A systematic review of reviews on comprehensive community initiatives to prevent or reduce alcohol and other drug harms

Abstract

Background and Aims

Comprehensive community initiatives (CCI) aimed at reducing or preventing alcohol or other drug (AOD) harms incorporate multiple initiatives delivered to whole communities to effect community‐level change on sociocultural and environmental factors. CCIs have gained in popularity and have been subject to extensive research; however, CCIs comprise multiple initiatives and evidence for effectiveness by substance type has been mixed. This umbrella review aimed to synthesise information from published reviews to describe the combination of CCIs with the most consistent evidence for impact for each substance type.

Method

We searched Embase/Medline, PsycINFO, Cochrane Database of Systematic Reviews, two online registries and hand searched references for English language reviews without date restriction and conducted an umbrella review using mixed methods synthesis (PROSPERO CRD42023432567). We considered all types of reviews focused on CCIs and addressing AOD use or harms. Two reviewers independently screened all articles and conducted full text review. Extraction of main results relating to CCI impact and quality assessment using AMSTAR‐2 and SANRA was completed by two independent reviewers and corrected covered area analysis conducted.

Results

We identified 87 reviews spanning three decades; 14 were rated high quality. Most reviews considered individual substances [alcohol (43 reviews) or tobacco (35 reviews) and rarely illicit drugs (16 reviews)], and some limited scope to ‘at‐risk’ community members [young people (26 reviews) and First Nations (8 reviews)]. Although the evidence did not meet criteria for consistent impact, communities should consider implementing school‐based (supported by 36 of 50 reviews) and parenting‐related (29 of 37 reviews) activities that are supported by media campaign where feasible (29 of 51 reviews). CCIs have the most consistent impact on alcohol (supported by 24 of 32 reviews) and tobacco‐related outcomes (22 of 35 reviews), though illicit drugs are yet to be adequately assessed.

Conclusions

Although the available evidence regarding comprehensive community initiatives is largely inconsistent, the addition of parenting‐related activities to existing school‐based education campaigns is likely to improve effectiveness. A media campaign may extend their reach to those outside school settings. Future evaluation of CCIs should measure impact of activities in isolation where possible and incorporate process measures to gauge community engagement and empowerment.

INTRODUCTION

Alcohol and other drug (AOD) use is a leading cause of the global burden of disease in both developing and developed regions [1]. Although AOD use is an individual's choice, this choice is made in the context of social and environmental factors that contribute to the risk of harmful AOD use [2]. With this understanding, prevention and harm minimisation initiatives should focus on both the individual‐level and the community‐level [3, 4]. Historically, responses to AOD harms were limited to the individual‐level, focussing largely on education programs and improving social skills [5]. However, the focus of responses has now broadened to include community‐level initiatives and these have started to show improved outcomes [6]. Indeed, the World Health Organization now recognises community‐level initiatives as one of ten suggested target areas for global action to reduce the harmful use of alcohol [7]. As there is a complex interplay between the multiple factors contributing to harmful AOD use operating on multiple levels, a comprehensive approach, that addresses as many of these factors as feasible, is thought to be the most viable way to prevent or minimise AOD harms [8, 9].

An AOD comprehensive community initiative (CCI) is defined as an approach aimed at preventing or reducing AOD harms through systemic change by collaborating with multiple sectors across a whole community to build capacity [10]. Community is a contested term, although generally refers to residents of a particular locality (such as a town or city) or to groups of people identifying as having a particular set of shared characteristics or identity (such as First Nations people or students) [11]. To have a community‐level impact, CCI typically involve multiple initiatives.

Although this definition of CCI is helpful to understand what community action is, it does not describe all of the specific initiatives that a CCI may implement. Indeed, despite a significant body of research spanning three decades [12], no ‘best practice’ CCI has emerged where specific initiatives have demonstrated consistent and statistically significant impact across different communities. Further, it remains unclear whether CCI in general has consistent evidence for impact. This may reflect the extant nature of CCI where combinations of initiatives that meet the definition of CCI include anything from a change in teaching practices across a small school district implemented alongside community‐based parenting workshops, through to a government funded community‐wide campaign involving media, alcohol service training, expanded treatment availability and increased compliance enforcement. Adding to the lack of clarity around the initiatives that comprise CCIs is the recognition that establishing the impact of a CCI is particularly challenging and often neglected because of difficulties with evaluation design and identifying and assessing outcome measures [13, 14, 15, 16, 17].

Moreover, because of the complex nature of AOD CCIs, reviews of the literature tend to focus on a subset of possible initiatives, such as information dissemination through education or media campaigns [5, 18, 19] or policy change [20, 21, 22]. Reviews also tend to focus on a particular community type, such as young people [23, 24] or First Nations [25, 26]. As such, those wanting to implement a best practice AOD CCI are likely to have incomplete evidence to guide their decisions.

A logical solution for bringing together such complexity and providing a more comprehensive overview of the significant knowledge from completed systematic reviews of AOD CCIs is to synthesise this information via a systematic review of reviews or umbrella review [27]. Our umbrella review has five aims:

1. Describe the characteristics of reviews (including the review type, number of evaluation studies, community setting, focus relating to CCI, study designs and quality);

2. Identify if there is evidence of consistent impact for CCI regardless of which initiatives it may comprise, by substance addressed;

3. Identify if there is evidence of consistent impact for a particular type of initiative that has been implemented as part of a CCI, by substance addressed;

4. Identify which combinations of initiatives have been reported to have greater impact when implemented together; and

5. Identify considerations related to methods, design and outcome selection linked with positive impact of a CCI approach.

METHODS

This umbrella review is registered with PROSPERO [28], although because of the large scope of initial aims was updated to be submitted in parts, with changes to the original registered protocol provided in Supporting information. This review adheres to the preferred reporting items for overview of reviews (PRIOR) [29], as depicted in Table S1.

Inclusion criteria

We considered all published English language reviews of any type as described by Grant and Booth [30] (critical, meta‐analytic, rapid etc.) without restriction by publication date. We then classified reviews as systematic under Phillips and Barker's definition [31] or narrative as per Sukhera's definition [32]. That is, systematic reviews had: (1) ‘transparent, reproducible methodology which indicate[d] how studies were identified and the criteria on which they were included or excluded’; and (2) an ‘element of evaluation’ provided among a ‘synthesis of…studies' findings’ [31]. For this umbrella review, this ‘synthesis’ was the reporting of CCI including a description the initiatives, outcomes and statistical significance from at least one evaluation study of AOD CCI. In contrast, narrative reviews ‘do not typically involve strict predetermined inclusion or exclusion criteria…[and] the initial scope may change through the review process…Such reviews [do] not provide an exhaustive, comprehensive review of the literature; however, they are useful for a rich and meaningful summary of a topic.’ [32]. For this umbrella review this ‘summary’ was regarding the impact of a CCI without the requirement of a description of evaluation studies.

As CCI is a polysemy concept, our definition per inclusion criteria was any community‐based AOD prevention approach that involved multiple initiatives designed to impact all the individuals of that community (any locality or group of people identifying as having a particular set of shared characteristics or identity).

CCI was included when it demonstrated focus on AOD by (1) clear description of a CCI that targeted any AOD‐related outcome such as use or related harms and impact (CCI on heart health were included where smoking was a primary risk factor); or (2) use of section headings; or (3) including a significant amount of discussion or results regarding the impact of CCI.

Search strategy

Our search strategy was conducted under Preferred Reporting Items for Systematic reviews and Meta‐Analyses (PRISMA) guidelines [33]. We searched three databases (Embase/Medline, PsycINFO and the Cochrane library) and two registries of reviews [the Database of Promoting Health Effectiveness Reviews (DoPHER) and the Database of Abstracts of Review Effects (DARE)], from inception until January 2024. We developed relevant keywords by reviewing the literature, through discussions with faculty librarian and our research group. The search strategy consisted of four primary elements: drug or alcohol use, related harms, community settings and community action (including MeSH terms relating to community) and applied filter for English language and review articles. The full search strategy is provided in Supporting information. In addition, the references of all included reviews were hand searched for additional relevant reviews.

Two authors (P.G. and A.Z.) independently screened titles and abstracts for eligibility and completed full text review of all articles using Covidence software [34]. Data extraction and review quality assessment was completed by two independent reviewers (P.G. and S.F.). Discrepancies were addressed by discussion and disagreements resolved by consensus.

Data extraction

To address aim 1, data extraction included review aims and details of the CCI under review, including initiatives implemented, evaluation study designs, substance(s) addressed and community settings involved. A summary of the included reviews' main discussion points and results regarding the impact of CCI in general (aim 2) and for each CCI initiative (aim 3) were extracted, by the substance(s) addressed. In addition, the included evaluation studies were noted to determine overlap across reviews and were assessed to resolve discrepancies across reviews. To address aim 4, extraction included data from results or discussions indicating which components may work most effectively in combination as part of a CCI. To address aim 5, extraction included factors relating to CCI design, approach or methods of implementation that were reported to be linked with positive significant impact.

Data analysis

Full analysis methods are provided in Supporting information. In brief, the activities described in the reviews as being part of a CCI were grouped into four ‘initiative types’ and 11 ‘initiative subtypes’ (see Figure 1). These four initiative types were: setting‐based, provider‐led, receiver‐based or activity‐based. The 11 initiative subtypes were: mobilisation, coalition‐led, policy‐related, school‐based, parenting‐related, police‐led, vendor‐based, workplace‐based, health service‐based, infrastructure‐related and media campaign. Each initiatives subtype comprised many different activities. For example, the school‐based initiative subtype could include any activity delivered within a school setting such as a curriculum change or extracurricular activities. In our review, we considered CCIs as comprising at least two initiative subtypes, which are made accessible to a whole community.

FIGURE 1.

Comprehensive community initiatives.

CCIs (any combination of initiative subtypes) and each initiative subtype were assessed to determine whether the reviews supported their implementation. We considered evidence for consistent impact, or certainty of the evidence, to be demonstrated where the ratio of the total number of reviews in support of CCI (aim 2) or a particular initiative subtype (aim 3) over the total number of reviews that did not support the CCI or CCI component was equal to or greater than 2:1 (reported by substance addressed). Systematic reviews were considered to support a CCI when the review included at least two CCI evaluation studies and reported that more than half of the evaluated CCIs had a significant positive impact on an AOD‐related outcome. If there was a discrepancy across reviews regarding the impact of CCI or the description of initiative subtypes, the relevant evaluation study was assessed and the reported results corrected. Narrative reviews were assessed for positive language and deductively coded as being supportive of CCIs in general, or of an initiative subtype. All reviews were assessed independently by two reviewers (P.G. and S.F.), discrepancies were addressed by discussion and disagreements resolved by consensus.

For aims 1 through 3 the results from the systematic reviews were considered first with any discrepancies noted by narrative review discussed second. For aims 4 and 5 any narrative discussion from either type of review was considered. Overlap of evaluation studies across reviews was assessed through corrected covered area (CCA) analysis [35], under guidelines proposed by Hennessy and Johnson [36].

To examine whether year of review publication impacted whether reviews demonstrated a CCI approach to have evidence for consistent positive impact, the year of publication was observed by approximately equal groups that reflected the relative distribution of reviews (33 reviews were published before 2005, 27 reviews published between 2005 and 2014 and 26 reviews published between 2015 and 2025).

Quality assessment

Review quality was assessed using highly cited and recommended tools—the 16‐item AMSTAR‐2 [37] for systematic reviews, and the 6‐item SANRA [38] for narrative reviews. We followed AMSTAR‐2 guidance for identifying reviews as high, medium, low or critically low quality [37]. SANRA provides a score from 0 through 12. For ease of reporting, we developed a system to narratively describe a SANRA score whereby a score of at least 9 with no score of 0 on any item was considered to be high quality; a score of between 6 and 9 with no more than one score of 0 on any item was considered to be moderate quality; a score of between 6 and 9 with more than one score of 0 on any item was considered to be low quality; and a score below 6 was considered to be of critically low quality. All reviews were included with review quality noted before result summary and results from high quality reviews presented separately.

RESULTS

Following duplicate removal, 4947 records were identified from our search strategy (see PRISMA flow‐chart) (Figure 2). A total of 4504 records were excluded by title and abstract review, and a further 356 records were excluded following full text review. This resulted in 87 included reviews, 49 of which were narrative reviews [3, 5, 6, 8, 9, 19, 20, 21, 22, 23, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76] and 38 were systematic reviews [4, 10, 11, 18, 24, 25, 26, 74, 75, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108]. Two narrative reviews were considered together [63, 64] as they were published in two parts.

FIGURE 2.

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) diagram of study selection.

Review characteristics (aim 1)

The characteristics of reviews (including the review type, number of evaluation studies, community setting, focus relating to CCI, study designs and quality) are provided in Table 1 and the AOD focus, initiative subtypes that comprised CCIs are in Table 2. Most reviews included all, or did not specify study designs, with some focused specifically on randomised controlled trial designs [77, 81, 82, 83, 87, 96, 98, 101, 105]. The systematic reviews included 264 evaluations. The proportion of CCI evaluation studies included by each review that were reported to have a significant positive impact on substance‐related outcomes is provided in the Supporting information.

TABLE 1.

Main characteristics and quality assessment of included reviews.

Review author, y, reference	Review type, (no. evaluation studies)	Relevant community setting(s)	Review focus relating to CCI	Included study designs	Quality rating
Aguirre‐Molina, 1996 [6]	N	Any	Community‐based approaches	All/unspecified	CL
Allen, 2014 [39]	N	First Nations (American Indian and Alaska Native)	Community level preventive interventions	All/unspecified	M
Allen, 2016 [77]	S (16)	Young People (adolescents, 10–19 years)	Interventions targeting adolescents and parents	RCT	L
Ballester, 2021 [78]	S (18)	Families	Universal family‐based drug prevention programs	All/unspecified	L
Baskerville, 2017 [79]	S (1)	LGBTQ+	Community‐based programs	All/unspecified	H
Bergen, 2014 [80]	S (10)	Any	Publicized sobriety checkpoint programs	All/unspecified	CL
Blewden, 1999 [40]	N	Young people (youth)	Comprehensive active enforcement strategies	All/unspecified	M
Brennan, 2011 [81]	S (5)	Venues	Premise‐based, community‐based, and multi‐level interventions	RCT, quasi‐experimental, minimum 2‐week follow‐up, at least one control group	CL
Bruce, 1999 [41]	N	Venues	Community‐based prevention	Any evaluation of ‘Smokebusters’	H
Burton, 2017 [22]	N	Any	Alcohol control policies, multi‐component community programmes	All/unspecified	H
Cairns, 2014 [18]	S (27)	Young people (11–18 y)	School and family ‘linked’ interventions	CBA with a matched control group	CL
Carson, 2013 [82]	S (23)	Young people (under 25 y)	Multi‐component community interventions	RCT, CCT, CBA, comparisons with no intervention/ single component/school‐based programme control group[s]	H
Casswell, 2000 [42]	N	Any	Community action	All/unspecified	CL
Cibich, 2023 [83]	S (10)	Young people (youth, 12–24 y)	Any rural or remote strategy implemented by a government or non‐government organization	Experimental and quasi‐experimental study designs, RCT, non‐randomised CBA, interrupted timeseries studies, analytical observational studies	CL
Coggans, 2015 [43]	N	Any	Community‐based drug education interventions	All/unspecified	CL
Cuijpers, 2002 [44]	N	Young people (students)	School‐based universal prevention	All/unspecified	M
Das, 2016 [23]	N	Young people (adolescents, 11–24 y)	Family and community‐based interventions, multi‐component interventions	All/unspecified	H
DeJong, 1998 [45]	N	Any	Strategies to reduce DUI (policies, mass communication, community approaches)	All/unspecified	M
Droste, 2014 [84]	S (4)	Any	Community‐level interventions involving hospital data	All/unspecified	CL
Evans, 2021 [46]	N	Any	Prevention programs designed to reduce crime/violence (excluding relationship or sexual violence) that also addressed substance use, implemented in multiple locations	All/unspecified	M
Fincham, 1992 [47]	N	Any	Community health promotion programs	All/unspecified	CL
Flay, 2000 [5]	N	Young people (students)	Substance use prevention utilising school curriculum plus social environmental change	All/unspecified	M
Foxcroft, 2012 [85]	S (20)	Young people (children and adolescents)	Multi‐component universal prevention interventions	Any randomised trial	CL
Gardener, 2018 [86]	S (1)	Any	Health promotion interventions including community campaigns	All/unspecified	CL
Gates, 2006 [87]	S (6)	Young people (under 25 y)	Non‐school based drug prevention, multi‐component community interventions	RCT	H
Geia, 2018 [88]	S (4)	First Nations (Australian Indigenous adolescents and young adults)	Community driven programs	All/unspecified	CL
Giesbrecht, 2014 [48]	N	Any	Local‐level, community‐based alcohol policies	All/unspecified	M
Gorman, 1996 [49]	N	Any	Community‐based initiatives	All/unspecified	M
Graham, 2000 [50]	N	Venues	Preventive interventions for on‐premise drinking	All/unspecified	CL
Haegerich, 2019 [89]	S (4)	Any	State, community and systems‐level opioid overdose prevention interventions	All/unspecified	CL
Hawkins, 2004 [51]	N	First Nations (American Indian and Alaska Native Youth)	Community‐oriented approaches to reduce substance use	All/unspecified	M
Hill, 2013 [90]	S (5)	Any	Tobacco control interventions with data on SES impact, community‐based programmes	All (excluding studies already included within published reviews)	CL
Hingson, 2002 [3]	N	Young people (college students)	Comprehensive community interventions	All/unspecified	M
Hutchison, 2021 [91]	S (9)	Young people (adolescent)	Community coalition efforts	All/unspecified	CL
Jackson, 2012 [24]	S (6)	Young people (11–25 y)	Whole‐school or multi‐setting programmes, community‐based interventions	All/unspecified	M
Jiwa, 2008 [52]	N	First Nations (American Indian)	Community‐based substance abuse services for aboriginal communities in an international context	All/unspecified	M
Jones, 2010 [92]	S (7)	Young people (5–19 y)	Alcohol education in community settings	All/unspecified	CL
Jones, 2011 [93]	S (6)	Venues	Interventions implemented in drinking environments, multi‐component programmes	All (excluding studies limited to sales or supply data)	CL
Jones, 2014 [94]	S (12)	Young people (‘underage’ youth)	Whole‐of‐community interventions	All/unspecified	CL
Kingsland, 2016 [95]	S (2)	Venues	Interventions in sports settings	All trials with a control comparison	H
Krakouer, 2022 [25]	S (1)	First Nations (Australian Indigenous)	Community‐based interventions	All/unspecified	L
Kreuter, 2000 [53]	N	Any	Community partnerships/coalitions for public health	All/unspecified	M
Langford, 2014 [96]	(15)	Young people (4–18 y)	WHO Health Promoting School framework	Cluster‐RCTs, where clusters were at the level of school, district or other geographical area	H
Lantz, 2000 [54]	N	Young people (youth)	Community interventions	All/unspecified	M
Le, 2023 [97]	S (6)	Any	‘Universal’ prevention interventions, community‐based programmes	All economic evaluations	M
Lin, 2020 [10]	S (6)	Young people (children and youth)	CCI	Experimental or quasi‐experimental designs with matched comparison groups	CL
Marshall, 2022 [98]	S (2)	African American	Preventative interventions among African Americans/Blacks populations in community settings	RCT, two‐arm non‐randomised and one‐arm CBA	L
McGrath, 2019 [11]	N	Any	Community engagement in decision‐making occurring within local government in the United Kingdom	All/unspecified	CL
Merzel, 2003 [99]	S (13)	Any	Community‐based approaches to health promotion	All/unspecified	CL
Minichiello, 2015 [100]	S (32)	First Nations (any Indigenous)	Commercial tobacco control interventions demonstrating community level change	All/unspecified	CL
Montoya, 2003 [55]	N	Young people (adolescents, 12–17 y)	Community‐based drug use prevention programs	All/unspecified	CL
Müller‐Riemenschneider, 2008 [101]	S (14)	Young people (children and youth)	Community‐based and multi‐sectorial behavioural prevention interventions	RCTs with a follow‐up duration of at least 12 months	CL
Nagorcka‐Smith, 2022 [102]	N	Any	Community‐based primary prevention initiatives using a coalition model	All (excluding trials, which did not measure the impact of coalition characteristics on outcomes)	CL
Nation, 2003 [56]	N	Any	Comprehensive programming (multiple interventions and multiple settings)	All/unspecified	M
Norman, 1993 [57]	N	Young people (adolescents)	Prevention programs with community involvement	All/unspecified	CL
O'Mara‐Eves, 2015 [58]	N	Disadvantaged groups	Community engagement in public health interventions	All trials with a control group (with no or minimal community engagement) that measured differential impact of social determinants of health	H
Pederson, 2000 [59]	N	African American	Church‐based and community‐based smoking cessation interventions	All/unspecified	H
Pentz, 2000 [20]	N	Any	Community‐based prevention through policy change	All/unspecified	CL
Pentz, 2003 [60]	N	Any	Evidence‐based prevention programs in community settings	All/unspecified	M
Petrie, 2007 [103]	S (19)	Young people (under 18 y)	Parenting prevention programmes in school, health or community settings	All controlled trials with blinded assessment of primary outcomes, ‘adequate’ follow‐up and ‘protection against contamination’	CL
Porthé, 2021 [4]	S (8)	Adults (over 19 y)	Community based interventions (a multi‐sector and multi‐disciplinary collaborative enterprise)	All/unspecified	CL
Power, 2009 [61]	N	First Nations (Australian Indigenous)	Tobacco cessation and prevention programs, community interventions	All/unspecified	H
Quinlan, 2015 [104]	S (3)	Any	Community‐based environmental strategies	All (excluding treatment studies or that focused on individual‐level strategies)	CL
Roussos, 2000 [62]	N	Any	Collaborative partnerships to improve community health	All/unspecified	M
Sanchez‐Puertas, 2021 [76]	S (2)	Young people (children and youth, under 20 y)	Community‐based prevention programs	All/unspecified	M
Secker‐Walker, 2008 [105]	S (35)	Adults (over 17 y)	Community interventions (co‐ordinated, multi‐dimensional programme)	RCT, controlled trials	M
Sellers, 1997 [74]	N	Any	Community heart health programs	Controlled trials with repeated cross‐sectional measurement	H
Shea, 1990a [63]	N	Any	Community‐based cardiovascular disease prevention programs	All/unspecified	CL
Shea, 1990b [64]	N	Any	Community‐based cardiovascular disease prevention programs	All/unspecified	CL
Shults, 2009 [106]	S (6)	Any	Multi‐component programs with community mobilization for reducing DUI	All/unspecified	CL
Sorensen, 1998 [8]	N	Any	Community‐based intervention trials	All/unspecified	M
Spoth, 2009 [9]	N	Young people (‘underage’ youth, under 21 y)	Universal, multi‐component, preventive interventions	All/unspecified	M
Stevenson, 2003 [65]	N	Any	Community‐wide collaboration in substance abuse prevention	All/unspecified	L
Stockings, 2018a [107]	S (24)	Any	Whole‐of‐community interventions involving more than one community‐based setting and multiple components	All trials with a comparison group (excluding trials that allocated individuals to experimental conditions)	M
Stockings, 2018b [66]	N	Any	Whole‐of‐community interventions involving more than one community‐based setting and multiple components	All trials with a comparison group (excluding trials that allocated individuals to experimental conditions)	M
Stoil, 2000 [67]	N	Any	Community‐based demonstrations in substance abuse prevention	All/unspecified	CL
Thomas, 2008 [75]	N	Any	‘Multifaceted’ population tobacco control interventions with data on social equity	All/unspecified	H
Thomson, 2006 [21]	N	Any	Comprehensive tobacco control programmes	All/unspecified	H
Toomey, 2011 [68]	N	Any	Community intervention studies	All/unspecified	CL
Treno, 1997 [69]	N	Any	Community mobilization	All/unspecified	CL
Van Hasselt, 1993 [70]	N	African American	Large‐scale prevention programs, community activity programs	All/unspecified	L
Wakefield, 2000 [71]	N	Young people (13–18 y)	Comprehensive tobacco control programmes	All/unspecified	H
Wakefield, 2003 [19]	N	Young people (youth)	Tobacco control programs with anti‐smoking advertising	All/unspecified	H
Walsh, 2015 [26]	S (2)	First Nations (American Indian and Alaska Native)	Multiple location prevention programs	All/unspecified	CL
Wandersman, 2003 [72]	N	Any	Community‐level interventions (multi‐component interventions that combine individual and environmental change strategies across multiple settings)	All/unspecified	L
Webb, 2010 [108]	S (2)	Hispanic	Community‐based interventions	All/unspecified	CL
Zakocs, 2006 [73]	N	Any	Community coalition efforts	All/unspecified	M

Abbreviations: CBA, controlled before and after trial; CCI, comprehensive community initiative; CCT, controlled clinical trial; CL, critically low quality; DUI, driving under the influence of a substance; H, high quality; L, low quality; LGBTQ+, lesbian, gay, bisexual, trans, queer; M, moderate quality; N, narrative review; RCT, randomised controlled trial; S, systematic review; SES, socio‐economic status; WHO, World Health Authority.

TABLE 2.

Assessment of the support for CCI components and CCI in general by substance type.

Review author, y, reference	Review type, quality	CM	CC	Policy	School	Parenting	Police	Vendor	Media campaign	WP	Health service	Infra.	CCI
Aguirre‐Molina, 1996 [6]	N, CL	–	U	U	C, U	C	u	a, t	A, T, C, U	–	–	–	A, T, C, U
Allen, 2014 [39]	N, M	–	–	–	–	–	–	–	–	–	–	–	u
Allen, 2016 [77]	S, L	A, T, c, u	A, T, c, u	a, t, c	a, T, C, U	a, T, C, U	‐	a, t, c	a, t, c	–	–	–	a, T, C, U
Ballester, 2021 [78]	S, L	a	–	–	A, T, C, i, U	A, T, C, i, U	–	–	–	–	–	–	A, T, C, i, U
Baskerville, 2017 [79]	S, M	–	–	–	–	–	–	t	t	–	t	–	t
Bergen, 2014 [80]	S, CL	–	–	–	–	–	A	‐	A	–	–	–	A
Blewden, 1999 [40]	N, M	–	T	T	–	–	T	T	T	–	–	–	T
Brennan, 2011 [81]	S, CL	a	a	a	–	–	a	a	a	–	a	–	a
Bruce, 1999 [41]	S	t	–	–	t	–	–	–	–	–	–	–	t
Burton, 2017 [22]	S	–	–	–	–	–	A	–	a	–	–	–	A
Cairns, 2014 [18]	S	A	A	a	A	A	a	a	a	a	–	–	A
Carson, 2013 [82]	S	t	t	T	t	t	t	t	t	t	t	–	t
Casswell, 2000 [42]	N	–	–	–	–	–	–	–	–	–	–	–	A
Cibich, 2023 [83]	S	A	A	A	A	A	A	a	A	–	A	–	A
Coggans, 2015 [43]	N	–	–	–	–	–	–	–	U	–	–	–	U
Cuijpers, 2002 [44]	S	–	U	–	U	U	–	–	U	–	–	–	U
Das, 2016 [23]	S	–	–	–	–	–	–	–	–	–	–	–	a, T, u
DeJong, 1998 [45]	N	–	A	A	–	–	A	‐	A	–	–	–	A
Droste, 2014 [84]	S	–	A	–	–	–	A	A	–	–	A	A	A
Evans, 2021 [46]	S	–	–	–	–	–	–	–	–	–	–	–	u
Fincham, 1992 [47]	N	–	T	–	–	–	–	–	T	–	T	–	T
Flay, 2000 [5]	N	–	–	–	Y	Y	–	–	–	–	–	–	u
Foxcroft, 2012 [85]	S	A	A	a	–	A	–	A	A	–	a	–	A
Gardener, 2018 [86]	S	–	t	–	t	–	–	–	–	–	–	–	t
Gates, 2006 [87]	S	c, u	c, u	–	C, i, u	C, i, u	–	c	–	–	–	–	C, i, u
Geia, 2018 [88]	S	a, u	a, u	–	A	–	–	–	–	–	u	–	a, u
Giesbrecht, 2014 [48]	N	a	a	A	–	–	–	A	–	–	–	–	A
Gorman, 1996 [49]	N	–	a	A	a	a	a	a	A	–	a	–	a
Graham, 2000 [50]	N	–	a	–	–	–	A	A	–	–	–	a	a
Haegerich, 2019 [89]	S	i	I	–	I	I	i	–	i	–	I	–	I
Hawkins, 2004 [51]	N	U	U	U	U	–	–	–	–	–	–	–	U
Hill, 2013 [90]	S	t	t	T	t	–	–	–	t	–	t	–	t
Hingson, 2002 [3]	N	a, t, C, u	A, T, C, u	a, T, C	A, T, C, u	A, T, C	a, t	a, t, c	A, T, c	t	a, t	–	a, t, C, u
Hutchison, 2021 [91]	S	a, T, c, i, u	A, T, c, I, U	a	T, c, i, U	A, T, c, i, U	‐	a	a, t, c	–	–	–	A, T, c, i, U
Jackson, 2012 [24]	S	a, t	a, t, c, i, u	–	a, T, c, u	a, t, C, i, u	–	–	–	–	–	–	a, t, c, i, u
Jiwa, 2008 [52]	N	–	–	–	–	–	–	–	–	–	–	–	A, U
Jones, 2010 [92]	S	a	a	–	a	a	–	a	a	a	–	–	a
Jones, 2011 [93]	S	a	a	A	a	–	A	A	a	–	a	–	A
Jones, 2014 [94]	S	a	a	A	–	A	a	a	a	a	a	a	a
Kingsland, 2016 [95]	S	–	a	–	a	–	–	a	a	–	–	–	a
Krakouer, 2022 [25]	S	t	t	–	–	–	t	–	–	t	t	–	t
Kreuter, 2000 [53]	N	–	U	–	–	–	–	–	–	–	–	–	U
Langford, 2014 [96]	S	a, T, c, u	a, T, c, u	a, t, c	T, c, u	a, T, c, u	–	a, t, c	a, T, c	–	t	–	a, T, c, u
Lantz, 2000 [54]	N	–	–	T	T	T	–	–	T	–	T	–	T
Le, 2023 [97]	S	a	A	–	A	A	A	a	a	–	–	–	A
Lin, 2020 [10]	S	a, T, c, i, u	a, T, c, i, u	–	a, T, c, i, u	A, T, c, i, u	–	–	a, t, i	–	–	–	a, T, c, i, u
Marshall, 2022 [98]	S	–	–	–	A	A	–	–	–	–	a	–	A
McGrath, 2019 [11]	S	–	a	a	a	–	–	–	–	–	–	–	a
Merzel, 2003 [99]	S	t, c, u	a, t, c, u	a, t, c	A, t, c, u	a, c, u	–	t, c, u	a, t, c, u	t, u	t, u	–	a, t, c, u
Minichiello, 2015 [100]	S	t	T	T	t	t	t	t	t	t	t	–	t
Montoya, 2003 [55]	N	–	–	U	U	U	–	–	U	–	–	–	U
Müller‐Riemenschneider, 2008 [101]	S	t	t	–	T	T	–	t	t	–	t	–	T
Nagorcka‐Smith, 2022 [102]	S	–	u	–	–	–	–	–	–	–	–	–	u
Nation, 2003 [56]	N	–	–	–	U	U	–	–	–	–	–	–	U
Norman, 1993 [57]	N	u	–	–	T, C	T, C	–	–	–	–	–	–	A, T, C, U
O'Mara‐Eves, 2015 [58]	S	–	–	–	–	–	–	–	–	–	–	–	u
Pederson, 2000 [59]	N	–	t	–	–	–	–	–	T	–	t	–	t
Pentz, 2000 [20]	N	–	U	U	U	–	–	–	U	–	–	–	U
Pentz, 2003 [60]	N	–	u	U	U	U	–	–	U	–	–	–	U
Petrie, 2007 [103]	S	a, T, u	a, t, c, u	A, t, c	T, C, i, u	A, T, C, i, u	a, t	a, t	A, T, C, i	–	a	–	A, T, C, i, u
Porthé, 2021 [4]	S	a	A	a	A	a	a	A	A	–	a	–	A
Power, 2009 [61]	N	–	–	–	–	–	–	t	–	–	T	–	T
Quinlan, 2015 [104]	N	–	c	c	C	–	c	–	C	–	–	–	C
Roussos, 2000 [62]	N	–	u	–	–	–	–	–	–	–	–	–	u
Sanchez‐Puertas, 2021 [76]	N	–	a	–	–	–	–	–	a	–	–	–	a
Secker‐Walker, 2008 [105]	S	t	t	t	t	t	‐	t	t	t	t	–	t
Sellers, 1997 [74]	S	–	T	T	–	–	–	–	–	–	–	–	T
Shea, 1990a,b [63, 64]	N	–	T	T	–	–	–	–	T	–	T	–	T
Shults, 2009 [106]	S	–	T	T, U	T, U	–	–	–	T	T	–	–	T, U
Sorensen, 1998 [8]	N	–	–	–	a	A	–	–	–	–	–	–	a
Spoth, 2009 [9]	N	–	U	U	–	–	–	–	–	–	U	–	U
Stevenson, 2003 [65]	N	a	A	a	a	a	A	a	A	–	A	–	a
Stockings, 2018a [107]	S	a, c, i	a, c, i, u	A, c, i	a, C, i	a, C, i	A, c, i, u	A, c, i, u	A, C, i, u	a	a, c, i	–	a, c, i, u
Stockings, 2018b [66]	N	u	U	–	U	U	–	–	–	–	–	–	U
Stoil, 2000 [67]	N	–	–	T	–	–	–	–	–	–	–	–	T
Thomson, 2006 [21]	N	–	–	T	–	–	–	–	T	–	–	–	T
Thomas, 2008 [75]		–	A	A	–	–	–	–	–	–	–	–	A
Toomey, 2011 [68]	N	–	A	A	–	–	–	–	A	–	–	–	A
Treno, 1997 [69]	N	a	A	a	A	a	a	A	A	–	a	–	A
Van Hasselt, 1993 [70]	N	U	–	–	U	U	–	–	–	–	–	–	U
Wakefield, 2000 [71]	N	–	T	T	T	–	–	t	T	–	t	–	T
Wakefield, 2003 [19]	N	–	–	T	T	–	–	–	T	–	–	–	T
Walsh, 2015 [26]	S	A, u	a, u	a, u	A	–	a, u	a, u	a	–	a, u	–	A, u
Wandersman, 2003 [72]	N	–	a, t	–	–	–	–	–	–	–	–	–	A, T
Webb, 2010 [108]	S	–	t	–	–	–	–	–	t	–	t	–	t
Zakocs, 2006 [73]	N	–	U	–	–	–	–	–	–	–	–	–	U

Note: Note that capital letters indicate that the review was considered to support the initiative subtype or CCI. Lower case letters indicate that the review was not considered to support the initiative subtype or CCI.

Abbreviations: A/a, alcohol; CC, community coalitions; C/c, cannabis; CCI, Comprehensive Community Initiative; CL, critically low quality; CM, community mobilisation; H, high quality; Infra., infrastructure; I/i, illicit drug (other than cannabis); L, low quality; M, moderate quality; N, narrative review; S, systematic review; T/t, tobacco; U/u: unspecified substances; WP, workplace‐based.

Most reviews included location based communities, although 26 reviews limited focus to young people [3, 5, 9, 10, 18, 19, 23, 24, 40, 44, 54, 55, 57, 71, 76, 77, 82, 83, 85, 87, 91, 92, 94, 96, 101, 103], eight on First Nations people (including American Indians, Alaskan Natives and Australian Indigenous people) [25, 26, 39, 51, 52, 61, 88, 100], three on African Americans [59, 70, 98], two on adults [4, 105], one on lesbian, gay, bisexual, transgender and queer youth [79] and one on Hispanics [108].

Four systematic reviews [82, 87, 95, 96] and 10 narrative reviews [19, 21, 22, 23, 41, 58, 59, 71, 74, 75] were of high quality. Full results of the quality analysis are available in Table S2 and S3.

Assessment of evidence for the consistent impact of a CCI approach, by substance (aim 2)

The 264 CCI evaluation studies included in the systematic reviews had little overlap (CCA of 1.33). Without regard to the type of AOD‐related outcome, over half of these reviews (21/38) were considered to support CCIs, however, this finding did not meet our criteria for consistency of evidence. A summary of whether each review was considered to support CCI or any assessed initiative subtype, for each substance is provided in Table 2.

Alcohol

CCI impact on alcohol‐related outcomes was assessed by 26 systematic reviews [4, 10, 11, 18, 24, 26, 81, 83, 84, 85, 88, 91, 92, 93, 94, 95, 96, 97, 98, 99, 103, 106, 107], and 17 narrative reviews [3, 6, 9, 22, 23, 42, 45, 48, 49, 50, 52, 57, 66, 68, 69, 72, 76]. The systematic reviews exploring the impact of a CCI approach on alcohol‐related outcomes did not demonstrate evidence for consistent positive impact as just over half these reviews (14/26) were considered to support CCI to target alcohol‐related outcomes. These reviews assessed 137 evaluation studies with little overlap (CCA of 2.36). This result varied across review quality as most of the critically low or low quality systematic reviews (13/21) were considered to support CCIs targeted at alcohol‐related outcomes, however, only one of the five systematic reviews of moderate or high quality was considered to support CCIs. Over half of the narrative reviews (10/17) were considered to support CCIs focused on alcohol‐related outcomes. This result also varied across review quality, as almost all critically low or low quality narrative reviews were considered to support CCI to target alcohol‐related outcomes (6/7), while just under half of the 10 moderate or high quality narrative reviews were considered to support CCI (4/10). The year of review publication did not appear to impact the proportion of reviews that were considered to support CCI to target alcohol‐related outcomes.

Tobacco

CCIs impact on tobacco‐related outcomes was assessed by 17 systematic reviews [10, 24, 25, 77, 78, 79, 82, 86, 90, 91, 96, 99, 100, 101, 103, 105, 108], and 18 narrative reviews [3, 6, 8, 19, 21, 23, 40, 41, 47, 54, 57, 59, 61, 64, 71, 72, 74, 75]. The systematic reviews on CCI approaches on tobacco‐related outcomes did not demonstrate evidence for consistent positive impact, with less than half (7/17) considered to support target tobacco‐related CCIs. These reviews assessed 138 evaluation studies with little overlap (CCA of 1.62). This result did not vary across review quality although reviews of moderate or high quality were less likely to be considered to support CCI to target tobacco‐related outcomes, as only one of the five systematic reviews met support criteria. In contrast, the narrative reviews demonstrated evidence for consistent positive impact, with 15 of 18 reviews considered to support CCIs targeted on tobacco‐related outcomes. This result did not vary by review quality. Notably, most of these narrative reviews were published before 2005, with the most recent published in 2016 [23]. In contrast, all but one systematic review was published after 2005, with the most recent published in 2022 [25]. As a result, the proportion of reviews considered to support CCIs targeted at tobacco‐related outcomes reduced across publication year groups.

Cannabis

CCI impact on cannabis‐related outcomes was assessed by 11 systematic reviews [10, 24, 77, 78, 87, 91, 96, 99, 103, 104, 107] and three narrative reviews [3, 6, 57]. The systematic reviews did not demonstrate evidence for consistent positive impact as less than half these reviews (5/11) were considered to support CCIs for cannabis‐related outcomes. These reviews assessed 32 evaluation studies with low to moderate overlap (CCA of 5). This overlap was because of two reviews identifying nine separate evaluation studies (18 total), with several of these studies also being assessed in other reviews. The proportion of reviews considered to support CCI to target cannabis‐related outcomes did not vary across review quality. In contrast, all three narrative reviews were considered to support CCI to target cannabis‐related outcomes. However, these reviews were all published before 2005 and focussed discussion on particular impactful CCI—Project Northland [109]. This was an intensive CCI involving several initiative subtypes and was also the most frequently assessed CCI by the systematic reviews.

Illicit drugs

CCI impact on illicit drug‐related outcomes (other than cannabis) was assessed by eight systematic reviews [10, 24, 78, 87, 89, 91, 103, 107] and no narrative reviews. These eight reviews assessed 14 evaluation studies with high overlap (CCA of 8.33). This overlap was because of one moderate quality 2018 review identifying seven evaluation studies [107] with four of these studies identified in other reviews. This review was not considered to support CCIs on target illicit‐drug related outcomes as just two of the seven CCIs were reported to have a significant positive impact. These reviews did not demonstrate evidence for consistent positive impact as just one was considered to support CCI to target illicit drug‐related outcomes—a critically low quality 2019 review of CCI intended to prevent opioid overdose [89]. All these CCIs were reported to have a significant positive impact, including ‘Project Lazarus’, which involved an intensive coordination of treatment services, schools and police [110].

Unspecified substances

CCI impact on unspecified substance‐related outcomes was assessed by 12 systematic reviews [10, 24, 26, 77, 78, 87, 88, 91, 96, 99, 103, 107] and 24 narrative reviews [3, 5, 6, 8, 20, 23, 39, 43, 46, 51, 52, 53, 55, 56, 57, 58, 60, 62, 65, 67, 70, 73, 102]. These reviews would typically refer to outcomes using generic terms such as ‘drug’. The 12 systematic reviews exploring the impact of a CCI approach on unspecified substance‐related outcomes did not demonstrate evidence for consistent positive impact as just three were considered to support CCI to target unspecified substance‐related outcomes. These reviews assessed 32 evaluation studies, with low overlap (CCA of 2.08). This result did not vary across review quality. In contrast, the 24 narrative reviews exploring the impact of a CCI approach on unspecified substance‐related outcomes did demonstrate evidence for consistent positive impact as 16 were considered to support CCI to target unspecified substance‐related outcomes. Notably, 14 of these 16 reviews were published before 2005. In contrast, the three systematic reviews considered to support CCI to target unspecified substance‐related outcomes were all published after 2015.

Assessment of evidence of consistent impact of initiative subtypes, by substance (aim 3)

Table 2 describes whether each review was considered to support each of the 11 initiative subtypes (when delivered in the context of CCI), by substance, with full results detailed in Supporting information.

Across the systematic reviews no initiative subtype was considered to demonstrate consistent evidence of impact. In contrast, among the narrative reviews, some initiative subtypes were considered to demonstrate consistent evidence of impact, as described by substance below.

Regarding alcohol, the narrative reviews described policy‐related, police‐led and media‐campaign initiative subtypes to demonstrate consistent evidence of impact. Most systematic reviews were considered to support school‐based, parenting‐related, police‐led and media campaigns as CCI subtypes to target alcohol‐related outcomes.

For tobacco, the narrative reviews described policy‐related, school‐based and parenting‐related initiative subtypes to demonstrate consistent evidence of impact, with most systematic reviews considered to support school‐based and parenting‐related initiative subtypes.

Regarding cannabis, the narrative and systematic reviews both described school‐based and parenting‐related initiative subtypes to demonstrate consistent evidence of impact. No initiative subtype was considered to demonstrate consistent evidence of impact for illicit drugs (other than cannabis).

Regarding unspecified substances, the narrative reviews described coalition‐led, policy‐related, school‐based, parenting‐related and media campaign initiative subtypes to demonstrate consistent evidence of impact.

Combinations of initiative subtypes reported to have greater impact when implemented together (aim 4)

Most review authors reported on specific combinations of initiative subtypes that were more effective when paired together than when implemented alone. In contrast, just one review supported single verse multiple interventions, although this result was not specific to substance use and also included other health outcomes such as breastfeeding, healthy eating and physical activity [58]. In addition, 13 reviews highlighted that there is yet to be consensus around which combinations of initiative subtypes are likely the most effective [9, 26, 39, 43, 57, 60, 66, 76, 83, 87, 88, 90, 108].

Effective combinations of initiative subtypes reported by authors included:

1. School‐based and parenting‐based (n = 13) [9, 18, 24, 77, 78, 82, 85, 89, 91, 96, 98, 101, 103];

2. Venue‐based and police‐led (education and training for responsible service that is monitored with ongoing police enforcement) (n = 7) [6, 50, 81, 84, 93, 94, 106], although one review noted that the cost implications of this combination were rarely considered [22];

3. Community coalition and media campaign (working to raise community engagement with the CCI objectives) (n = 6) [4, 45, 69, 82, 94, 105];

4. Community coalition and policy‐related (community advocacy for policy change) (n = 4) [18, 68, 69, 94];

5. Policy‐related and media campaign (using media to raise community support for policy change) (n = 4) [69, 94, 104, 105];

6. Police‐led and media campaign (publicised roadside checkpoints and patrols) (n = 4) [4, 22, 45, 80];

7. School‐based and media campaign (a ‘supplementary approach’ to target youth) (n = 2) [6, 104];

8. Health service‐based and parenting‐related (social services working in association with health services) (n = 1) [78];

9. Health service‐based and community coalition (using emergency department data to inform allocation of coalition resources) (n = 1) [84]; and

10. Health service‐based and media campaign (using media to raise awareness of health services) (n = 1) [79].

Considerations relating to methods, design and outcome selection linked with positive impact (aim 5)

Theoretical approaches

Reviews reported that CCIs should be based on theoretical underpinnings that illustrate how the approach will reach its desired effect [3, 8, 56, 58, 72, 92, 99, 106]. Across reviews, social learning and social influence theories were most common. The theory tenets related to building positive relationships [4, 5, 6, 43, 44, 47, 54, 56, 57, 58, 60], and understanding social influences to build refusal skills [57, 60]. Additionally important was diffusion of innovation theory [20, 47, 60, 63, 64], which encourages a focus on community norms [20, 22, 43, 57, 60] and improving community readiness and support for the initiative objectives. Finally, community empowerment theory was often used, which focuses on increasing community ownership of the approach and the promoting movements toward social change [6, 8, 49, 51]. With this understanding of theory, the chain of logic, which connects the approach to desired effects should be illustrated in a program logic and clearly communicated [8, 9, 53, 62, 65, 106].

Community engagement and participation

CCIs involving the community in the design of initiatives were reportedly more effective [3, 4, 8, 42, 46, 48, 49, 51, 58, 74, 93, 100, 102, 107], especially when involving community leaders or ‘champions’ [4, 20, 65, 82, 100]. As statistically significant positive impact was often associated with the community's awareness of the CCI and the ‘dose’ received, authors recommended that CCI planners should include methods to increase community participation [56, 82, 105]. In addition, authors reported that CCIs should be aware of the target community and initiatives should be culturally appropriate [25, 51, 56, 58, 60, 92, 100]. This is especially the case when implementing a health‐service based component with First Nations communities [25, 51, 100].

Involvement of professionals

Reviews reported that CCIs should be delivered by adequately trained professionals, especially those involving education and service delivery [20, 56, 65, 92].

Outcome selection

CCIs with more than one outcome of interest (such as crime rates, drink driving and consumption) were found to be less effective than those with a single substance‐related outcome (such as risky drinking) as resources and, therefore, impact may be diluted [3, 94]. One review also suggested that the single outcome should be any substance use as opposed to a particular substance [60].

DISCUSSION

Over the last three decades there have been close to 100 reviews of AOD CCIs to prevent or reduce AOD use and harms. Over time, these reviews have consistently stated that further research is needed to identify a gold standard approach and to date the most effective combination of components is not established [9, 26, 39, 43, 57, 60, 66, 76, 83, 87, 88, 90, 108]. This is unsurprising as CCIs can comprise any combination of initiatives and effectiveness can be measured in as many ways. We aimed to synthesise the available information from published reviews to describe the combination of initiatives with the most consistent evidence for impact for each substance type.

Evidence for consistent positive impact of CCIs was based on a ratio of at least 2:1 of reviews that were considered to support CCI. Under this definition, the systematic reviews did not identify evidence for consistent positive impact of CCIs on any substance‐related outcome, nor any particular set of activities within a CCI (referred to as an initiative subtype). This finding was noteworthy given that there was very low overlap across reviews in the 264 evaluation studies that were assessed and was consistent across review quality. However, we did note that just over half of systematic reviews were considered to support CCI to target alcohol or tobacco, although this was not the case regarding cannabis, other illicit drugs or unspecified substances. This may reflect a relative low impact of CCI on illicit drug‐related outcomes or a lack of primary focus on illicit drugs by any of the reviews with the exception of one review on community initiatives to prevent opioid overdose [89]. Impact on illicit drug use was typically measured by CCI aiming to prevent any substance use and very few CCI were described to have illicit drug use as a primary focus. In addition, as the vast majority of CCI target young people, measuring impact on illicit drug use is constrained by relatively low samples of users and would require lengthy follow‐up. Taken together, the impact of CCI on illicit drugs is unclear.

In contrast, when based on the proportion of narrative reviews considered to support CCI, there was evidence for consistent positive impact of CCI when targeting tobacco, cannabis and unspecified substances. This difference may be explained by the relatively small samples of evaluation studies that would have been available to many narrative reviews, as 32 of the 48 were published before 2005. In addition, many of these reviews were describing a few prominent and impactful early examples of CCI, including Project Northland [109], and the Community Trials Project [111], where additional studies would have been included with a systematic design.

Several specific factors were reported to increase the likelihood that a CCI would have impact. First, including activities to address multiple risk or protective factors [3, 4, 8, 20, 43, 49, 56, 60, 66, 92, 93] and to increase rates of participation [56, 82, 105]. Second, communities should lead or be involved in CCI design, through community consultation [3, 8, 42, 46, 48, 49, 51, 58, 74, 93, 100, 102, 107], to identify initiatives that could address local issues and also to ensure appropriateness in terms of the community: culture [25, 51, 56, 58, 60, 92, 100], equity (gender, social class and ethnicity) [58, 90, 96, 102] and readiness [20, 47, 60, 63, 64]. Third, there should be a clear connection between the CCI objectives and the initiatives, with this connection underpinned by an appropriate theory [3, 8, 56, 58, 72, 92, 99, 106] and demonstrated by a program logic [8, 9, 53, 62, 65, 106].

Based on the narrative reviews, several initiative subtypes did demonstrate evidence for consistent impact, some of which were supported by over half of the systematic reviews. Taken together, the available evidence indicates that CCIs targeting alcohol should first consider police‐led activities and a media campaign and potentially include school‐based and parenting‐related activities. CCIs targeting tobacco should first consider school‐based and parenting‐related activities and potentially include policy‐related activities relating to creating smoke free environments. CCIs targeting cannabis should consider school‐based and parenting‐related activities. Further research is needed to clarify which initiative subtypes may be most effective to target illicit drugs. In addition, including a media campaign working to raise engagement and awareness and correct normative beliefs was consistently thought to improve impact when combined with most other initiative subtypes. In contrast, some reviews cautioned against community mobilisation (specifically holding one‐off community events for information dissemination or alternative activities as a substitute for AOD use) because of the resources required and lack of impact [39, 66, 82, 95, 107].

The evidence for consistent impact was very clearly weighted to school‐based and parenting‐related activities. A caveat here is that the school and parenting initiatives should involve interactive components and incorporate social learning, building positive relationships and refusal skills [4, 5, 6, 43, 44, 47, 54, 56, 57, 58, 60, 77, 82, 101, 103]. In addition, although these activities benefit from the fact that most children and their families have contact with schools, they have a limited reach or applicability to those not in school such as disengaged youth or adults. Regardless, this finding is noteworthy as school programs that address substance use and involve parents may be one of the less resource intensive activities assessed, and schools and families are perhaps the most ubiquitous component of any community. These activities were the most consistently assessed across reviews, followed closely by coalition‐led activities.

However, the reported impact from coalition‐led activities was mixed across reviews, although the narrative reviews did demonstrate consistent evidence for impact regarding unspecified substances. Notably, coalition‐led activities are often presented as the structural backbone to the funding, development and delivery of CCIs [62, 65, 73, 91, 102], although their evaluation is recognised to be resource intensive and complex [53, 62]. Some authors argued that the value of coalitions relates to community engagement and empowerment [102], as well as interagency collaboration across multiple community sectors [62, 65, 73]. It follows that measuring the impact of coalitions should include process measures, such as effectiveness of community consultation or connections across services. As a result, the importance of coalition‐led activities was likely downplayed by our limited focus on substance‐related outcomes.

As with any review, this umbrella review suffers from the same limitations of included studies and could only reflect what has been published [112]. This was particularly evident regarding the impact of CCI on illicit drug‐related outcomes and adults, as well as the impact of workplace‐based, health service‐based and infrastructure related activities within CCI. The majority of reviews were of critically low quality, with only four systematic reviews [82, 87, 95, 96] and 10 narrative reviews [19, 21, 22, 23, 41, 58, 59, 71, 74, 75] rated as high quality. Moreover, just two of these systematic reviews considered CCI of any type [82, 87] as opposed to having a limited focus on particular initiative subtypes such as school‐based activities [96] or vendor‐based activities [95]. Major contributors to the low quality of reviews were a lack of detail of the included studies and failing to consider risk of bias. In addition, almost all the evaluation studies included in the reviews were published before 2015 and may be outdated. The impact of review quality and publication date was assessed with differences reported in results.

It is imperative to make recommendations on CCI with evidence for effectiveness as decision makers have historically funded single component initiatives such as education or information campaigns that have been demonstrated to be ineffective [94]. Although the available evidence regarding CCI was largely inconsistent, our findings demonstrate that the addition of parenting‐related activities to be part of such existing school‐based education campaigns is likely to improve effectiveness. Where resources allow, a media campaign is likely to support these activities by extending their reach to those outside school settings. Future work should build on what is demonstrated here to be best associated with effectiveness by resourcing evaluation of CCI, especially those targeting adults and illicit drug use, and noting the comparative impact of any chosen activities.

AUTHOR CONTRIBUTIONS

Peter Gates: Conceptualization (equal); data curation (equal); formal analysis (equal); writing—original draft (lead); writing—review and editing (lead). Andrea Zocco: Data curation (supporting); investigation (supporting); writing—review and editing (supporting). Sara Farnbach: Conceptualization (equal); data curation (equal); formal analysis (supporting); investigation (supporting); methodology (supporting); resources (lead); supervision (equal); validation (supporting); writing—original draft (equal); writing—review and editing (equal).

DECLARATION OF INTERESTS

The authors declare they have no potential conflicts of interest, financial interests, or competing interests directly or indirectly arising from this research. Ethical approval was not required for this paper. P.G., A.Z. and S.F. have no interests to declare.

This research was produced in whole or part by University of New South Wales (UNSW) Sydney researchers and is subject to the UNSW Intellectual property policy. For the purposes of Open Access, the author has applied a Creative Commons Attribution CC‐BY licence to any Author Accepted Manuscript (AAM) version arising from this submission.

Supporting information

Table S1. Preferred reporting items for overview of reviews (PRIOR) Checklist

Table S2. Systematic Review Quality Assessment – AMSTAR 2

Table S3. Narrative Review Quality Assessment – SANRA

Table S4. Systematic review results by initiative subtype and CCI overall: impact on alcohol‐related outcomes

Table S5. Narrative review results by initiative subtype and CCI overall: impact on alcohol‐related outcomes

Table S6. Systematic review results by initiative subtype and CCI overall: impact on tobacco‐related outcomes

Table S7. Narrative review results by initiative subtype and CCI overall: impact on tobacco‐related outcomes

Table S8. Systematic review results by initiative subtype and CCI overall: impact on cannabis‐related outcomes

Table S9. Narrative review results by initiative subtype and CCI overall: impact on cannabis‐related outcomes

Table S10. Systematic review results by initiative subtype and CCI overall: impact on illicit drug‐related outcomes

Table S11. Systematic review results by initiative subtype and CCI overall: impact on unspecified substance‐related outcomes

Table S12. Narrative review results by initiative subtype and CCI overall: impact on unspecified substance‐related outcomes

Gates PJ, Zocco AC, Farnbach S. A systematic review of reviews on comprehensive community initiatives to prevent or reduce alcohol and other drug harms. Addiction. 2026;121(2):241–260. 10.1111/add.70153

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.