We used the Joanna Briggs Institute (JBI) methodological guidance for scoping reviews to inform our scoping review methods [5]. The Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) [6] was used to guide the reporting of this rapid scoping review, and the checklist is provided in Additional file 1. At the time of conducting this review, a PRISMA extension to rapid reviews was under development.
Protocol and registration
The protocol for this rapid scoping review was developed a priori and registered on the Open Science Framework [7] on August 21, 2018 [8].
Eligibility criteria
The eligibility criteria were developed using the PECOS components, as follows:
P - Population
We included individuals of all ages, across all geographic areas to explore GD and IGD in all populations.
E - Exposures
We included populations that had ICD–11-defined GD [4] or DSM–5-defined IGD [1], according to our objectives. All types of gaming behaviours (e.g., online/offline, mobile/console/computer, single/multi-player) were relevant for our review. For an IGD diagnosis, five or more of the nine criteria needed to be met [2]. As this was a scoping review, we included all studies that applied this definition, so that we could chart all of the various methods and diagnostics tests that have been reported in the literature.
We excluded studies with populations that were defined using adapted criteria from conditions in older versions of the DSM (e.g., DSM-IV) as these did not fit the specific definitions of interest for our review. We also excluded studies with populations who only had gambling conditions or disorders related to gambling games (e.g., casino or poker games) as gambling-related conditions were considered separate from GD or IGD, and outside of the scope of this review.
C - Comparators
In order to be inclusive of any study that reported prevalence, we included studies with any type of comparator or studies without a comparator.
O - Outcomes
We included studies that reported the prevalence of GD or IGD, according to our objectives. Prevalence is defined as a measure of the occurrence of the health condition or exposure of interest; it is the total number of individuals who have the condition (cases) at a particular time (or during a particular period) divided by the number of persons in the population at a specified time [9]. To capture the various methods of reporting GD or IGD in the literature, studies that reported GD or IGD scores on diagnostic scales were also included.
We included studies that reported the prevalence or scores of severe cases of GD or IGD. Severe GD or IGD was defined as cases where an entire population with GD or IGD was undergoing treatment for GD or IGD.
S - Study designs
We included primary quantitative studies such as randomized controlled trials (RCTs), controlled before-after studies, uncontrolled before-after studies, controlled after studies, quasi-experimental studies, observational studies, case-control studies, and cross-sectional studies. Studies that were categorized as observational [9] were non-experimental studies where there was not enough information from the article to determine the specific study design.
We included mixed methods studies if they were qualitative studies that reported quantitative data on our outcomes of interest, otherwise qualitative studies were excluded. We also excluded books, case studies, case series, and reviews in our rapid review; however, a list of potentially relevant reviews is provided in Appendix A (Additional file 2).
Year of Publication, Publication Status, and Language
We included full-text publications from all years of publication. Grey literature (i.e., difficult to locate or unpublished material) in the form of theses/dissertations and conference abstracts were included if a full-text article was available. We limited the included studies to those in the English language, due to the short timeline to conduct the review; however, a list of potentially relevant non-English studies is provided in Appendix B (Additional file 2).
Literature search strategy
The literature search strategy was drafted by an experienced librarian (BS) with input from the research team, and peer-reviewed by a second experienced librarian (HM) using the Peer Review of Electronic Search Strategies (PRESS) checklist [10]. An experienced library technician (AE) searched MEDLINE, EMBASE, PsycINFO, and the Cochrane Library on July 9, 2018, exported the literature search results into EndNote, and discarded duplicates. To ensure saturation, we assessed studies that were identified by content experts. The final literature search strategy for MEDLINE, EMBASE and PsycINFO is found in Appendix C (Additional file 2).
Selection of sources of evidence
A standardized screening form to determine eligibility of studies for title and abstract screening was developed, and pilot-tested independently by team members on the same 25 citations in Synthesi.SR, proprietary software developed by the Knowledge Translation Program of St. Michael’s Hospital [11]. After one pilot, ≥ 80% inter-rater agreement was reached, whereby the include/exclude status of at least 80% of the studies was unanimously agreed upon by the entire team. The standardized form was modified as needed, and the remaining titles and abstracts were screened by one reviewer (VN, JPS, MG, AD, SL, NT). A second reviewer verified excluded citations (ND, CCL, VN, JPS, MG, RC).
A standardized screening form to determine eligibility of full-text articles was developed, and pilot tested independently by team members on the same 25 citations in Synthesi.SR [11]. After two pilots of 25 studies each (50 in total), ≥ 80% inter-rater agreement was reached, whereby the include/exclude status of at least 80% of the studies was unanimously agreed upon by the entire team. The standardized form was modified as needed, and each remaining full-text article was screened by one reviewer (ND, AR, RC, VN, JPS, MG). A second reviewer verified excluded articles (ND, AR, VN, JPS, MG).
Data charting process and data items
A standardized charting form was pilot-tested independently by team members on a sample of five studies in Excel. After two pilots of five studies each (10 in total) sufficient agreement was reached. The standardized form was modified, as required, in an iterative process [6]. Each study was charted by one reviewer (VN, JPS, MG, RB) and verified by a second (ND, AR). Due to the time limitations of this rapid scoping review, authors of included studies were not consulted to confirm data. We extracted the following data for general, clinical, and severe populations: country, study design, population description, sample size, age, gender/sex proportions, gaming time outcomes, GD or IGD reporting method (self-report, health professional, etc.), GD or IGD measure/instrument/assessment, GD or IGD prevalence numerator/denominator (number of people with GD or IGD, divided by total population at risk), GD or IGD prevalence estimate and 95% confidence interval (CI), GD or IGD mean score estimate and standard deviation (SD), and the list of variables (e.g., mental health outcomes) reported.
We charted data for three groups:
- General: Populations that were not seeking treatment for GD or IGD, did not have GD or IGD at recruitment, and were not undergoing treatment for GD or IGD
- Clinical: Populations who sought treatment, and populations with GD or IGD at recruitment
- Severe: Populations undergoing treatment for GD or IGD. Severity could have been reported by the authors of included studies in the following four ways: i) authors distinguished between GD/IGD and severe GD/IGD using a test/scale to categorize participants into a separate severe group, ii) a portion of the population was undergoing treatment for GD or IGD, iii) the entire population with GD or IGD was undergoing treatment for GD or IGD, and iv) a portion of the population sought treatment for GD or IGD.
For objectives 1 and 2, we charted prevalence results that were defined by having met at least five IGD criteria. In one case where studies separated outcomes based on whether or not participants answered “somewhat agree” versus “strongly agree” on a polytomous scale [12], we charted the outcomes where a response of “strongly agree” was indicative of a criterion being endorsed. For objective 3, we charted all reported variables only in the included studies that reported prevalence or scores for GD or IGD. We considered variables to be all demographic data that study authors reported (e.g., education, income) and all outcomes of interest to study authors (e.g., depression, brain imaging characteristics).
For studies where multiple prevalence or score results were presented, we charted the results from the main analysis (or from a secondary analysis if the main analysis results were not reported as prevalence or scores). Since most studies reported only on gender or only on sex and did not clarify which of these two were measured and how, we charted gender or sex as reported in the article and included it as one data item. We reported the proportions of gender/sex as presented by authors; if authors did not report data for a particular gender/sex category, we did not chart any data for that category. Gender/sex proportions were reported as percentages rounded to the nearest whole number. In cases where authors of included studies only provided raw data for prevalence or gender/sex proportions, we calculated these so that they would be reported consistently across studies.
Risk of bias appraisal of individual sources of evidence
We did not appraise risk of bias of individual sources of evidence, which is consistent with established scoping review methods [5, 6].
Quantitative analyses
We did not conduct summary quantitative analyses (e.g., meta-analyses) of prevalence data, which is consistent with established scoping review methods [5, 6].
Synthesis of results
To synthesize the rapid scoping review findings, we provide the prevalence or score estimates for GD, IGD, severe GD, and severe IGD for each study in tabular format and summarize the ranges of prevalence estimates in the text. We do not provide ranges for mean score data (between test comparison), as the tests differed widely and their numerical values could not be compared and summarized together. We also do not provide the ranges of prevalence or mean scores by each of the measures (within test comparison), as most of the tests had results from only one included study, the populations in these included studies were different, and the results may have been on different scales. In addition, we provide a summary of the frequencies of variables reported in included studies. Where appropriate, we summarize findings by the following subgroups: WHO geographic region [13], gender/sex categories, and age groups. Age groups were defined as follows: children (0–19 years old), adolescents (10–19 years old), and adults (18 years and older) [14]. All adolescent-specific data were also included in the children group. WHO geographic regions [13] included the following: African Region, Eastern Mediterranean Region, European Region, Region of the Americas, South-East Asia Region, and Western Pacific Region. In addition, based on the included studies, we created a map using the Mapchart tool [15] to show the number of articles found and prevalence ranges by WHO regions.
Deviations from study protocol
Due to time and resource constraints of this rapid scoping review (the WHO requested results within five months), we were unable to scan the reference lists of included studies and relevant systematic reviews or comprehensively chart quantitative comorbidity results as part of our third objective. Instead, we charted all variables that were reported by study authors in included studies, to describe the variables that researchers in the field of IGD and GD measure when studying these populations.