School of Health Sciences, University of Salford, Manchester,
UK. 9Wigan Council, Manchester, Wigan, UK.
1
Received: 15 February 2018 Accepted: 5 April 2018
References
1. World Health Organization. Global status report on alcohol and health.
Geneva: World Health Organization; 2014.
2. Barbor T, Caetano R, Casswell S, Edwards G, Giesbrecht N, Graham K, Grube
J, Hill L, Holder H, Homel R, et al. Alcohol: no ordinary commodity: research
and public policy. 2nd ed. Oxford: Oxford University Press; 2010.
3. Cook PA, Phillips-Howard PA, Morleo M, Harkins C, Briant L, Bellis MA. The
big drink debate: perceptions of the impact of price on alcohol
consumption from a large scale cross-sectional convenience survey in north
West England. BMC Public Health. 2011;11:664.
4. Campbell CA, Hahn RA, Elder R, Brewer R, Chattopadhyay S, Fielding J,
Naimi TS, Toomey T, Lawrence B, Middleton JC. Task force on community
preventive services: the effectiveness of limiting alcohol outlet density as a
means of reducing excessive alcohol consumption and alcohol-related
harms. Am J Prev Med. 2009;37:556–69.
5. Fone D, Dunstan F, White J, Webster C, Rodgers S, Lee S, Shiode N, Orford
S, Weightman A, Brennan I, et al. Change in alcohol outlet density and
alcohol-related harm to population health (CHALICE). BMC Public Health.
2012;12:428.
6. Fone DL, Farewell DM, White J, Lyons RA, Dunstan FD. Socioeconomic
patterning of excess alcohol consumption and binge drinking: a crosssectional study of multilevel associations with neighbourhood deprivation.
BMJ Open. 2013;3.
7. Bellis MA, Hughes K, Nicholls J, Sheron N, Gilmore I, Jones L. The alcohol
harm paradox: using a national survey to explore how alcohol may
disproportionately impact health in deprived individuals. BMC Public Health.
2016;16:111.
8. O'Donnell A, Anderson P, Newbury-Birch D, Schulte B, Schmidt C, Reimer J,
Kaner E. The impact of brief alcohol interventions in primary healthcare: a
systematic review of reviews. Alcohol Alcohol. 2014;49:66–78.
9. de Vocht F, Heron J, Angus C, Brennan A, Mooney J, Lock K, Campbell R,
Hickman M. Measurable effects of local alcohol licensing policies
on population health in England. J Epidemiol Community Health. 2016;70:
231–7.
10. de Vocht F, Heron J, Campbell R, Egan M, Mooney JD, Angus C, Brennan A,
Hickman M. Testing the impact of local alcohol licencing policies on
reported crime rates in England. J Epidemiol Community Health. 2017;71:
137–45.
11. D'Onofrio G, Degutis LC. Preventive Care in the Emergency Department:
screening and brief intervention for alcohol problems in the emergency
department: a systematic review. Acad Emerg Med. 2002;9:627–38.
12. Mitchell S: An evaluation of a pilot programme to prepare and support
health trainers as peer educators to deliver IBA to offenders in community
settings. London: CB Offender Health; 2010.
13. Bellis MA, Phillips-Howard PA, Hughes K, Hughes S, Cook PA, Morleo M,
Hannon K, Smallthwaite L, Jones L. Teenage drinking, alcohol availability
and pricing: a cross-sectional study of risk and protective factors for alcoholrelated harms in school children. BMC Public Health. 2009;9:380.
14. Morleo M, Cook PA, Bellis MA, Smallthwaite L. Use of fake identification to
purchase alcohol amongst 15-16 year olds: a cross-sectional survey
Cook et al. BMC Public Health (2018) 18:522
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
examining alcohol access, consumption and harm. Subst Abuse Treat Prev
Policy. 2010;5:12.
Popova S, Giesbrecht N, Bekmuradov D, Patra J. Hours and days of sale and
density of alcohol outlets: impacts on alcohol consumption and damage: a
systematic review. Alcohol Alcohol. 2009;44:500–16.
Martineau F, Tyner E, Lorenc T, Petticrew M, Lock K. Population-level
interventions to reduce alcohol-related harm: an overview of systematic
reviews. Prev Med. 2013;57:278–96.
Bryden A, Roberts B, McKee M, Petticrew M. A systematic review of the
influence on alcohol use of community level availability and marketing of
alcohol. Health Place. 2012;18:349–57.
Holder HD, Gruenewald PJ, Ponicki WR, Treno AJ, Grube JW, Saltz RF, Voas
RB, Reynolds R, Davis J, Sanchez L, et al. Effect of community-based
interventions on high-risk interventions on high-risk drinking and alcoholrelated injuries. JAMA. 2000;284:2341–7.
Glasgow Centre for Population Health. Strengthening the community
voice in alcohol licensing decisions. In: Glasgow Centre for population
health; 2014.
Home Office: Next steps following the consultation on delivering the
Government's alcohol strategy. 2013.
Morgan A, Ziglio E. Revitalising the evidence base for public health: an
assets model. IUHPE - Promot Educ. 2007;17:17–22.
National Institute for Health and Care Excellence. Behaviour change: general
approaches. In: National Institute for health and care excellence; 2007.
Foot J, Hopkins T: A glass half-full: how an asset approach can improve
community health and well-being improvement and development
agency; 2010.
Marmot M: Fair society, healthy lives: the marmot review. 2010.
Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I, Petticrew M:
Developing and evaluating complex interventions: new guidance. London:
Medical Research Council; 2008.
Craig P, Cooper C, Gunnell D, Haw S, Lawson K, Macintyre S, Ogilvie D,
Petticrew M, Reeves B, Sutton M, Thompson S: Using natural experiments to
evaluate population health interventions: guidance for producers and users
of evidence. London: Medical Research Council; 2011.
Lamont T, Barber N, de Pury J, Fulop N, Garfield-Birkbeck S, Lilford R, Mear L,
Raine R, Fitzpatrick R. New approaches to evaluating complex health and
care systems. BMJ. 2016;352:i154.
Moore G, Audrey S, Barker M, Bond L, Bonell C, Cooper C, Hardeman W,
Moore L, O'Cathain A, Tinati T, et al. Process evaluation in complex public
health intervention studies: the need for guidance. J Epidemiol Community
Health. 2014;68:101–2.
National Institute for Health and Care Excellence: Community engagement:
improving health and wellbeing and reducing health inequalities. 2016.
Lorenc T, Oliver K. Adverse effects of public health interventions: a
conceptual framework. J Epidemiol Community Health. 2014;68:288–90.
Bonell C, Jamal F, Melendez-Torres GJ, Cummins S. Dark logic': theorising
the harmful consequences of public health interventions. J Epidemiol
Community Health. 2015;69:95–8.
Public Health England: Local Health Profiles for England (March 2016
update). London: PHE; 2016.
Hussey MA, Hughes JP. Design and analysis of stepped wedge cluster
randomized trials. Contemp Clin Trials. 2007;28:182–91.
de Vocht F, Campbell R, Brennan A, Mooney J, Angus C, Hickman M.
Propensity score matching for selection of local areas as controls for
evaluation of effects of alcohol policies in case series and quasi case-control
designs. Public Health. 2016;132:40–9.
Abadie A, Diamond A, Hainmueller J. Synthetic control methods for
comparative case studies: estimating the effect of California’s tobacco
control program. J Am Stat Assoc. 2010;105:493–505.
Egan M, Brennan A, Buykx P, De Vocht F, Gavens L, Grace D, Halliday E,
Hickman M, Holt V, Mooney JD, Lock K. Local policies to tackle a national
problem: comparative qualitative case studies of an English local authority
alcohol availability intervention. Health Place. 2016;41:11–8.
Brodersen KH, Gallusser F, Koehler J, Remy N, Scott SL. Inferring causal
impact using Bayesian structural time-series models. Ann Appl Stat.
2015;9:247–74.
Scott SL, Varian HR: Predicting the present with Bayesian structural time
series. Int J Math Model Numberical Optimisation 2014, 5.
Abadie A, Gardeazabal J. The economic costs of conflict: a case study of the
Basque country. Am Econ Rev. 2003;93:113–32.
Page 11 of 11
40. Brannstrom L, Trolldal B, Menke M. Spatial spillover effects of a community
action programme targeting on-licensed premises on violent assaults:
evidence from a natural experiment. J Epidemiol Community Health. 2016;
70:226–30.
41. Cost Benefit Analysis [http://www.neweconomymanchester.com/our-work/
research-evaluation-cost-benefit-analysis/cost-benefit-analysis].
42. Drummond M, Sculpher M, Claxton K, Stoddart G, Torrance G. Methods for
the economic evaluation of health care Programmes. 4th ed. Oxford: Oxford
University Press; 2015.
43. Brazier J, Ratcliffe J, Salomon J, Tsuchiya A. Measuring and valuing health
benefits for economic evaluation. 2nd ed. Oxford: Oxford University Press;
2017.
44. Ritchie J, Lewis J, McNaughton Nicholls C, Ormston R. Qualitative research
practice: a guide for social science students and researchers. London: Sage;
2003.
45. Moon JA. A handbook of reflective and experiential learning: theory and
practice. Routledge Falmer: London and New York; 2004.
46. Audrey S, Holliday J, Parry-Langdon N, Campbell R. Meeting the challenges
of implementing process evaluation within randomized controlled trials: the
example of ASSIST (a stop smoking in schools trial). Health Educ Res. 2006;
21:366–77.
47. Friedli L. What we’ve tried, hasn’t worked’: the politics of assets based public
health1. Crit Public Health. 2013;23:131–45.
48. South J, Meah A, Bagnall A, Kinsella K, Branney P, White J, Gamsu M: People
in public health - a study of approaches to develop and support people in
public health roles. 2010.
49. Brunton G, Thomas J, O'Mara-Eves A, Jamal F, Oliver S, Kavanagh J.
Narratives of community engagement: a systematic review-derived
conceptual framework for public health interventions. BMC Public Health.
2017;17:944.
50. Petersen A. Community development in health promotion: empowerment
or regulation. Aust J Public Health. 1994;18:213–7.
51. MacLeod MA, Emejulu A. Neoliberalism with a community face? A critical
analysis of asset-based community development in Scotland. J Community
Pract. 2014;22:430–50.
52. Bonnington O, Harris M. Tensions in relation: how peer support is
experienced and received in a hepatitis C treatment intervention. Int J Drug
Policy. 2017;47:221–9.
53. Masters R, Anwar E, Collins B, Cookson R, Capewell S. Return on investment
of public health interventions: a systematic review. J Epidemiol Community
Health. 2017;71:827–34.
54. Lavoie D. Alcohol identification and brief advice in England: a major plank
in alcohol harm reduction policy. Drug Alcohol Rev. 2010;29:608–11.
55. Johnson M, Jackson R, Guillaume L, Meier P, Goyder E. Barriers and
facilitators to implementing screening and brief intervention for alcohol
misuse: a systematic review of qualitative evidence. J Public Health (Oxf).
2011;33:412–21.