Introduction

Over the past decade, cannabis use prevalence has increased substantially across the globe [1,2,3]. The United Nation’s World Drug Report, 2020 estimates that in 2018, 192 million people used cannabis in the past year, equating to a global prevalence of roughly 3.9% [4]. North America, Australia and New Zealand, and West and Central Africa have substantially higher cannabis use prevalence, at 14.6%, 10.6%, and 9.3%, respectively [4]. Upticks in cannabis use prevalence are also seen among women of reproductive age, including pregnant and postpartum women [5,6,7]. Large increases in North America, including the United States and Canada have been seen the past decade [2, 8, 9]. In the United States, estimates of past-month cannabis use among non-pregnant women have increased from 11.0% to 2016 to 14.7% in 2019 [10]. Canada has seen a similar trend, with prevalence of cannabis use in women nearly doubling from 6.6 to 11.1% from 2004 to 2017 [11]. Yet, evidence on the etiology of cannabis use, including reasons for and influences of cannabis use among women remains largely unknown [12,13,14].

There is a growing body of evidence exploring women’s cannabis-related knowledge, attitudes, and perceptions of cannabis use [12, 13, 15,16,17,18]. Assessing antecedents of cannabis use among women of reproductive age throughout critical life stages (e.g., adolescence, preconception, prenatal, postpartum) is imperative for the development of tailored and effective cannabis use prevention efforts. However, a robust, in-depth assessment, including psychometric evaluation, of existing measures of these potential reasons for cannabis use in women of reproductive age has not yet been performed. Such a systematic mapping of available measures on antecedents of cannabis use would undoubtedly aid researchers and clinicians in identifying the best measure for their respective purpose and population. This evidence gap, in combination with increasing prevalence of cannabis use among women of reproductive age [2, 8, 9], supports an urgent need to examine the depth and breadth of existing instruments to measure cannabis-related knowledge, attitudes, perceptions, motivations, and influences among women of reproductive age. Further, this gap may also hinder the strength of epidemiologic studies examining women’s cannabis use.

Thus, we aimed to systematically map existing evidence on measures of cannabis-related knowledge, attitudes, perceptions, motivations, and influences among women of reproductive age, including pregnant and postpartum women. This scoping review will also serve as a necessary precursor to determine if a systematic review on this topic should be performed [19].

Methods

This scoping review is directly aligned with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist [20].

Protocol and registration

We utilized the scoping review framework by Arksey and O’Malley (2005), as well as recent guidance to increase rigor and reporting of scoping reviews [19,20,21]. The a priori protocol for this review was drafted using the PRISMA extension for Scoping Reviews [20]. Due to the rapid nature of this review, the protocol for this review was not published, but can be accessed by contacting the authors.

Eligibility criteria

To be included in the review, studies needed to examine or report on the development, utilization, or limitations of, measures of cannabis-related knowledge, attitudes, perceptions, motivations, and influences among women of reproductive age. Further, studies were eligible if they focused on women of reproductive age, including, but not limited to women during the preconception (12 months prior to pregnancy), prenatal (during pregnancy), and postpartum (the 12 months after birth) periods. Peer-reviewed studies written in English from any geographical location were included if they were published between 2010 and 2021. Quantitative studies were eligible; mixed-methods studies that included quantitative studies were eligible, but we extracted only quantitative information to be included in the analysis. We also included systematic reviews, with or without meta-analysis, and reviews of the literature if they included quantitative studies. We excluded studies where the population was not women of reproductive age (e.g., biological men, older adults, mixed gender populations) as well as studies that were published before 2010, published as conference abstracts or book chapters, and published in a language other than English. We also excluded studies that did not measure cannabis-related knowledge, attitudes, perceptions, motivations, or influences, as well as studies assessing and reporting on self-reported cannabis use prevalence only. Finally, we excluded reviews that included only qualitative studies.

Information sources

To identify potentially relevant studies, we searched the following databases from January 2010 to March 2021: PubMed, PyschINFO, CINAHL, and Google Scholar. We also included the first 200 results from Google Scholar, when sorted via relevance. We limited our search from 2010 onward due to the changing nature of cannabis, including legalization, so that we captured only contemporary measures in this review. We developed the final search strategy using terms for instruments that have been previously used in systematic reviews (e.g., “questionnaire”, “instrument”, “tool”) [22], incorporating additional terms specific for our population (e.g., “women”, “prenatal”, “pregnant”, “perinatal”, “postpartum”, “breast feeding”) and topic of interest (e.g., “cannabis”, “marijuana”) [23, 24]. We piloted our search strategy for each database to ensure effectiveness in producing relevant articles. After piloting search strategies in each database, we adapted the initial search terms to exclude terms that failed to yield relevant results, which included the following terms: “survey”, “evaluation”, “assessment”, “weed”, and “CBD”. The final search strategy utilized for this scoping review is presented in Additional file 1.

Selection of sources of evidence

We used Covidence Software, an online systematic review management tool, to streamline and manage the review process (Covidence Systematic Review Software, Veritas Health Innovation, Melbourne, Australia). As part of the import process, Covidence automatically de-duplicated citations based on a match of the citation author, title, and date. After search results were imported into Covidence, the review team performed a two-stage review process (title and abstract screening and full-text screening) to screen and identify references eligible for inclusion. Two study team members (KS and ED) piloted the screening process in Covidence on 20 citations, during which we examined both the screening process and reviewer agreement. Then, two members of the research team performed title and abstract screening independently in duplicate. Upon completion of title and abstract screening, we screened potentially relevant studies in their full-text form. We resolved disagreement between reviewers at any stage using consensus and discussion. Studies meeting all inclusion criteria moved forward for data extraction. For each included study, we carried out forward and backward citation searches to identify any potential articles not included via database searching.

Data extraction

The research team developed a detailed data extraction form, which was piloted within Covidence. Two reviewers independently extracted study data and achieved consensus on each item. Using recommendations on relevant data fields for scoping reviews, we extracted the following data: (1) study information (e.g., author, geographic location, dates, purpose, funding); (2) population and context (e.g., study population, setting, method of recruitment); (3) measure/tool/instrument-related data (e.g., tools, measures, psychometric properties); (4) results of pilot or feasibility testing of the measure/tool/instrument; (5) limitations of the study; (6) recommendations for future research; and (7) study conclusions. Given the overall purpose of this scoping review, we did not perform quality assessment on included studies.

Synthesis of results

Based on expected variability in how measure-related information is presented in included studies, we analyzed data both narratively and quantitatively, reporting summary of findings tables that map results in a meaningful manner. For tabular presentation of results, we first stratified by country in which the study was conducted, noting the overall sample, setting, aim, results, and conclusions of each included study. We initially intended to further stratify results by the type of measure used (e.g., knowledge, attitude, perception, motivation, influence). However, as many included studies tapped into multiple domains, this was not feasible. Additionally, we aimed to present, via tabular form and narrative synthesis, findings based on psychometric testing, differentiating between those measures for which validity and reliability have been established versus those measures that did not undergo psychometric testing. Due to the lack of psychometric testing of included measures, this was not possible. We synthesized survey characteristics in both tabular and narrative form, summarizing existing measures based on specific period(s), if any, that the measure was given (e.g., all women of reproductive age, pregnant women, breastfeeding women). We synthesized recommendations for future results and present them in tabular form.

Results

Out of 927 unique citations screened, 11 studies were eligible for inclusion in this review. Figure 1 details the systematic study selection process in accordance with PRISMA guidelines.

Fig. 1
figure 1

PRISMA flow diagram

Full size image

Key features of included studies

Table 1 describes key features of included studies, including study sample, setting, aim, results, and conclusions. We identified one systematic review [25] and 10 original studies [12,13,14, 26,27,28,29,30,31,32]. Of included studies, 3 were conducted in Canada [13, 26, 27] and 8 were conducted in the United States [14, 25, 28,29,30,31,32]. About 81% of studies (n = 9) were published in the past 4 years [12, 13, 25,26,27,28, 30,31,32]. Studies included samples of women in various life stages; two studies included non-pregnant women only [29, 32] and one included women experiencing infertility only [27]. A majority of studies included women in the perinatal period (n = 8), with some studies specifically focusing on pregnant women (n = 4) [12, 13, 28, 30] or postpartum women (n = 3) [14, 26, 31]. Most studies (n = 8) recruited women from clinics or hospital settings [12,13,14, 26, 28,29,30, 32], with 2 studies performing secondary data analyses reporting data from national surveillance systems in the United States [30, 31].

Table 1 Characteristics of included studies (N = 11)
Full size table

Measure characteristics

Characteristics of measures, including measurement domain(s), recruitment methods, population, administration modality, and a brief description of the instrument are presented in Table 2. Measurement domains among included studies were perceptions (n = 8) [13, 14, 25, 27,28,29,30,31,32], knowledge (n = 3) [12, 25, 26], attitudes (n = 2) [26, 28], intentions [12], and motivations [12]. More specifically, studies aimed to measure how information received from health care providers influenced cannabis-related decision making [13], attitudes and experiences about using cannabis during childbirth or labor [26], perceptions of cannabis use on infertility [27], risk perceptions of cannabis use [14, 28, 30], negative expectancies associated with cannabis use [29], perceptions of cannabis use and the sexual experience [32], views on cannabis legalization [12, 28], potential influence of legalization on cannabis use [12], knowledge of potential harms [12, 28], and motivations for cannabis cessation [12]. Of studies examining risk perceptions of cannabis use, 4 examined perceptions associated with prenatal cannabis use [12, 14, 28, 30], and one examined risk perceptions among postpartum, breastfeeding women [31].

Table 2 Measure characteristics among included studies (N = 11)
Full size table

Psychometric evaluation

Most studies created their own survey on women’s cannabis-related knowledge, attitudes, perceptions, and motivations (n = 7) [12,13,14, 27, 28, 30, 32] and did not report evaluating psychometric properties of included measures. However, only one study mentioned piloting the survey for validation purposes [27]. Ng et al. (2020) mention reviewing their survey for readability statistics, but do not mention other methods of psychometric evaluation [28]. Two studies used measures from United States surveillance systems [30, 31] and one study utilized the Marijuana Effect Expectancies Questionnaire (MEEQ), a previously validated instrument [29].

Research and practice recommendations

Included studies had numerous recommendations for both future research and practice (Table 3). Surprisingly, only one study mentioned psychometric evaluation of measures for future research [28]. Common areas of future research to address existing knowledge gaps included future studies with a more robust design (e.g., controlling for co-substance use, homogenous populations) and studies examining the etiology of cannabis use among women, including how women’s attitudes, beliefs, and perceptions play a role in the cannabis-related decision making [12,13,14, 29, 30]. Another recommendation area for future research was examining health care providers’ motives for and influences of cannabis-related recommendations [30, 31].

Most practice recommendations centered on the role of health care providers in preventing potential adverse health outcomes. Specifically, studies recommended that health care providers counsel women about risks of cannabis use to the mother and fetus during pregnancy and postpartum [27, 31, 33, 34]. A single study highlighted that cannabis as a labor analgesic should not be recommended, given absence of safety data [35]. Two studies reiterated the need for public health campaigns that reflected contemporary evidence of risks of prenatal cannabis use [33, 36].

Table 3 Research and practice recommendations of included studies
Full size table

Discussion

This is the first review, to our knowledge, to comprehensively examine the breadth of research on measures of antecedents of cannabis use among women of reproductive age. We identified 11 studies reporting on measures of cannabis-related knowledge, attitudes, perceptions, and motivations. We found risk perceptions among pregnant women was the most frequent construct assessed and that most studies were conducted with English-speaking women from hospital or clinic settings. A single study measured the role that health care providers play in women’s cannabis-related decision making. Surprisingly, there were no studies measuring social influences of cannabis use in women. Overall, there was a paucity of evidence, with little to no discussion of psychometric properties of these measures. Thus, we have identified several measurement gaps in this field which future research should aim to address.

In this review, we found a lack of valid, reliable measures to assess antecedents of cannabis use in women of reproductive age, including important maternal health periods, such as the preconception, prenatal, and postpartum periods. With increasing surveillance and research being conducted on women’s cannabis use, the importance of using psychometrically sound measures cannot be understated. Many measures to assess cannabis-related knowledge, perceptions, and motivations in broader, heterogenous populations exist [37,38,39,40]. Undoubtedly, future research could look to validate and test for reliability these existing instruments in subpopulations of women. Future research should also prioritize addressing existing measurement-related gaps of cannabis use among women via the creation of psychometrically sound measures to assess antecedents of cannabis use throughout the life span (e.g., adolescence, young adult, preconception, prenatal, postpartum, parenthood), as these may drastically change over time. Importantly, as most prior research was conducted with English-speaking women in health care settings, future research should look how health disparities and health inequities contribute to prenatal cannabis use. As a start, researchers could aim to examine the psychometric properties of instruments or measures included in this review, which would provide a solid foundation from which future research could build.

The lack of available research on measures of antecedents in women of reproductive age poses a challenge to current and future epidemiologic studies that aim to assess cannabis use. Validated and reliable measures of substance use are critical in the success of longitudinal substance use studies, such as the Adolescent Brain Cognitive Development (ABCD) cohort study [41]. Additionally, the lack of evidence on psychometric properties of existing measures is worrisome, as sound psychometric properties are a necessary prerequisite for utilization of any measure [42]. Several included studies used data from nationally-based surveillance systems in the US. However, many of these measures used have yet to be examined for reliability and validity—yet another important area that future research should examine. There has been much qualitative work conducted in this area [15,16,17, 43]; now researchers should transition to the development and evaluation of quantitative measures. Only after psychometrically sound measures are developed can future work aiming to address associations between antecedents of cannabis use and uptake and continuation of cannabis use begin.

An aim of this scoping review was to elucidate the need for a systematic review on the measures of antecedents of cannabis use among women of reproductive age. Although this is an expanding field, it appears that there is not yet enough empirical evidence to undertake a systematic review. However, researchers could look to conduct a systematic review in this area after this research area has had time to develop and expand. As this is a rapidly growing area of research, we recommend that another scoping review be conducted in 1-2 years and the need for a systematic review be re-evaluated.

Limitations

There are some limitations of this scoping review. First, we excluded gray literature and studies not published in English, which in turn, could have resulted in failure to identify potentially relevant studies. Second, we utilized date restrictions to capture measures with contemporary relevancy. In doing so, we may have missed in-press or recently published articles yet to be indexed or older articles that may be relevant. Lastly, we attempted to extract psychometric information to include in tabular form in this scoping review but given the lack of psychometric assessment and reporting among included studies, we were unable to do so.

Conclusions

Amid rapidly changing societal norms and policies regarding cannabis use, those aiming to examine and understand women’s attitudes, perceptions, motivations, and influences of cannabis use uptake and use patterns need measures that are valid, reliable, and easy to use. In this scoping review, however, we found a paucity of evidence in this area, with existing measures limited by breadth, depth, and psychometric soundness, posing a measurement challenge. Ideally, psychometrically sound measures of key constructs should be developed prior to the start of cannabis prevention efforts. Thus, the overarching conclusion of this scoping review is that measurement of women’s cannabis-related knowledge, attitudes, perceptions, motivations, and influences should be a focus of this emerging research agenda.