International Journal of Drug Policy 112 (2023) 103958

Contents lists available at ScienceDirect

International Journal of Drug Policy

journal homepage: www.elsevier.com/locate/drugpo

Review

What is a public health approach to substance use? A qualitative

systematic review and thematic synthesis
Jean-François Crépault a,b,∗, Cayley Russell a,c, Tara Marie Watson a, Carol Strike b, Sarah Bonato a,
Jürgen Rehm a,b,c,d,e,f
a
Centre for Addiction and Mental Health (CAMH), 1001 Queen Street, Toronto, Ontario, M6J 1H4, Canada
b
Dalla Lana School of Public Health, University of Toronto, 6th Floor, 155 College Street, Toronto, Ontario, M5T 3M7, Canada
c
Institute of Medical Science, University of Toronto, 1 King’s College Circle, Toronto, Ontario, M5S 1A8, Canada
d
Department of Psychiatry, University of Toronto, 8th Floor, 250 College Street, Toronto, Ontario, M5T 1R8, Canada
e
Institute of Clinical Psychology and Psychotherapy & Center for Clinical Epidemiology and Longitudinal Studies, Technische Universität Dresden, Chemnitzer Str. 46,
D-01187 Dresden, Germany
f
Center for Interdisciplinary Addiction Research (ZIS), Department of Psychiatry and Psychotherapy, University Medical Center Hamburg-Eppendorf (UKE),
Martinistraße 52, 20246 Hamburg, Germany

a r t i c l e i n f o a b s t r a c t

Keywords: Background: Psychoactive substance use and the regulations that govern it both have the potential to lead to
Public health approach harm. A ‘public health approach’ (PHA) is frequently invoked as a means of addressing these harms, but the term
Public policy is used in inconsistent and contradictory ways. This study systematically reviewed the English-language academic
Drug policy
literature to understand how a public health approach to substance use is defined and described.
Substance use
Systematic review Methods: This review employed thematic synthesis, a methodology designed to rigorously synthesize qualitative
Thematic synthesis evidence. Eligible articles were published in peer-reviewed journals, in the English language, with full text avail-
able, and focused primarily on substance use. There were no limits on year of publication. Original research,
opinion/commentary, and reviews were included. The searches were conducted in October 2021 in CINAHL,
Embase, Medline, PAIS Index, PsycINFO, Scopus, Sociological Abstracts, and Web of Science.
Results: 272 articles from 25 countries, published between 1950 and 2021, were synthesized. Definitions of a
PHA have changed over time and differ by substance. The most commonly cited characteristics of a PHA were:
for alcohol, regulation, e.g. of price and availability (54% of articles); for cannabis: regulation (68%); for illicit
drugs: that a PHA is distinct from a criminal justice approach (63%); for opioids: substance use disorder treatment
(55%); and for tobacco: regulation (62%).
Conclusion: There is no consensus on the definition of a public health approach to substance use, but there is
substantial agreement when it comes to PHAs to specific substances. There are also similarities in how they
are described for legal substances versus illicit ones. This review found areas of disagreement regarding the
extent to which PHAs should focus on individual-level factors. Policymakers, academics, and others developing
or implementing PHAs to substance use should be explicit about their aims and objectives – as well as the premises
and assumptions underlying them.

Introduction consumption and harm (Babor et al., 2018). In recent years, govern-
ments have increasingly invoked the concept of a ‘public health ap-
Psychoactive substance use and the regulations that govern it both proach’ as a means of addressing these harms. When the Canadian
have the potential to lead to harm. As an example of the former, the government legalized non-medical cannabis in 2018, it referred to its
World Health Organization (2022a, 2022b, 2022c) estimates that alco- strategy as a public health approach (Health Canada, 2018). Both the
hol, tobacco, and illicit drugs account for more than 11 million deaths a American and Canadian governments have claimed to be guided by a
year. As examples of the latter, the criminalization of some substances public health approach to opioids (Centers for Disease Control and Pre-
negatively impacts HIV prevention and treatment (DeBeck et al., 2017), vention, 2019; Health Canada, 2021). And at the UN General Assem-
whereas the commercialization of others is associated with increased bly Special Session on the World Drug Problem (UNGASS) in 2016,

∗
Corresponding author.
E-mail address: JeanFrancois.Crepault@camh.ca (J.-F. Crépault).

https://doi.org/10.1016/j.drugpo.2023.103958

0955-3959/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

numerous countries stated their support for a public health approach Aim
to substance use (Volkow et al., 2017). However, these “public health
approaches” seem to have little in common. Government strategies Using a modified version of the population, intervention, com-
to reduce opioid-related harms in the United States and Canada dif- parator, and outcome (PICO) tool for formulating qualitative system-
fer markedly, particularly in the area of harm reduction: while super- atic review questions (Munn et al., 2018) (see the study protocol
vised injection sites are a cornerstone of Canada’s response, they are Crépault et al., 2021 for a brief discussion), the problem under consider-
illegal in the U.S. (Leung, 2019). In addition, both governments’ ad- ation is substance use / drug policy;1 the phenomenon of interest is the
herence to a public health approach are disputed by some policymak- definition of a PHA; and the context is the English-language academic
ers and non-governmental organizations (Human Rights Watch, 2021; literature. Thus, our research question is as follows: How is a PHA to
Wyton, 2022). Finally, following UNGASS, the term has been used by substance use defined and described in the English-language academic
some governments as a “smokescreen” to justify drug policies focused literature?
on incarceration and other punitive measures rather than public health
(Csete & Wolfe, 2017). While it is not surprising that there would be Synthesis methodology
some variation in public health approaches across substances and coun-
tries, these strategies are so different as to call into question the meaning This review employs thematic synthesis, which was designed by
of the concept. They can also be expected to have very different impacts Thomas and Harden (2008) as a means of rigorously and transparently
on public health. synthesizing research in the area of public health (also see Barnett-Page
Calls for a public health approach are common in the academic lit- & Thomas, 2009, and Noyes et al., 2020). It has three stages of analysis:
erature as well. They have been proposed as a way to reduce alcohol- 1) coding text, 2) grouping codes into descriptive themes, and 3) inter-
related harms (Querido, 1954), as a model for legalizing and regulat- preting those to generate analytic themes (Thomas & Harden, 2008).
ing cannabis (Haden & Emerson, 2014), and as a means of addressing In that third stage, the objective is to “‘go beyond’ the primary studies
North America’s opioid poisoning crisis (Kolodny et al., 2015), to name and generate new interpretive constructs, explanations or hypotheses”
just three, but here too the term is used in inconsistent and contradic- (Thomas & Harden, 2008). Per the Cochrane Handbook for Systematic
tory ways. Proponents of public health approaches differ in their gen- Reviews of Interventions (Noyes et al., 2020, p. 534), “Findings may be
eral views of substance use, ranging from broad disapproval (Gostin & presented as a narrative, or summarized and displayed as tables, in-
Lazzarini, 1997) to the suggestion of potential benefits (Jonas, 1994); fographics or logic models.” This flexibility in how to present review
accordingly, some calls for a public health approach focus on prevention findings makes thematic synthesis well suited to our research question.
and treatment (McLellan, 2017), while others emphasize harm reduction
(Sweanor et al., 2007). Even among calls for a public health approach Inclusion criteria
to opioids in North America, definitions range from those narrowly
focused on population-level prevention via surveillance and research Eligible articles met the following criteria: they were published in
(Joranson & Gilson, 2006) to expansive ones that include a commitment peer-reviewed journals, in the English language, with full text available,
to the principles of social justice, human rights, and equity (Emerson and focused primarily on substance use. There were no limits on year of
& Haden, 2021). So while public health approaches to substance publication. Original research, opinion/commentary, and reviews were
use are frequently called for, what they are—or should be—remains all included.
unclear. As we were specifically interested in the academic literature, grey
This study systematically reviews the English-language academic lit- literature was excluded from this review (for a discussion of this deci-
erature to understand what is meant and understood when the term sion, see the study protocol Crépault et al., 2021.) Book reviews and
‘public health approach’ (PHA) is used in the context of substance use. conference proceedings were excluded, as were responses to or com-
It employs thematic synthesis to explore how definitions of a PHA have mentaries on other articles—unless the definition of a public health ap-
changed over time and differ by substance. It concludes that although proach was central to the response or commentary. Also excluded were
there is no across-the-board consensus on the definition, characteristics, articles where substance use was not the primary focus (e.g., an article
and components of a PHA to substance use, there is a substantial level about public health approaches to HIV that mentions intravenous drug
of agreement when it comes to PHAs to specific substances or classes of use, or one proposing a PHA to gambling by drawing on the substance
substances. use literature).
We believe this paper is the first to systematically review the aca-
Data sources
demic literature for definitions of a public health approach. In addition
to shedding light on how the term has been used historically, we hope it
The searches were conducted on October 1, 2021, in eight databases:
will contribute to a shared understanding of the premises and objectives
CINAHL, Embase, Medline, PAIS Index, PsycINFO, Scopus, Sociological
of a PHA to substance use in general, as well as the components of PHAs
Abstracts, and Web of Science. These databases were selected to capture
to particular substances. We believe this will be of value to policymakers
literature from a wide range of medical and public health journals as
as well as public health and drug policy workers.
well as journals in a range of social sciences.

Search strategy
Methods
The team’s librarian (SB) designed a comprehensive search strategy
This report follows two guidelines: Preferred Reporting Items for Sys- to capture all available journal articles addressing the topic of a ‘public
tematic Review and Meta-Analysis (PRISMA) (Page et al., 2021) and
Enhancing Transparency in Reporting the Synthesis of Qualitative Re- 1
We use the terms “substance” and “drug” interchangeably in this document
search (ENTREQ) (Tong et al., 2012). This methods section is structured
to refer to licit and illicit psychoactive substances such as alcohol, cannabis, opi-
according to ENTREQ guidelines; for a PRISMA checklist, see Supple-
oids, psychedelics, stimulants (excluding caffeine), and tobacco. By drug policy,
mentary Material. we mean public policy that is concerned with psychoactive substances: “laws,
The protocol for this systematic review was prospectively registered regulatory measures, funding priorities, and other courses of action concerning
at PROSPERO (CRD42021270632) (National Institute for Health Re- psychoactive drugs that are promoted and implemented by a governmental en-
search, 2020) and a more detailed version was published in a journal tity and its representatives” (European Monitoring Centre for Drugs and Drug
(Crépault et al., 2021). Addiction, 2013).

2
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

Table 1 Table 2
PRISMA flow diagram. Year of publication of synthe-
sized articles.
4483 references imported for screening
↳ 2228 duplicates removed n %
2255 studies screened against title and abstract
↳ 1650 studies excluded 1950–1959 3 1.1%
605 studies assessed for full-text eligibility 1960–1969 3 1.1%
↳ 333 studies excluded 1970–1979 6 2.2%
↳ 186 Irrelevant 1980–1989 11 4.0%
↳ 49 Duplicate 1990–1999 37 13.6%
↳ 40 Wrong article type 2000–2009 46 16.9%
↳ 37 Full text unavailable 2010–2019 123 45.2%
↳ 21 Wrong language 2020– 43 15.8%
272 studies included

Table 3
Subject matter of synthesized articles.
health approach.’ This term was searched using adjacency / proxim-
n %∗
ity operators to capture slight phrasing variations (e.g., “public health-
based approach” or “public health-oriented approach”). This strategy Alcohol 66 24.3%
General / unspecified 56 20.6%
leaves out approaches that might have the same objectives and char-
Illicit drugs∗∗ 49 18.0%
acteristics but different terminology, e.g. “public health framework” or Opioids 39 14.3%
“public health strategy.” Searches were adapted for each database indi- Tobacco 31 11.4%
vidually, and the search strategies included both subject headings and Cannabis 27 9.9%
keywords to include a broad range of psychoactive substances. The Med- Other 11 4.0%

line search is included as an example in Supplementary Material. ∗

doesn’t sum to 100% because some
articles were about more than one sub-
stance
Study screening methods ∗∗
articles were placed in this cate-
gory if they were about illegal drugs
Screening took place in two stages: title / abstract and full text. Four broadly, with no focus on a specific
team members were involved. Title / abstract screening was conducted substance or class of substances
independently by two reviewers (JFC and TMW), with conflicts resolved
a third party (CS). Full-text screening was also conducted independently Table 4
by two reviewers (JFC and CR), with conflicts again resolved by a third Country of origin of synthesized articles∗∗ .
party (TMW). Proportionate agreement was 88.6% (Kappa 0.72) for ti- n %
tle / abstract screening and 86.4% (Kappa 0.73) for full-text screening,
United States 148 54.4%
indicating “substantial” inter-rater reliability (Landis & Koch, 1977).
Canada 37 13.6%
Australia 19 7.0%
United Kingdom 19 7.0%
Data extraction New Zealand 9 3.3%
Other (Europe) 25 9.2%
All eligible articles were obtained as—or in some cases converted Other (Asia) 7 2.6%
Other (Africa) 4 1.5%
into—searchable Portable Document Format (PDF) files. To find the Other (South America) 4 1.5%
phrase “public health approach” in each article, one reviewer (JFC) ran
∗∗ First author’s listed country of origin
separate searches for the words “public” and “health” using Adobe Ac-
robat’s “Find” function. We searched this way to avoid missing text due
to word breaks or line breaks. All sentences containing the phrase ‘pub-
lic health approach’ were extracted, as well as adjacent sentences when Appraisal
appropriate. (See Supplementary Material for sample extractions.) We
refrained from making inferences, relying instead on explicit statements A qualitative systematic review would generally appraise the in-
about the characteristics of a public health approach. Approximately cluded studies and report the approach and process (Page et al., 2021;
20,000 words were extracted for coding in this way. Tong et al., 2012). Since our review synthesizes opinion, not evidence,
The following background information was also extracted from each this is not applicable here. However, we extracted all articles’ reported
article: full reference, article type, substance(s) discussed, reported fund- funding sources and conflicts of interest to explore the possibility that
ing sources and conflicts of interest, and any sources the authors cited industry or government funding or affiliation might impact authors’ def-
in support of their definitions of a PHA. initions of a PHA. Eleven articles were notable in that regard, but there
were no differences in those articles’ definitions that would impact the
synthesis (see Supplementary Material for a summary).
Selection results and study characteristics

Selection results are summarized in Table 1. Software

Of the 272 eligible articles (see Supplementary Material for the
full list), 169 were original articles, research (quantitative or quali- Screening and extraction were conducted using Covidence
tative), or reviews (e.g. systematic or scoping); 100 were commen- (Veritas Health Innovation, 2022), a web-based collaboration soft-
tary/opinion/perspective, editorials, responses to other articles, or let- ware platform designed for systematic and other literature reviews.
ters to the editor; and 3 were “other” (two study protocols and one lec- Coding was completed using Taguette (Rampin & Rampin, 2021), an
ture). Tables 2-4 summarize the year of publication, subject matter, and open-source qualitative research tool that allows for text to be entered
country of origin of these articles. and collaboratively coded.

3
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

Coding Different interpretations of the term ‘public health approach’

Coding was inductive, with codes developed and evolving iteratively Part of the rationale for this systematic review was the observation
as the text of each article was assessed line by line. The main coder that while the concept of a public health approach is increasingly ap-
(JFC) coded all the articles, creating 121 codes. To verify, the second plied to substance use, this term is used in inconsistent and contradic-
coder (CR) separately used the resulting codebook to code 48 randomly tory ways—and often simply undefined—in the academic literature. Our
selected articles (25% of the sample), after which the two compared re- findings confirm this observation. Out of the 272 included articles:
sults. As a result of this verification, no codes were changed or removed,
• 47 used the term “public health approach” without defining it. Many
but five codes were added to four articles. Examples of coding are given
in the Supplementary Material file, as are the codebook and the coding mentioned PHAs only once, e.g. in the abstract or conclusion. Others
results for every included article. used the phrase more than once but did not explain its meaning,
perhaps assuming it to be self-explanatory.3
• While most articles used the term “public health approach” with ap-
Derivation of themes
proach meaning a strategy or a particular way of conceptualizing or
doing something, 32 instead used the term PHA to mean a measure
Themes emerged during coding. The main coder (JFC) distilled the
or intervention. For example: “Screening, Brief Intervention, Refer-
121 codes in the final codebook into 65 descriptive themes that emerged
ral and Treatment (SBIRT) is a comprehensive, integrated, public
during coding. For example, descriptive themes such as “A PHA includes
health approach to the delivery of early intervention and treatment
limits on advertising,” “A PHA includes limits on availability (location,
services for persons with substance use disorders or who are at risk”
hours or sale),” “A PHA includes regulations on packaging and labels
(McCance-Katz & Satterfield, 2012, p. 176).4
(e.g. plain packaging, health warning labels),” and nine others were • It was common for authors to define PHAs in part by contrasting
combined to form a single analytical theme: “A PHA includes regulation”
these with other approaches, e.g. a clinical or criminal justice ap-
(see Supplementary Material for the full list of analytical and descriptive
proach. For example: “The clinical approach provides intensive, ef-
themes). The main coder counted the number of articles that included
ficacious interventions to smokers who seek help, whereas the public
each theme, and, along with another reviewer (JR), explored associa-
health approach provides lower intensity interventions to a broader
tions between these themes, as well as their distribution over time and
spectrum of the smoking population” (Curry, 1993, p. 970). How-
across substances, via cross-tabulation and calculation of Pearson cor-
ever, 7 articles defined a PHA only in terms of what it is not.
relation coefficients. • 7 pointed out that the definition of a PHA is unclear or contested.
To discuss and agree on analytical themes, the team used an adapta-
For example: “What do we mean by the public health approach?
tion of the Nominal Group Technique, a “group process model” designed
Surprisingly enough, if one were to ask a group of public health
to ensure equal participation in idea generation and decision-making in
professional workers what the public health approach is, few of them
the context of small groups (Delbecq et al., 1975). The main author sent
would give the same answer” (Cross, 1967, p. 956).
the raw data, initial observations, and draft analytical themes to four re- • 5 noted that the definition of a PHA varies. For example: “Various
viewers (JR, CR, CS, TMW), along with three discussion questions. The
public health approaches to drugs exist. The model discussed here is
reviewers answered the questions via email, and two weeks later, met
a harm-reduction approach” (Neill, 2014, p. 375).
virtually. The meeting was planned to have three stages: 1) participants
answer each question one at a time in “round robin” format; 2) par- This last point is especially important. Four articles in this review
ticipants discuss the points of view presented in the previous step and (Berridge & Bourne, 2005; Mold, 2018; Room, 2004; Stenius, 2018) dis-
work towards consensus; 3) any disagreements or competing viewpoints cuss the evolution of public health approaches over time, in particular
/ interpretations are solved via vote. places, or for specific substances. They suggest that the meaning of this
Consensus on analytical themes was achieved during the meeting concept shifts over time, but also by substance and by geographic area.
without the need for a vote. In the words of Mold (2018, p. 94), there is “no single ‘public health’
approach to substance use: this can mean different things at different
Results times and in different places.” Accordingly, this section will begin by
reporting the results of the systematic review overall, then over time,
The term “public health approach” (PHA) has been used in the con- by substance, and finally by country.
text of substance use in the English-language academic literature at least
as far back as 1950 (see McCarthy, 1950). It has become more common Characteristics of a public health approach (overall)
over the years; the median year of publication of the 272 articles in-
cluded in this review is 2014. Subject matter evolves with time: From each article we extracted any definition or description of a
PHA and its characteristics. As mentioned, 47 articles used the term
• The articles from the 1950s and 1960s are about alcohol. PHA without explanation, leaving nothing to extract, and 32 used the
• The first article about illicit drugs appears in 1970; this topic be- term differently, taking “approach” to mean a measure or intervention
comes more common from 1990 onwards. rather than a strategy. This left 193 articles from which to synthesize
• Opioids first appear in 1973, but 90% of articles about opioids are definitions of a PHA. The most frequently mentioned characteristics and
from 2015 or later. components of a PHA are listed in Table 5 (for the full list, see Supple-
• Tobacco first appears in 1988.
• Cannabis first appears in 2008, but 74% of these articles are from
2014 or later. the over-representation of Canada in articles about cannabis mirrors that coun-
try’s debates about whether and how to legalize cannabis, which the federal
Most of the articles are from the US (54%) and Canada (14%). The US government proposed in 2015 and carried out in 2018.
3
is particularly over-represented in articles about opioids (77%), while The proportion of articles mentioning PHAs but making no attempt to define
Canada is over-represented in articles about cannabis (63%).2 them goes from 4% in 1950–1989 to 16% in 1990–2019 and 30% from 2020
onwards. This may suggest that as the term becomes more common, authors
increasingly assume it to be self-explanatory.
2 4
The over-representation of the United States in articles about opioids and Of these, 16 were referring to SBIRT as a “public health approach,” 5 were
the preponderance of articles on that topic from 2015 onwards likely reflect the referring to harm reduction, and 3 were referring to health warnings on alcohol
gravity and trajectory of the opioid overdose crisis in that country. Meanwhile or cannabis labels.

4
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

Table 5
Most commonly cited characteristics of a public health approach (n=193).

Number of articles

PREMISES & PRINCIPLES

What it rests on
A PHA is, or should be, evidence-based 46
A PHA is pragmatic / realistic / not moralistic about substance use 20
In a PHA, problematic substance use / SUDs / addiction defined as a health issue or disease 12
What it focuses on
A PHA focuses on populations 38
A PHA focuses primarily on environmental factors 22
A PHA does not centre / focus on individuals 20
What it is not
A PHA is distinct from a criminal justice approach 54
A PHA is distinct from a commercial approach 21
A PHA is at odds with the commercialization of psychoactive substances 14
A PHA involves countering industry (e.g. its political influence and discrediting research) 10
A PHA is distinct from a clinical approach 18
A PHA is not punitive / coercive 17
A PHA is distinct from / an alternative to prohibition 15
OBJECTIVES
A PHA attempts to reduce the harm associated with substance use 29
but not use per se 12
A PHA aims to get individuals to reduce or stop their substance use 24
A PHA aims to reduce aggregate / population-level consumption 21
A PHA aims to reduce risk, e.g. by focusing on modifiable risk factors 21
A PHA aims to protect / improve / promote health 17
A PHA aims to reduce death and disability and increase longevity 14
A PHA aims to reduce substance “misuse” / addiction 14
A PHA aims to delay the onset of substance use 11
COMPONENTS
A PHA includes regulation 70
regulating the price of psychoactive substances 36
limits on availability (location, hours of sale) 33
limits on advertising 29
minimum legal age 19
regulation of public use 15
regulation of products themselves (e.g. potency) 12
measures to address impaired driving 12
regulations on packaging and labels 10
A PHA includes SUD treatment 63
A PHA includes medication-assisted treatment, e.g. opioid agonist treatment 13
A PHA includes education 58
…targeting social norms around substance use / promoting “healthy behaviours” 21
Prevention efforts are part of a PHA 55
Primary prevention efforts are part of a PHA 14
Secondary prevention efforts are part of a PHA 15
A PHA includes harm reduction measures 39
A PHA includes needle / syringe exchange programs 13
A PHA ensures access to naloxone (overdose prevention) 10
A PHA includes screening and brief interventions 21
A PHA includes interventions for groups deemed to be at higher risk 18
A PHA addresses social determinants of health 17
PHA measures (e.g. education, prevention, treatment) should be based in the community 16
A PHA involves support for research 14
A PHA ensures access to services (e.g. primary care, mental health) 13
A PHA includes early intervention 12
A PHA includes support for “rehabilitation” and recovery 12
A PHA includes monitoring and surveillance 11
A PHA has a (limited) role for law enforcement 11
A PHA involves school-based education / prevention 10

mentary Material). They are sorted into three categories: premises and • A PHA includes substance use disorder treatment (n=63). Most arti-
principles, objectives, and components. cles mentioned that treatment generally is a part of PHAs, but oth-
Among the synthesized articles, the five most frequently cited char- ers specified certain types, e.g. medication-assisted or psychosocial
acteristics and components of a PHA were the following (also see Table 6 treatment.
for illustrative quotes): • A PHA includes education (n=58). This category consists of interven-
tions intended to communicate information to the public, for ex-
• A PHA includes regulation (n=70). Some articles simply mentioned ample through media campaigns, public service announcements, or
that a PHA involves regulating psychoactive substances, while oth- school-based education. Most often the desired content of this ed-
ers gave specific examples; the most commonly cited types of regu- ucation was not discussed, but 21 articles stated that PHAs involve
lation were price controls (n=36), limits on availability (n=33), and targeting social norms around substance use and promoting “healthy
advertising restrictions (n=29). behaviours.”

5
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

Table 6
Most commonly cited characteristics of a public health approach – illustrative quotes.

Characteristic Illustrative quote

Regulation “The key levers for reducing or preventing alcohol-related problems in these models are regulating the sales and distribution of alcohol (e.g., through
price controls, restrictions on sales venues and hours) and limiting the demand for it (e.g., through curtailing advertising).” (Herd, 2010, p. 1228)
Treatment “A comprehensive public health approach should offer accessible evidence-based prevention, treatment, and recovery options to drug users, and engage
those who commit criminal offences in evidence-based treatment during and following, or in lieu of, incarceration, to prevent relapse and recidivism.”
(Volkow et al., 2017, p. 213)
Education “Public health approaches to preventing smokeless tobacco use should include development of skills for responding to social influences to use tobacco,
as well as education on short-and long-term health consequences.” (Tomar & Giovino, 1998, pp. 24–25)
Prevention “A public health approach seeks to reduce the total burden of alcohol problems to society through primary, secondary and tertiary prevention.” (Sobell
& Sobell, 1995, p. 1150)
“Not criminal “A criminal justice approach focuses on assigning blame whereas a public health approach focuses on minimizing harm and preventing the spread of the
justice” problem.” (Weisheit & Klofas, 1998, p. 200)

• Prevention efforts are part of a PHA (n=55). Prevention is a broad Phase 1: 1950–1980
category.5 Most articles mentioned only broad types of prevention This period sees the application of basic public health princi-
(e.g. primary or secondary) that they envision as part of a PHA, but ples—particularly a population-level focus and attention to the social
a few also offered examples, e.g. early detection of substance use environment—to substance use. For example, Querido (1954) states:
disorders as a form of secondary prevention.
What are the characteristics in a public-health approach to any prob-
• A PHA is distinct from a criminal justice approach (n=54). Here, articles
lem? Our first answer to this question might be that this approach is
presented PHAs as a contrast or alternative to drug policies that are
scientific—that is, based on the methods of natural science, based on
centred around criminalization and incarceration.
the laws of cause and effect.... In the second place, the public-health
The most commonly cited characteristic of a PHA, regulation, ap- approach is based on the firm belief that any pattern of behavior
pears in only 36% of coded articles, which would seem to suggest a low shown by man is the result of an interaction of various forces. Since
level of agreement overall. But only a minority of the articles in this re- man is a social being, his behavior can only be the result of his indi-
view provided an in-depth discussion of PHAs. Many were brief in their vidual drives as they are expressed in and through the social milieu
descriptions; 25 articles offered only a single characteristic of a PHA and in which he lives…. In fact, the chief purpose of public-health activ-
another 23 gave only two, accounting for a quarter of the coded arti- ity, should it be successful, will be to analyze the various conditions
cles. On the other hand, 42 articles offered 10 or more characteristics and to trace [these] interrelations. (pp. 470-471)
of a PHA. If we focus on those more detailed articles, the five most fre-
Similarly, per Khanna (1965, p. 342), “Alcoholism should be consid-
quently cited characteristics of a PHA are the same, but with a higher
ered a public health problem and as such should be subjected to meth-
level of agreement:
ods applicable to other health problems involving chronic diseases and
• Regulation: n=29 (69%) mental disorders.” Those methods include monitoring and surveillance,
• Treatment: n=26 (62%) early diagnosis, and education.
• Education: n=26 (62%) In the 1970s we see the first article on illicit drugs. Here the emphasis
• Prevention: n=23 (55%) is on prevention, especially primary. Writing about the United States,
• Not criminal justice: n=18 (43%) Brown (1975, p. 115) calls primary prevention “the weak suit in the
national public health approach to narcotic addiction and drug abuse”
Characteristics of a public health approach (over time) due to an emphasis on treatment. He states that “The chief prevention
target in the future must logically be the nation’s adolescents, for whom
As alluded to above, the subject matter of articles mentioning PHAs and with whom we must develop convincing drug abuse educational
changes over time, with alcohol most commonly mentioned early on programs and alternative pursuits which are more appealing than the
and cannabis and opioids occurring more frequently from the mid-2010s drug experience” (Brown (1975, p. 119).
onwards (see Table 7). As seen in Table 8 above, the most commonly cited components of a
Our analysis suggests an evolution over time in the definition of a PHA in this period are prevention, treatment, and support for research.
PHA that can be roughly divided into three phases:
• Phase 1, 1950–1980: the application of basic public health concepts Phase 2: 1981–2010
(e.g. prevention, population focus) to substance use. In the 1980s we see an increased focus on product regulation as a
• Phase 2, 1981–2010: increased emphasis on environmental factors. means of primary prevention—often with the caveat that a PHA should
• Phase 3, 2011–present: heightened focus on evidence-based policy not be focused on individual behaviour. For example see Lieber (1982):
and harm reduction. In the field of alcoholism, policy makers have been sharply divided
Table 8 shows the most commonly cited characteristics of a PHA between two camps: on the one side are the advocates of a conven-
by period, and these phases are discussed and illustrated with quotes tional public health approach against alcoholism through control of
below. the offending agent, namely alcohol (which they blame for the dam-
ages of alcoholism); on the other side are those who espouse the
5
thesis that it is the minority of drinkers who cannot handle alcohol
Many measures included in this review, from education to screening to treat-
that are at the core of the alcoholism problem. (p. 171)
ment, can be defined as prevention. We did not make inferences about the au-
thors’ intentions or “translate” these concepts; items were coded “prevention” Similarly, see Beauchamp (1985):
(or primary, secondary, or tertiary prevention) only if the authors used those
words. Similarly, tertiary prevention may be considered to include treatment, The strategy of the public health approach is not to alter individual
but if an author mentioned tertiary prevention as part of a PHA without speci- capacities or to instill a particular moral code for individuals. It is
fying whether that involves treatment, we coded that statement as tertiary pre- rather to employ legal and regulatory means to check consumption
vention and did not code it as treatment. of alcohol by drinkers generally, and to limit the discretion of the

6
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

Table 7
Subject matter of coded articles over time∗ .

Total 1950 – 1980 1981 – 2010 2011 – present

(n=193) (n=12) (n=81) (n=100)

Alcohol 48 8 21 19
General / unspecified 41 1 25 15
Illicit drugs 41 2 17 22
Opioids 29 1 2 26
Tobacco 21 0 14 7
Cannabis 19 0 3 16
∗
columns don’t sum to 100% because four articles are about more than one substance

Table 8 lication for articles mentioning it as part of a PHA is 2006, as opposed

Most commonly cited characteristics of a public health approach over to 2011 for the 193 coded articles.
time.

1950 – 1980 (n=12) 1981 – 2010 (n=81) 2011 – present (n=100) Phase 3: 2010–present
Prevention (8) Regulation (34) Regulation (34) Over half of the coded articles are from this time period. The ar-
Treatment (7) Education (33) Treatment (33) ticles here are more diverse (by substance discussed as well as coun-
Research (5) Treatment (23) Not criminal justice (31) try of origin) and thus more challenging to characterize, but among
Education (3) Prevention (23) Evidence-based (27)
the noticeable changes is an increased focus on harm reduction. Per
Population focus (3) Not criminal justice (21) Harm reduction (25)
Yeo et al. (2021), a PHA to opioids treats substance use disorders as
a health issue rather than a legal one, moving from criminalization to
harm reduction. They note that ultimately “upstream measures (tackling
alcohol industry to sell and promote the consumption of its products. the social determinants of health) are more effective public health strate-
(p. 34) gies to control the epidemic. Nevertheless, in the meantime, evidence-
based harm reduction strategies have to be employed to mitigate the
In the 1990s we see the first calls for an across-the-board PHA to harm” (Yeo et al., 2021, p. 245).
substance use that includes both legal and illicit drugs, beginning with The quote above mentions that harm reduction strategies should be
Erickson (1990) and Mosher and Yanagisako (1991). These authors and evidence-based; this too is a trend that accelerates in this time period.
others revisit the agent / host / environment model of disease cau- According to Nolan et al. (2020):
sation (also known in public health as the “epidemiologic triad”—see Often the public health approach is not clearly described, resulting
Centers for Disease Control and Prevention, 2012). Per Erickson (1990, in a definition of what it is not (e.g. non-punitive, non-coercive)
p. 564), “The key concepts in the public health model are the agent, rather than what it is. The absence of a clearly articulated approach
host and environment. Prevention is accomplished by reducing exposure grounded in public health practice has allowed for responses that,
to the agent (drug) either by modifying the environment (community) while non-punitive, are not evidence-based or data-driven…. A pub-
or by altering the susceptibility of the host (person/user).” Mosher & lic health approach [should] be grounded in science and based on
Yanagisako (1991, p. 309) make this point as well but add that “current the best available evidence. (p. 959)
priorities should be substantially redefined to target environmental risk
factors rather than agent and non-health host factors.” Room (2015, p. 367) also points to the importance of science, stating
In this period, PHAs are also increasingly presented as a pragmatic that “a strong strand in [public health approaches] has been the focus
or realistic alternative to prohibitionist drug policy. For instance Jonas on using evidence to inform policies affecting the health of the people.”
(1994, p. 248) contrasts PHAs with prohibition and states that “rather There is also more of an emphasis on treatment in this period than
than reaching for the unreachable, the Public Health Approach to the the one before. This may be due to the marked increase in articles about
drug problem sets as its goal the reachable: broad-based harm reduc- opioids (they account for 3% of articles in 1981–2010 but 26% from
tion for users, their contacts, and society as a whole.” Similarly, the 2011 onwards)—most of which mention treatment as a component of a
National Association for Public Health Policy (1999, pp. 268, 275) states PHA. This brings us to an important point: this observed evolution over
that its proposed PHA to illicit drugs “is grounded in objective reali- time appears to be linked to, or at least partly a function of, changes in
ty…. a policy of compassion and understanding, with the ultimate aim the dominant substance use challenges of the period in question.
of improving the human condition,” and characterizes it as “a basis for
mitigating the disastrous consequence of the ‘War on Drugs.’” Characteristics of a public health approach (by substance)
In this period we also see the first hints of a broader debate about
the objectives of PHAs. In 1999 we see expressed for the first time the It makes intuitive sense that definitions of a PHA might differ de-
idea that a PHA is primarily concerned with the harms associated with pending on the substance(s) being discussed. For example, regulation
substance use, and not necessarily with use per se (see Riley, 1999, and may not be top of mind to someone writing about illicit substances, and
NAPHP, 1999). On the other hand, there are in total 24 articles in our similarly, the idea that substance use should not be treated as a crim-
sample that argue a PHA involves encouraging people to stop or reduce inal justice issue may not occur to someone writing about alcohol. If
their substance use; 17 of these were published between 1988 and 2009. we examine substances separately, each drug category has at least one
According to these authors, a PHA counts as its goals “to persuade in- characteristic of a PHA that was mentioned by more than half of the
dividuals to modify some behaviors to reduce harmful consequences” coded articles:
(Weisheit & Klofas, 1998, p. 202) and to “change norms and attitudes
community-wide” (Jason et al., 2005, p. 286). Public education cam- • Alcohol: Regulation, 26 of 48 (54%)
paigns and various forms of denormalization are proposed as the means • Cannabis: Regulation, 13 of 19 (68%)
to achieve this end. As seen in Table 8 above, education is the second • Illicit drugs: Not criminal justice, 26 of 41 (63%)
most commonly cited component of a PHA in this period. Education de- • Opioids: Treatment, 16 of 29 (55%)
clines in importance from this point onwards; the median year of pub- • Tobacco: Regulation, 13 of 21 (62%); Education: 12 of 21 (57%)

7
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

Table 9
Correlations between frequently cited characteristics of a public health approach and substance.

POSITIVE NEGATIVE

General Prevention r = .21, p = .004

Alcohol Not centred on individuals r = .35, p < .001 Harm reduction r = -.29, p < .001
Reduce aggregate use r = .34, p < .001 “Not criminal justice” r = -.25, p = .001
Environmental focus r = .32, p < .001 Treatment r = -.17, p = .018
Population focus r = .29, p < .001
Regulation r = .21, p = .002
Tobacco Reduce individual use r = .27, p < .001 “Not criminal justice” r = -.18, p = .012
Education r = .21, p = .004 Treatment r = -.17, p = .017
Regulation r = .19, p = .009
Illicit drugs “Not criminal justice” r = .41, p < .001 Regulation r = -.29, p < .001
Harm reduction r = .21, p = .003 Population focus r = -.23, p = .002
Treatment r = .18, p = .013 Reduce aggregate use r = -.18, p = .012
Opioids Harm reduction r = .22, p = .002
Treatment r = .20, p = .005

Cannabis Reduce risk r = .28, p < .001

“Not commercialization” r = .28, p < .001
Regulation r = .22, p = .002

DF=191; significance tests two-tailed

Table 10
Most commonly cited characteristics of a public health approach to alcohol, by period.

1950 – 1980 (n=8) 1981 – 2010 (n=21) 2011 – present (n=19)

Treatment (6) Regulation (13) Regulation (11)

Prevention (5) Population focus (10) Population focus (7)
Education (3) Environmental focus (8) Reduce aggregate use (6)
Environmental focus (3) Not centred on individuals (8) Not centred on individuals (5)
Screening (3) Education (7) “Not commercialization” (5)
Prevention (7)
Reduce aggregate use (7)

To explore associations between PHA definitions and particular sub- (Raninen & Livingston, 2020).8 Education also declines from around
stances, we conducted a basic quantitative analysis. Pearson correla- 35% in the earlier two periods to just 1% from 2011 onwards. In fact,
tion coefficients were calculated for frequently cited PHA characteris- during this later period we also see articles arguing that education aimed
tics, which we defined as any item cited by 20 or more articles, versus at individuals is a relatively weak measure likely to be preferred by in-
substance discussed.6 Several weak to moderate correlations were ob- dustry over more effective measures like regulation of the product and
served; those significant at the p < .05 level are displayed in Table 9. how it is sold (see for instance Moewaka Barnes et al., 2016). This also
This table shows a high degree of variation between substances, coincides with the appearance of articles documenting industry opposi-
supporting the notion of substance-specific differences in definitions of tion to public health approaches that would tackle the commercializa-
a PHA. There are also similarities in how PHAs are described for le- tion of alcohol (see for instance Parry, 2014). In sum, definitions of a
gal / regulated substances (alcohol, tobacco, and in some jurisdictions, PHA to alcohol show a clear evolution over time.
cannabis) versus illicit ones.7 This raises the question of whether the ob-
served changes over time can simply be attributed to substance-specific
Characteristics of a public health approach (by country)
differences in definitions of a PHA. To explore this question, we can
look at changes over time for a particular substance. There is just one
It might also be expected that definitions of a PHA would differ by
substance – alcohol – for which there are enough articles across time pe-
country. Since the United States accounts for over half of the reviewed
riods to conduct this comparison. Table 10 shows the most commonly
articles, with Canada second at just under 14%, this is more difficult to
cited characteristics of a public health approach to alcohol by period.
assess. Given the significant differences by substance described above,
The most evident evolution here is the reduced importance of treat-
we would ideally want to control for that factor when comparing coun-
ment, prevention, and education over time. Treatment was cited by
tries, but the small number of cases makes any meaningful analysis im-
75% of articles in 1950-80, but only 10% in 1981-2010 and 5% from
possible.9 Here, we can simply reiterate that variations in PHA defini-
2011 onwards. This can be explained by the rise of the “new pub-
tions are likely linked to changes in the dominant substance use chal-
lic health approach” to alcohol that emerged from Scandinavia in the
lenges of the period and place in question. We can also point to three
1970s and, as explained by Room (2004, p. 338), proposed “refocus-
ing on rates of alcohol-related harm in the population as a whole,”
with a concomitant emphasis on regulations like price controls and 8
As mentioned in the Methodology section, in conducting this review we ex-
availability restrictions over individual-level or clinical interventions. tracted any sources the authors cited in support of their definitions of a PHA.
This approach was profoundly influential in the field of alcohol policy The two most cited sources are 1) a 1975 book outlining the tenets of this "new
public health approach" to alcohol (Bruun et al., 1975), and 2) a well-known
book now in its third edition (Babor et al., 2022) that features many of the same
6
This excludes regulation of price and limits on availability and advertising, authors and follows “in the Bruun et al. tradition” (Room, 2004, p. 339). For
as those items are included in the “regulation” variable. the list of most cited sources, see Supplementary Material.
7 9
It seems plausible that most of the negative correlations are due to authors For example, there are 22 American articles about alcohol but only 4 from
not considering those characteristics relevant, rather than disagreeing with them Canada; conversely, there are 12 Canadian articles about cannabis but only 3
as part of the definition of a PHA. from the United States.

8
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

articles captured in this review that discuss the form PHAs have taken less frequently cited as a feature of a public health approach. Reg-
in particular places: Mold (2018) and Berridge and Bourne (2005) for ulation has been the most commonly cited characteristic from the
the United Kingdom, and Stenius (2018) for Nordic countries. 1980s onwards.
• Definitions of a public health approach differ by substance. There is no
Areas of disagreement across-the-board consensus on the definition of a PHA to substance
use, but there is a substantial level of agreement when it comes to
This paper has focused primarily on finding commonalities in def- PHAs to specific substances or classes of substances. There are also
initions of a PHA, but there are also important differences of opinion similarities in how PHAs are described for legal / regulated sub-
among the reviewed articles. The first is the question of whether a stances (alcohol, tobacco, and in some jurisdictions, cannabis) com-
PHA should aim to influence or change individual behaviour, and if pared to illicit ones.
so, how. There are four main perspectives on this issue, arguing that a
PHA should: These findings reflect the work of others, notably Berridge and
Bourne (2005), Mold (2018), and Room (2004), who have documented
1) encourage individuals to reduce or stop their substance use (24 ar- changes over time and by substance in definitions of public health ap-
ticles) proaches to substance use. One unique contribution of this paper is a
2) focus on harm, rather than use per se (12) summary of the characteristics of a public health approach for which
3) reduce consumption by targeting social norms, e.g. by promoting there is a high level of agreement, by substance, in the academic litera-
healthy behaviours/ “lifestyles” (21) ture. Table 11 summarizes those characteristics.
4) reduce per-capita consumption through impersonal population-level These findings will be of value to policymakers as well as public
measures (21) health and drug policy workers as they can shed light on the policy chal-
Time and substance play a role here. For example, the first perspec- lenges mentioned in the introduction. According to this consensus defi-
tive seems to be an older viewpoint (median year of publication: 1999) nition summarized in Table 11, a public health approach to illicit drugs
most closely associated with tobacco. And while examples of all four and opioids would, at a minimum, 1) move away from treating sub-
perspectives can be found for alcohol, the median year of publication stance use as a criminal justice issue, 2) be evidence-based, and include
for articles stating that a PHA to alcohol aims to reduce consumption both 3) SUD treatment (including medication-assisted treatment) and 4)
by targeting social norms is 1988, while for articles stating that a PHA harm reduction measures (e.g. needle / syringe exchange programs). For
to alcohol should reduce per-capita consumption through impersonal jurisdictions that have legalized cannabis or are considering doing so,
population-level measures it is 2014. a public health approach would 1) rest on the best available evidence,
In addition, there are nine articles explicitly including a strong role 2) aim to reduce harm by focusing on and addressing modifiable risk
for the state in the implementation of public health approaches. Room factors, e.g. age of initiation and intensity of use, 3) minimize the com-
(2015, p. 367) notes that a more controversial aspect to PHAs has been mercialization of cannabis, and 4) include regulation, e.g. of price, avail-
the “tendency to privilege the population level over the individual level ability, and marketing / promotion of cannabis. (In the case of cannabis,
in public health policymaking…. Often this has been at the expense of it is worth noting that this differs greatly from the policy models seen in
individual choices, liberties and interests.” He argues that in a PHA, the US states that have legalized cannabis (Blanchette et al., 2022), and
action can be “taken for the common good, in disregard of individual Canada, despite its federal government’s stated commitment to a pub-
preferences.” Along similar lines, MacGregor (2012, p. 435) states that lic health approach to cannabis, lacks some of these measures as well
“Public health approaches can involve considerable state intervention (Crépault & Jesseman, 2022).) These characteristics of a PHA are only
in the lives of citizens (including redistribution of resources) if this is the ones for which there was the highest level of agreement. The longer
considered to be for the general good.” But five articles argue that a list of commonly cited characteristics, presented in Table 5, could be
PHA balances public health with freedom and liberty, e.g. the National looked at as a menu of possibilities for a public health approach to sub-
Association for Public Health Policy’s call for a public health approach stance use, with the choice of strategies and measures depending on the
to “drug abuse” in which “respect for individuals’ civil liberty and their substance and its current – or desired – legal status.
right to the pursuit of happiness” are “inherent” (NAPHP, 1999, p. 268). Twenty of the articles in this review included criticism of public
Similarly, Kalant (1999, p. 25) states that in a PHA, “the maximum pro- health approaches. The most common criticism (n=7) was that PHAs,
tection of the public from harm would be balanced with the least possi- with their population-level focus, neglect individual factors like gen-
ble restriction of freedom.” While these points of view are not entirely der, racialization, and class, and thus “potentially play upon and exacer-
contradictory, they would appear to be in tension. Time may play a role bate social divisions and inequality” (Marron, 2017, p. 883). Conversely,
here; the articles advocating a strong role for the state tend to be more three articles argue that PHAs are overly focused on individuals, to the
recent than those calling for balance (median years of publication: 2015 detriment of structural factors. As an example, Collins et al. (2019) argue
and 1999, respectively). that research conducted with people who use drugs “has demonstrated
how the social-structural conditions (e.g. social, political, and economic
institutions) of individuals’ environments produce or mitigate harm…
Discussion
such elements were unaddressed by public health approaches that had
emphasized individual-focused behaviour change.” Finally, four articles
High-level findings of this study, which we believe to be the first to
criticize PHAs for ignoring potential benefits of substance use. Stimson
systematically review the academic literature for definitions of a public
(2007, p. 68) notes, for instance, that “The public health approach to
health approach, can be summarized as follows:
reducing harms from illicit drugs and alcohol has yet to engage with the
• The concept of a public health approach to substance use has gained issue of drugs and pleasure.”
currency. The term has been used more frequently over time, with What links all of these critiques is that they pertain to the extent to
2014 being the median year of publication of the articles in this which PHAs should in theory – or do in actuality – focus on individual-
review. At the same time, it is often—and increasingly—undefined. level factors. That is also the key issue at play in the two areas of dis-
17% of articles did not define the term, and this percentage rises to agreement discussed in the previous section: whether PHAs should aim
30% among articles written since 2020. to influence or change individual behaviour around substance use, and
• Definitions of a public health approach have changed over time. Most no- the extent to which the state should implement measures that privilege
tably, the past decade has seen an increased emphasis on evidence- public health over individual preferences. Whether there are “correct”
based measures and harm reduction. Conversely, education has been positions here is an open question, and indeed, the proper balance likely

9
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

Table 11
Consensus definition – characteristics of a public health approach, by substance.

PREMISES & PRINCIPLES

What it rests on
A PHA is, or should be, evidence-based
What it focuses on
A PHA focuses on populations Alcohol
A PHA focuses primarily on environmental factors Alcohol
A PHA does not centre / focus on individuals Alcohol
What it is not
A PHA is distinct from a criminal justice approach Illicit drugs
A PHA is distinct from a commercial approach Cannabis

OBJECTIVES
A PHA aims to reduce aggregate / population-level consumption Alcohol
A PHA aims to reduce risk, e.g. by focusing on modifiable risk factors Cannabis
A PHA aims to get individuals to reduce or stop their substance use Tobacco
COMPONENTS
A PHA includes regulation Alcohol, Cannabis, Tobacco
A PHA includes SUD treatment Illicit drugs, Opioids
A PHA includes education Tobacco
Prevention efforts are part of a PHA General
A PHA includes harm reduction measures Illicit drugs, Opioids

depends on the substance, the time, and the place – but there is a need Conclusion
for greater clarity here.
Public health approaches to substance use are frequently called for
but inconsistently defined. This systematic review and thematic syn-
Limitations thesis of the English-language academic literature found that the term
has been used at least as far back as 1950 but has gained currency
This systematic review has important limitations. Because we re- in recent years, and that definitions of a public health approach have
frained from making inferences, relying instead on explicit statements changed over time and differ by substance. The most commonly cited
from the authors, we likely missed material the authors intended as def- characteristic of a PHA for alcohol, cannabis, and tobacco was regula-
initions of a public health approach. The review’s focus on a particular tion, e.g. of price and availability; for illicit drugs, the most commonly
phrase, “public health approach,” de facto leaves out approaches that cited characteristic of a PHA was that it is distinct from a criminal jus-
might have the same objectives and characteristics but different termi- tice approach. Even by substance, definitions evolve over time: in the
nology, e.g. “public health framework” or “public health strategy.” This past decade, there has been an increased emphasis on evidence-based
could explain the fact that this review’s earliest article on tobacco is measures and harm reduction, while education has been less frequently
from 1988, despite public health involvement in tobacco control having cited as a feature of a public health approach. Overall, while there is no
begun in earnest more than two decades earlier (Alberg et al., 2014). across-the-board consensus on the definition of a public health approach
We believe this is acceptable, as our goal was to understand a specific to substance use, there is a substantial level of agreement when it comes
phrase that is now in common use. to PHAs to specific substances or classes of substances, and similarities
More significantly, by focusing on English-language literature, this in how such approaches are described for legal, regulated substances
systematic review neglects drug policy views from most of the world. versus illicit ones.
85% of the articles in this review have a first author from the United This review also found areas of disagreement and criticism regard-
States, Canada, United Kingdom, Australia, or New Zealand. It also ex- ing the extent to which public health approaches should, or do, focus on
cluded grey literature, for reasons described above. For these reasons, individual-level factors – notably whether PHAs should aim to influence
our findings cannot be assumed to be representative of global or civil- or change individual behaviour around substance use, and the extent to
society drug policy debates and should not be generalized beyond the which the state should implement measures that privilege public health
English-language academic literature. over individual preferences. Policymakers, academics, and others devel-
Coding was conducted by one researcher, with a second researcher oping or implementing public health approaches to substance use should
verifying. Krippendorff (2004, p. 277) identifies three types of coding be explicit about their aims and objectives, as well as the premises and
reliability: stability (coder behaviour remaining consistent over time), assumptions underlying them.
replicability (multiple coders achieving the same results), and accuracy
(assessing coding according to a pre-agreed codebook). Having coding
done primarily by a single researcher immersed in every stage of the Ethics approval
research can mean higher levels of stability and accuracy, but lower
levels of replicability. The authors declare that the work reported herein did not require
Finally, while thematic synthesis is considered a strong methodol- ethics approval because it did not involve animal or human participa-
ogy well suited to synthesizing qualitative evidence and generating hy- tion.
potheses, it has been criticized for masking differences in the quality
of included articles, which can affect confidence in the resulting syn- Declarations of Interest
thesized findings (Lucas et al., 2007; Noyes et al., 2020). Since expert
opinion and conjecture are in fact the object of this study, we find this The authors declare that they have no known competing financial
acceptable. The review also does not shed any light on the impacts or interests or personal relationships that could have appeared to influence
effectiveness of the measures authors proposed as part of a PHA. the work reported in this paper.

10
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

Supplementary materials Herd, D. (2010). Community mobilization and the framing of alcohol-related prob-
lems. International Journal of Environmental Research and Public Health, 7(3), 1226.
10.3390/ijerph7031226.
Supplementary material associated with this article can be found, in Human Rights Watch. (2021). Nearly 100 groups oppose Biden’s plan to make
the online version, at doi:10.1016/j.drugpo.2023.103958. misguided temporary drug criminalization policy permanent. Retrieved from
https://www.hrw.org/news/2021/10/22/nearly-100-groups-oppose-bidens-plan-ma
ke-misguided-temporary-drug-criminalization
References Jason, L., Pokorny, S., Muldowney, K., & Velez, M. (2005). Youth tobacco sales-to-minors
and possession-use-purchase laws: A public health controversy. Journal of Drug Edu-
Alberg, A. J., Shopland, D. R., & Cummings, K. M. (2014). The 2014 Surgeon General’s cation, 35(4), 275–290. 10.2190/1f1r-kgtl-pvde-v9ej.
Report: Commemorating the 50th anniversary of the 1964 Report of the Advisory Jonas, S. (1994). A public health approach to reducing harm from drug use. American
Committee to the US Surgeon General and updating the evidence on the health con- Journal of Health Promotion, 8(4), 247–251. 10.4278/0890-1171-8.4.247.
sequences of cigarette smoking. American Journal of Epidemiology, 179(4), 403–412. Joranson, D. E., & Gilson, A. M. (2006). Wanted: A public health approach to prescription
10.1093/aje/kwt335. opioid abuse and diversion. Pharmacoepidemiology and Drug Safety, 15(9), 632–634.
Babor, T. F., Casswell, S., Graham, K., Huckle, T., Livingston, M., Österberg, E., Rehm, J., 10.1002/pds.1293.
Room, R., Rossow, I., & Sornpaisarn, B. (2022). Alcohol: No ordinary commodity (3rd Kalant, H. (1999). Differentiating drugs by harm potential: The rational versus the feasible.
ed.). Oxford, UK: Oxford University Press. Substance Use & Misuse, 34(1), 25–34. 10.3109/10826089909035633.
Babor, T. F., Caulkins, J., Fischer, B., Foxcroft, D., Humphreys, K., Medina-Mora, M. E., Khanna, J. M. (1965). A public health approach to alcoholism. Canadian Journal of Public
Obot, I., Rehm, J., Reuter, P., Room, R., Rossow, I., & Strang, J. (2018). Drug policy Health, 56, 335.
and the public good (2nd ed.). Oxford, UK: Oxford University Press. Kolodny, A., Courtwright, D. T., Hwang, C. S., Kreiner, P., Eadie, J. L., Clark, T. W., &
Barnett-Page, E., & Thomas, J. (2009). Methods for the synthesis of qualitative research: A Alexander, G. C. (2015). The prescription opioid and heroin crisis: A public health
critical review. BMC Medical Research Methodology, 9, 59. 10.1186/1471-2288-9-59. approach to an epidemic of addiction. Annual Review of Public Health, 36(1), 559–
Beauchamp, D. E. (1985). Alcohol and the republic. Journal of Drug Issues, 15(1), 29–38. 574. 10.1146/annurev-publhealth-031914-122957.
10.1177/002204268501500104. Krippendorf, K. (2004). Content analysis: An introduction to its methodology (4th ed.). Thou-
Berridge, V., & Bourne, S. (2005). Illicit drugs, infectious disease and public health: A sand Oaks, CA: Sage.
historical perspective. Canadian Journal of Infectious Diseases & Medical Microbiology, Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical
16(3), 193. 10.1155/2005/530160. data. Biometrics, 33(1), 159–174. 10.2307/2529310.
Blanchette, J. G., Pacula, R. L., Smart, R., Lira, M. C., Boustead, A. E., Caulkins, J. P., Leung, M. (2019). The opioid crisis: The United States and Canada’s
Kilmer, B., Kerr, W. C., Treffers, R., & Naimi, T. S. (2022). Rating the com- fentanyl epidemic. Retrieved from https://www.wilsoncenter.org/
parative efficacy of state-level cannabis policies on recreational cannabis markets article/the-opioid-crisis-the-united-states-and-canada
in the United States. International Journal of Drug Policy, 106, Article 103744. s-fentanyl-epidemic
10.1016/j.drugpo.2022.103744. Lieber, C. S. (1982). A public health approach for the control of the dis-
Brown, B. S. (1975). Drugs and public health: Issues and answers. Annals ease of alcoholism. Alcoholism: Clinical & Experimental Research, 6(2), 171–177.
of the American Academy of Political and Social Science, 417, 110–119. 10.1111/j.1530-0277.1982.tb04958.x.
10.1177/000271627541700111. Lucas, P. J., Baird, J., Arai, L., Law, C., & Roberts, H. M. (2007). Worked examples of alter-
Bruun, K., Edwards, G., Lumio, M., Mäkelä, K., Pan, L., Popham, R. E., Room, R., native methods for the synthesis of qualitative and quantitative research in systematic
Schmidt, W., Skog, O.-J., Sulkunen, P., & Österberg, E. (1975). Alcohol control policies reviews. BMC Medical Research Methodology, 7, 4. 10.1186/1471-2288-7-4.
in public health perspective. Helsinki, Finland: Finnish Foundation for Alcohol Studies. MacGregor, S. (2012). Do we need a European drugs strategy? Drugs: Education, Prevention
Centers for Disease Control and Prevention. (2012). Introduction to epi- & Policy, 19(6), 429–435. 10.3109/09687637.2012.713411.
demiology, section 8: Concepts of disease occurrence. Retrieved from Marron, D. (2017). Smoke gets in your eyes: What is sociological about cigarettes? Socio-
https://www.cdc.gov/csels/dsepd/ss1978/lesson1/section8.html logical Review, 65(4), 882–897. 10.1111/1467-954X.12404.
Centers for Disease Control and Prevention. (2019). CDC’s efforts to pre- McCance-Katz, E. F., & Satterfield, J. (2012). SBIRT: A key to integrate prevention and
vent opioid overdoses and other opioid-related harms. Retrieved from treatment of substance abuse in primary care. American Journal on Addictions, 21(2),
https://www.cdc.gov/opioids/pdf/strategic-framework-factsheet_jan2019_508.pdf 176. 10.1111/j.1521-0391.2011.00213.x.
Collins, A. B., Boyd, J., Cooper, H. L. F., & McNeil, R. (2019). The intersectional risk en- McCarthy, R. G. (1950). Public health approach to the control of alcoholism. American
vironment of people who use drugs. Social Science and Medicine, 234, Article 112384. Journal of Public Health, 40(11), 1412–1417. 10.2105/ajph.40.11.1412.
10.1016/j.socscimed.2019.112384. McLellan, A. T. (2017). Substance misuse and substance use disorders: Why do they matter
Crépault, J.-F., & Jesseman, R. (2022). Regulating the legal cannabis market: How is Canada in healthcare? Transactions of the American Clinical and Climatological Association, 128,
doing?. Toronto, ON: Centre for Addiction and Mental Health. 112–130.
Crépault, J.-F., Watson, T. M., Strike, C., Bonato, S., & Rehm, J. (2021). What is a public Moewaka Barnes, H., McCreanor, T., Goodwin, I., Lyons, A., Griffin, C., & Hutton, F
health approach to substance use? Protocol for a qualitative systematic review. BMJ (2016). Alcohol and social media: Drinking and drunkenness while online. Critical
Open, 11(10), Article e055991. 10.1136/bmjopen-2021-055991. Public Health, 26(1), 62–76. 10.1080/09581596.2015.1058921.
Cross, J. N. (1967). Epidemiologic studies and control programs in alcoholism. I. Pub- Mold, A. (2018). Framing drug and alcohol use as a public health problem
lic health approach to alcoholism control. American Journal of Public Health and the in Britain: Past and present. Nordisk Alkohol & Narkotikatidskrift, 35(2), 93.
Nation’s Health, 57(6), 955–964. 10.2105/AJPH.57.6.955. 10.1177/1455072518765836.
Csete, J., & Wolfe, D. (2017). Seeing through the public health smoke-screen in drug Mosher, J. F., & Yanagisako, K. L. (1991). Public health, not social warfare: A public health
policy. International Journal of Drug Policy, 43, 91–95. 10.1016/j.drugpo.2017.02.016. approach to illegal drug policy. Journal of Public Health Policy, 12(3), 278–323.
Curry, S. J. (1993). Self-help interventions for smoking cessation. Journal of Consulting and Munn, Z., Stern, C., Aromataris, E., Lockwood, C., & Jordan, Z. (2018). What kind of
Clinical Psychology, 61(5), 790–803. 10.1037/0022-006X.61.5.790. systematic review should I conduct? A proposed typology and guidance for systematic
DeBeck, K., Cheng, T., Montaner, J. S., Beyrer, C., Elliott, R., Sherman, S., Wood, E., reviewers in the medical and health sciences. BMC Medical Research Methodology, 18,
& Baral, S. (2017). HIV and the criminalisation of drug use among peo- 5. 10.1186/s12874-017-0468-4.
ple who inject drugs: A systematic review. The Lancet HIV, 4(8), e357–e374. National Association for Public Health Policy. (1999). A public health approach to miti-
10.1016/S2352-3018(17)30073-5. gating the negative consequences of illicit drug abuse. Journal of Public Health Policy,
Delbecq, A. L., Van deVen, A. H., & Gustafson, D. H. (1975). Group techniques for program 20(3), 268–281. 10.2307/3343401.
planning: A guide to Nominal Group and Delphi processes. Glenview, IL: Scott, Foresman National Institute for Health Research. (2020). PROSPERO: International prospective reg-
& Co. ister of systematic reviews. Retrieved from https://www.crd.york.ac.uk/prospero/
Emerson, B., & Haden, M. (2021). A public health based vision for the management Neill, K. A. (2014). Tough on drugs: Law and order dominance and the neglect of
and regulation of opioids. International Journal of Drug Policy, 91, Article 103201. public health in U.S. drug policy. World Medical and Health Policy, 6(4), 375–394.
10.1016/j.drugpo.2021.103201. 10.1002/wmh3.123.
Erickson, P. G. (1990). A public health approach to demand reduction. Journal of Drug Nolan, M. L., Allen, B., & Paone, D. (2020). Commentary on Hoots et al. (2019):
Issues, 20(4), 563–575. 10.1177/002204269002000406. The gap between evidence and policy calls into question the extent of a pub-
European Monitoring Centre for Drugs and Drug Addiction. (2013). Drug policy and law. lic health approach to the opioid overdose epidemic. Addiction, 115(5), 959–960.
Retrieved from http://www.emcdda.europa.eu/policy-and-law doi:10.1111/add.14958.
Gostin, L. O., & Lazzarini, Z. (1997). Prevention of HIV/AIDS among injection drug users: Noyes, J., Booth, A., Cargo, M., Flemming, K., Harden, A., Harris, J., . . . Thomas, J. (2020).
The theory and science of public health and criminal justice approaches to disease Chapter 21: Qualitative evidence. In Higgins et al.. Cochrane handbook for sys-
prevention. Emory Law Journal, 46, 587–696. tematic reviews of interventions version 6.1 updated September 2020Retrieved from
Haden, M., & Emerson, B. (2014). A vision for cannabis regulation: A public health ap- https://training.cochrane.org/cochrane-handbook-systematic-reviews-interventions .
proach based on lessons learned from the regulation of alcohol and tobacco. Open Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mul-
Medicine, 8(2), e73–e80. row, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R.,
Health Canada. (2018). Health Canada releases summary of com- Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W.,
ments from cannabis regulatory consultations. Retrieved from Mayo-Wilson, E., McDonald, S., . . . Moher, D. (2021a). The PRISMA 2020 statement:
https://www.canada.ca/en/health-canada/news/2018/03/health-canada-releases-su An updated guideline for reporting systematic reviews. British Medical Journal, 372,
mmary-of-comments-from-cannabis-regulatory-consultations.html n71. 10.1136/bmj.n71.
Health Canada. (2021). 2020–21 Departmental plan. Retrieved from Parry, C. D. H. (2014). African experience supports view that the global alcohol industry
https://www.canada.ca/en/health-canada/corporate/transparency/corporate-mana should have no role in the formulation of public health policies. Addiction, 109(7),
gement-reporting/report-plans-priorities/2020-2021-report-plans-priorities.html 1211–1212. 10.1111/add.12582.

11
J.-F. Crépault, C. Russell, T.M. Watson et al. International Journal of Drug Policy 112 (2023) 103958

Querido, A. (1954). Alcoholism as a public health problem. Quarterly Journal of Studies on Tomar, S. L., & Giovino, G. A. (1998). Incidence and predictors of smokeless to-
Alcohol, 15(3), 469–476. 10.15288/qjsa.1954.15.469. bacco use among US youth. American Journal of Public Health, 88(1), 20–26.
Rampin, R., & Rampin, V. (2021). Taguette: Open-source qualitative data analysis. Journal 10.2105/AJPH.88.1.20.
of Open Source Software, 6(68), 3522. 10.21105/joss.03522. Tong, A., Flemming, K., McInnes, E., Oliver, S., & Craig, J. (2012). Enhancing transparency
Raninen, J., & Livingston, M. (2020). The theory of collectivity of drinking cul- in reporting the synthesis of qualitative research: ENTREQ. BMC Medical Research
tures: How alcohol became everyone’s problem. Addiction, 115(9), 1773–1776. Methodology, 12, 181. 10.1186/1471-2288-12-181.
10.1111/add.15057. Veritas Health Innovation. (2022). Covidence systematic review software. Retrieved from
Riley, D., Sawka, E., Conley, P., Hewitt, D., Mitic, W., Poulin, C., Room, R., Single, E., & https://www.covidence.org
Topp, J. (1999). Harm reduction: Concepts and practice. A policy discussion paper. Volkow, N. D., Poznyak, V., Saxena, S., & Gerra, G. (2017). Drug use disorders: Impact
Substance Use & Misuse, 34, 9–24. 10.3109/10826089909035632. of a public health rather than a criminal justice approach. World Psychiatry, 16(2),
Room, R. (2004). Alcohol and harm reduction, then and now. Critical Public Health, 14(4), 213–214. 10.1002/wps.20428.
329–344. 10.1080/09581590400027536. Weisheit, R. A., & Klofas, J. M. (1998). The public health approach to illicit drugs. Criminal
Room, R. (2015). Public health, the public good, and drug policy. Nordic Studies on Alcohol Justice Review, 23, 197–207. 10.1177/073401689802300204.
and Drugs, 32(4), 367–370. 10.1515/nsad-2015-0036. World Health Organization. (2022a). Alcohol. Retrieved from
Sobell, M. B., & Sobell, L. C. (1995). Controlled drinking after 25 years: https://www.who.int/health-topics/alcohol
How important was the great debate? Addiction, 90(9), 1149–1153. World Health Organization. (2022b). Tobacco. Retrieved from
10.1111/j.1360-0443.1995.tb01077.x. https://www.who.int/health-topics/tobacco
Stenius, K. (2018). Public health is not an innocent concept. Nordisk Alkohol & World Health Organization. (2022c). Drugs (psychoactive). Retrieved from
Narkotikatidskrift, 35(2), 100. 10.1177/1455072518765838. https://www.who.int/health-topics/drugs-psychoactive
Stimson, G. V. (2007). Harm reduction – Coming of age”: A local move- Wyton, M. (2022). Experts reject BC’s safe supply claims. The Tyee Retrieved from
ment with global impact. International Journal of Drug Policy, 18(2), 67–69. https://thetyee.ca/news/2022/03/03/experts-reject-bc-safe-supply-claims/ .
10.1016/j.drugpo.2006.12.012. Yeo, Y., Johnson, R., & Heng, C. (2021). The public health approach to the worsening
Sweanor, D., Alcabes, P., & Drucker, E. (2007). Tobacco harm reduction: How rational opioid crisis in the United States calls for harm reduction strategies to mitigate the
public policy could transform a pandemic. International Journal of Drug Policy, 18(2), harm from opioid addiction and overdose deaths. Military Medicine, 187(9-10), 244–
70–74. 10.1016/j.drugpo.2006.11.013. 247. 10.1093/milmed/usab485.
Thomas, J., & Harden, A. (2008). Methods for the thematic synthesis of qualita-
tive research in systematic reviews. BMC Medical Research Methodology, 8, 45.
10.1186/1471-2288-8-45.

12