Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Nov 1.
Published in final edited form as: Clin Psychol Rev. 2023 Sep 9;105:102336. doi: 10.1016/j.cpr.2023.102336

Cognitive-behavioral treatment for gambling harm: Umbrella review and meta-analysis

Rory A Pfund a,b,*, Meredith K Ginley a,c, Hyoun S Kim d, Cassandra L Boness e, Tori L Horn a,b, James P Whelan a,b
PMCID: PMC11059187  NIHMSID: NIHMS1986731  PMID: 37717456

Abstract

The aim of the current umbrella review and meta-analysis was to evaluate the methodological rigor of existing meta-analyses on cognitive-behavioral treatment (CBT) for gambling harm. The Cochrane Database of Systematic Reviews, PsycINFO, and PubMed were searched for meta-analyses of CBT for gambling harm among individuals aged 18 years and older. The search yielded five meta-analyses that met inclusion criteria, representing 56 unique studies and 5389 participants. The methodological rigor for one meta-analyses was rated high, two were moderate, and two were critically low. Including only moderate- to high-quality meta-analyses, a robust variance estimation meta-analysis indicated that CBT significantly reduced gambling disorder severity (g = −0.91), gambling frequency (g = −0.52), and gambling intensity (g = −0.32) relative to minimal and no treatment control at posttreatment, suggesting 65%–82% of participants receiving CBT will show greater reductions in these outcomes than minimal or no treatment controls. Overall, there is strong evidence for CBT in reducing gambling harm and gambling behavior, and this evidence provides individuals, clinicians, managed care companies, and policymakers with clear recommendations about treatment selection.

Keywords: CBT, Evidence-based practice, Gambling disorder, Problem gambling

1. Introduction

Gambling, defined as placing a bet on an event where the outcome is at least partially due to chance, has increasingly become a socially acceptable form of entertainment across the world. According to Statista (2023), the United Kingdom leads all online gambling markets in gross gaming revenue with 12.5 billion United States dollars, followed by the United States ($11.0 billion) and Australia ($6.5 billion). The social acceptability of gambling will only continue to proliferate as there continue to be more opportunities to gamble legally throughout the world. A recent example is the United States, where a Supreme Court ruling allowed states to decide the legality of sports gambling. Since that court ruling, sports gambling has been legalized in 33 states and Washington D.C. as of June 2023 (American Gaming Association, 2023).

Between 0.1% and 5.8% of adults worldwide are estimated to experience gambling harm (Calado & Griffiths, 2016), corresponding to a major public health concern involving adverse consequences associated with decrements in well-being (Langham et al., 2015). Such adverse consequences include but are not limited to, lost productivity, mental and physical health difficulties, interpersonal problems, suicide risk, and financial concerns (American Psychiatric Association, 2013; Hofmarcher, Romild, Spångberg, Persson, & Håkansson, 2020; Langham et al., 2015; Latvala, Lintonen, & Konu, 2019). The estimated social cost of these consequences is approximately $9393 USD per year for each individual experiencing gambling harm (Grinols, 2011). A more recent study conducted in Sweden estimated that gambling harm resulted in a societal cost of 1.42 billion Euros in 2018 alone (Hofmarcher et al., 2020). It has also been estimated that an individual’s gambling problems negatively affects six other people, especially family members (Goodwin, Browne, Rockloff, & Rose, 2017).

Within an evidence-based practice framework that integrates the best available research evidence, clinical experience, and patients’ characteristics, culture, and preferences (American Psychological Association, 2006), cognitive-behavioral treatment (CBT) is the treatment with the most empirical evidence for reducing gambling harm and gambling behavior (Petry, Ginley, & Rash, 2017). CBT is grounded in social cognitive theory that suggests individuals are active agents of change within a dynamic and complex social environment, and individuals can influence and be influenced by their environments (Bandura, 2001). From this theory, individuals participating in CBT might identify high-risk situations that precipitate gambling, expose themselves to gambling cues, engage in other values-consistent activities and learn new coping strategies, set limits on the amount of time spent gambling or the amount of money gambled, create new ways of thinking, and/or plan for future high-risk situations to reduce gambling behavior and harms (Grant et al., 2009; Pfund & Ginley, 2019). Subsequently, individuals who participate in these CBT activities increase their self-efficacy, and these increases in self-efficacy are associated with reductions in gambling disorder severity (Winfree, Ginley, Whelan, & Meyers, 2015).

Despite the support of multiple systematic reviews and meta-analyses, the research on CBT for gambling harm is fragmented. Extant reviews have examined specific CBT variations (e.g., exposure therapies, Bergeron, Giroux, Chrétien, & Bouchard, 2022) and formats (e.g., in person, Cowlishaw et al., 2012). Other reviews have examined the effect of CBT on specific outcomes. Most recently, a meta-analysis reviewed the effect of CBT on gambling outcomes across 29 published studies and indicated that CBT significantly reduced the number of days gambled, the amount of money gambled, and gambling disorder severity at posttreatment relative to minimally treated (e.g., brief advice) and inactive controls (e.g., waitlist) (Pfund et al., 2023). A follow-up meta-analysis reviewed a subset of those same 29 studies that reported the effect of CBT for gambling harm on common co-occurring problems and found that CBT also significantly reduced anxiety and depression outcomes, but not substance use outcomes, at posttreatment relative to minimally treated and inactive controls (Pfund et al., 2023). The fragmented research in this area requires a synthesis of all knowledge to assist individuals seeking treatment, clinicians, managed care companies, and policymakers in selecting empirically supported treatment options for gambling harm.

Professional organizations have used knowledge syntheses, such as umbrella reviews, to facilitate the selection of empirically supported treatment options. For example, the American Psychological Association Society of Clinical Psychology’s (SCP) uses umbrella reviews to maintain a list of empirically supported treatments for various psychological disorders (https://div12.org/psychological-treatments/). However, as of August 2023, the SCP did not recognize the diagnosis of gambling disorder on their website and did not provide any recommendations about empirically supported treatments for gambling harm. Professional organizations that specifically focused on resources for gambling harm also provided little guidance on the selection of empirically supported treatments. As of August 2023, the National Council on Problem Gambling provided information about where to find treatment providers but did not provide recommendations on specific treatment options for gambling harm.

The purpose of the present study was to conduct an umbrella review based on procedures from the American Psychological Association’s SCP (Boness et al., 2021). Specifically, this review synthesized the extant meta-analyses on CBT for gambling harm and evaluated their methodological rigor. This umbrella review would provide at least one professional organization with a definitive recommendation on CBT for gambling harm that could be disseminated broadly to multiple stakeholders.

2. Method

The method of the present study was adapted from procedures from the SCP manual for the evaluation of empirically supported treatments according to the Tolin criteria (Boness et al., 2021) and an application of the Tolin criteria for contingency management for drug use disorder (Pfund et al., 2023). Specifically, this study included a systematic article search for meta-analyses of CBT for gambling harm, an evaluation of the methodological rigor of the included meta-analyses, a meta-analysis of CBT for gambling harm, and a recommendation for CBT as an empirically supported treatment. Methods were also consistent with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (Page et al., 2021) guidelines as well as the reporting standards of the American Psychological Association (Appelbaum et al., 2018). This umbrella review was not pre-registered and did not include a protocol.

2.1. Search strategy

The Cochrane Database of Systematic Reviews, PsycINFO, and PubMed were searched to identify meta-analyses of CBT among individuals experiencing gambling harm in September 2021. A search using the same search terms and databases was also conducted in May 2023 to identify meta-analyses that may have been published since September 2021. A comprehensive list of terms that were used to search each database can be found in Supplemental Table 1.

Meta-analyses were included if they were: (1) a quantitative review, (2) that examined the effect of CBT, (3) among individuals experiencing gambling harms, and (4) who were aged 18 years or older. All CBT formats (i.e., in-person, Internet, self-help workbook) were eligible for inclusion. Meta-analyses were excluded if the overall effect of CBT versus control was blended with overall effect of another treatment (e.g., motivational interviewing, personalized feedback interventions) versus control. There were no restrictions on the type of CBT (e.g., mindfulness-based), types of control groups (i.e., active treatment, inactive treatment), publication year, language, or type of publication (e.g., conference presentations, dissertations, grey literature).

2.2. Meta-analysis selection

Fig. 1 displays a flowchart of the study selection process from the searches conducted in September 2021 and May 2023. After the removal of duplicates, the literature search yielded a total of 38 articles. Using a codebook, two authors (RAP and MKG) independently coded these articles at the title and abstract level, as well as 19 articles at the full-text level. All coding discrepancies were resolved via discussion. The two authors agreed upon the inclusion of five meta-analyses. A list of articles with the primary reason for exclusion can be found in Supplemental Table 2.

Fig. 1.

Fig. 1.

Open in a new tab

Flowchart for the search process. This figure illustrates the search process for locating reviews eligible for inclusion in the evaluation of CBT for gambling disorder. CBT = cognitive-behavioral treatment.

2.3. Data extraction and coding

For all five included meta-analyses, two authors (RAP and MKG) independently coded the population, intervention, comparison, outcomes, timeline, and setting (PICOTS) (Schardt, Adams, Owens, Keitz, & Fontelo, 2007). They also extracted effect sizes for all available outcomes, including gambling disorder severity (i.e., number of diagnostic symptoms or gambling problems), gambling frequency (i.e., number of days gambled), gambling intensity (i.e., amount of money gambled), gambling duration (i.e., amount of time gambled), gambling beliefs, gambling craving, anxiety, depression, substance use, and quality of life. Effect sizes were extracted at posttreatment (i.e., immediately following the conclusion of CBT), as well as follow-up (i.e., in the weeks and months after the termination of CBT). Effect size values reported in a form other than Hedges’s g, like Cohen’s d (Cowlishaw et al., 2012), were converted to Hedges’s g using the Comprehensive Meta-Analysis software.

Numerical values of study-level effect sizes and their lower and upper limit of the 95% confidence interval (CI) were prioritized for extraction. These values were extracted from the Tables and Figures (e. g., forest plots) presented in meta-analyses. However, if numerical values were not available, the WebPlotDigitizer (https://automeris.io/WebPlotDigitizer/) was used to approximate numerical values. Studies support the interrater reliability and validity of the data extracted using this tool (Drevon, Fursa, & Malcolm, 2017).

2.4. AMSTAR-2 ratings

Using the Assessing the Methodological Quality of Systematic Reviews 2 (AMSTAR-2; Shea et al., 2017) system, two authors (CLB and HSK) independently rated the methodological quality of all five included meta-analyses and then provided an overall confidence rating for the quality of each meta-analysis. This system critically evaluates systematic reviews and meta-analyses across 16 methodological domains, and each domain is coded “Yes,” “Partial Yes,” or “No.” Several domains were deemed critical based on the procedures outlined in the SCP manual (Boness et al., 2021), including (a) adequate descriptions of the PICOTS in the research questions and inclusion criteria, (b) use of a comprehensive search strategy, (c) adequate descriptions of the PICOTS for each study included in the meta-analysis, (d) use of appropriate methods to conduct meta-analysis, (e) consideration of the risk of bias in individual studies, and (f) explanation for any heterogeneity observed in the results. A code of “No” indicated on a critical domain was deemed a critical weakness, and a code of “No” on the other domains was deemed a noncritical weakness.

Based on the ratings across the 16 methodological domains, an overall quality rating was then assigned to each meta-analysis. A rating of “High” was assigned if the meta-analysis had zero or one noncritical weakness, “Moderate” if the meta-analysis had no critical weaknesses and more than one noncritical weakness, “Low” if the meta-analysis had one critical weakness with or without noncritical weaknesses, or “Critically Low” if the meta-analysis had more than one critical weakness with or without noncritical weaknesses. Codes were made using a codebook and any discrepancies were resolved via discussion.

2.5. Data analysis plan

Considering only the moderate- and high-quality meta-analyses as determined by AMSTAR-2 ratings, a robust variance estimation meta-analysis was conducted using the “robumeta” and “clubSandwich” packages in R version 4.2.1 (Pfund, Ginley, et al., 2023). Robust variance estimation where ρ was assumed 0.8 was used because some studies included multiple treatment groups and multiple effect sizes within those treatment groups (Hedges, Tipton, & Johnson, 2010). The robust variance estimation meta-analysis used meta-regression models to estimate the weighted Hedges’s g effect size for each outcome. A random effects model was used because there was expected heterogeneity in recruitment methods, formats of CBT (e.g., in-person versus remote), and study methods. Heterogeneity was assessed using τ2, I2, and 95% prediction intervals (PI). Results were only presented for models with adequate degrees of freedom after accounting for small sample size adjustments (Tipton, 2015).

To aid in the interpretation of effect sizes from the present umbrella review, several guidelines were considered. Hedges’s g values of 0.20 were considered small changes in outcomes, 0.50 medium changes, and 0.80 large changes (Cohen, 1988). Negative Hedges’s g values indicated reductions in outcomes relative to controls, and positive Hedges’s g values indicated increases in outcomes relative to controls at post-treatment and follow-up. Hedges’s g values were also converted to Cohen’s U3 values that represented the percentage of participants in the treatment group with better scores on an outcome measure than the average score on an outcome measure among participants in the control group (Cohen, 1988).

2.6. GRADE rating

The Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system was used (Guyatt, Oxman, Schünemann, Tugwell, & Knottnerus, 2011; Schünemann et al., 2023) to assess the quality of the evidence for each outcome included in the moderate- and high-quality meta-analyses. Consistent with this system, critical and important outcomes were identified (Guyatt et al., 2011), and the determination of outcomes as critical versus important was guided using the Banff, Alberta Consensus (Walker et al., 2006). Critical outcomes were disorder-specific symptoms and behavior (i.e., gambling disorder symptom severity, gambling frequency, gambling intensity, gambling duration), and important outcomes were psychological functioning (i.e., anxiety, depression, quality of life, substance use).

Four domains of quality were considered for each outcome: inconsistency, imprecision, indirectness, and publication bias. Inconsistency was assessed using I2 values where 25%, 50%, and 75% represented low, moderate, or high heterogeneity, respectively (Higgins, Thompson, Deeks, & Altman, 2003). Outcomes had serious inconsistency if heterogeneity was high, and outcomes had no serious inconsistency if heterogeneity was low or moderate. Imprecision was assessed based on the significance of the outcome and the sample size required to detect a clinically meaningful effect size with a power of 0.80 at α = 0.05 (Guyatt et al., 2011). Outcomes had serious imprecision if either the effect size was not significant or the sample sizes were smaller than needed to detect a clinically meaningful effect size, and outcomes had no serious imprecision if both the effect size was significant and the sample size exceeded what was needed to detect a clinically meaningful effect size. Indirectness represented differences in populations, treatments, outcomes, and comparison from those of interest in the present review (Guyatt et al., 2011). Publication bias was detected if funnel plots were asymmetrical and/or statistical tests supported the asymmetry of effect sizes (Guyatt et al., 2011).

Based on the assessment of quality for each domain, outcomes were assigned an overall rating of high, moderate, low, or very low to represent confidence that the true effect lies close to the estimated effect size (Balshem et al., 2011). Consistent with the recommendations of Tolin, McKay, Forman, Klonsky, and Thombs (2015), several contextual factors, such as the effect of CBT relative to other treatments, the dose of CBT needed to generate a clinically meaningful effect, the evidence for mechanisms of CBT, the effects of CBT with marginalized groups, and the allegiance of researchers investigating the effects of CBT, were considered to increase or decrease the overall quality ratings.

3. Results

3.1. Summary of Selected Meta-Analyses

Across the five meta-analyses, there were 87 individual studies. However, some of these studies were included in multiple meta-analyses (see Supplemental Table 3), with one study included in five meta-analyses, seven studies included in three meta-analyses, and 11 included in two meta-analyses. In total, there were 56 unique studies that were included in the present evaluation of CBT for gambling harm, representing 5389 participants.

Table 1 presents the five meta-analyses that were included in the present evaluation of CBT for gambling harm. All five meta-analyses comprised a sample of individuals experiencing problem gambling or gambling disorder, and none of the meta-analyses specified a treatment setting. Three meta-analyses examined the effect of CBT relative to treatment as usual (e.g., referral to Gamblers’ Anonymous, brief advice) and inactive controls (e.g., waitlist, assessment only). One meta-analysis examined the effect of CBT versus active (e.g., motivational interviewing) and inactive controls, as well as pre-post effects of CBT. One-meta-analysis examined the effect of CBT using pre-post comparisons only.

Table 1.

Meta-analyses included in the evaluation of cognitive-behavioral treatment as an empirically supported treatment for gambling harm.

Study k Population Intervention(s) Comparison condition
(s)		 Setting Outcome(s) Time points
Bergeron et al. (2022) 5 Individuals with PG or GD CBT (in-person) None (Pre-post only) NS Severity; Duration; Beliefs; Craving Posttreatment (0 months); Follow-up (6 and 12 months)
Cowlishaw et al. (2012) 11 Individuals with PG or GD CBT (in-person) Treatment as usual; Inactive control NS Severity; Frequency; Intensity; GD diagnosis; Anxiety; Depression Posttreatment (0–3 months); Follow-up (9–12 months)
Gooding and Tarrier (2009) 25 Individuals with PG or GD CBT (in-person, workbook) Active control; Inactive control; None (Pre-post only) NS Severity; Frequency; Duration; Abstinence; Craving Posttreatment (0–3 months); Follow-up (6 months)
Pfund, Forman, et al. (2023) 29 Individuals with PG or GD CBT (in-person, Internet, workbook) Treatment as usual; Inactive control NS Severity; Frequency; Intensity; Duration Posttreatment (0 months); Follow-up (1–22 months)
Pfund, King, et al. (2023) 10 Individuals with PG or GD CBT (in-person, Internet, workbook) Treatment as usual; Inactive control NS Anxiety; Depression; Quality of life; Substance use Posttreatment (0 months)
Open in a new tab

Notes. CB = cognitive-behavioral treatment; GD = gambling disorder; k = number of studies, NS = not specified; PG = problem gambling.

Table 2 presents the outcomes and timepoints when these outcomes were measured across the five meta-analyses. At posttreatment, four meta-analyses reported gambling disorder severity, three reported gambling frequency, two meta-analyses reported gambling intensity, gambling duration, anxiety, and depression, one meta-analysis reported gambling abstinence, gambling disorder diagnosis, gambling craving, gambling beliefs, substance use, and quality of life. At follow-up, three meta-analyses reported gambling disorder severity and gambling frequency, two meta-analyses reported gambling duration and gambling cravings, and one reported gambling intensity, gambling abstinence, and gambling beliefs.

Table 2.

Outcomes of meta-analyses included in the evaluation of cognitive-behavioral treatment.

Study Outcome Time Point k Hedges’s g 95% CI I 2
Bergeron et al. (2022) Severity Posttreatment (0 months) 4 −1.09 −1.52, −0.64 NR
Follow-up (6 & 12 months) 2 −1.69 −2.74, −0.62 NR
Duration Posttreatment (0 months) 5 −1.92 −2.77, −1.07 NR
Follow-up (6 & 12 months) 2 −2.45 −5.33, 0.43 NR
Beliefs Posttreatment (0 months) 2 −0.65 −1.32, 0.05 NR
Follow-up (6 & 12 months) 2 −1.31 −2.01, −0.61 NR
Craving Posttreatment (0 months) 3 −0.96 −1.76, −0.12 NR
Follow-up (6 & 12 months) 2 −1.01 −1.50, −0.49 NR
Cowlishaw et al. (2012) Severity Posttreatment (0–3 months) 7 −1.82 −2.61, −1.02 90%
Follow-up (9–12 months) 1 −0.11 −0.43, 0.22 NA
Frequency Posttreatment (0–3 months) 7 −0.78 −1.11, −0.45 61%
Follow-up (9–12 months) 1 −0.12 −0.45, 0.20 NA
Intensity Posttreatment (0–3 months) 7 −0.52 −0.71, −0.33 0%
Follow-up (9–12 months) 1 −0.15 −0.47, 0.18 NA
GD diagnosis Posttreatment (0–3 months) 2 0.13 0.05, 0.31 0%
Anxiety Posttreatment (0–3 months) 4 −0.64 −0.90, −0.37 0%
Depression Posttreatment (0–3 months) 4 −0.66 −0.93, −0.39 0%
Gooding and Tarrier (2009) Severity Posttreatment (0–3 months) 3 −0.58 −1.20, 0.04 NR
Frequency Posttreatment (0–3 months) 6 −0.79 −1.21, −0.36 NR
Follow-up (6 months) 5 −0.23 −0.66, 0.20 NR
Duration Posttreatment (0–3 months) 2 −0.68 −1.70, 0.35 NR
Follow-up (6 months) 2 −0.65 −1.50, 0.21 NR
Abstinence Posttreatment (0–3 months) 3 −1.87 −1.54, −0.20 NR
Craving Follow-up (6 months) 3 −1.05 −1.63, −0.46 NR
Pfund, Forman, et al. (2023) Severity Posttreatment (0 months) 20 −1.14 −1.68, −0.60 91%
Follow-up (1–22 months) 8 −0.58 −1.52, 0.36 90%
Frequency Posttreatment (0 months) 18 −0.54 −0.80, −0.27 79%
Follow-up (1–22 months) 7 −0.10 −0.30, 0.10 17%
Intensity Posttreatment (0 months) 14 −0.32 −0.51, −0.13 34%
Pfund, King, et al. (2023) Anxiety Posttreatment (0 months) 7 −0.44 −0.70, −0.18 42%
Depression Posttreatment (0 months) 7 −0.35 −0.69, −0.01 66%
Quality of life Posttreatment (0 months) 4 0.43 0.16, 0.70 11%
Substance use Posttreatment (0 months) 2 −0.40 −0.82, 0.03 NA
Open in a new tab

Notes. k = number of studies; NR = not reported; NA = not applicable.

Bolded Hedges’s g values indicate that these values are significant.

Across all outcomes, the magnitude of effect sizes varied considerably. The posttreatment effect sizes ranged from small (Hedges’s g = −0.11) to large (Hedges’s g = −1.92). This variation was similar for follow-up effect sizes ranging from small (Hedges’s g = −0.10) to large (Hedges’s g = −2.45). Seventeen of the 21 posttreatment effect sizes (81%) were significant, and 4 of the 12 (33%) follow-up effect sizes were significant. Heterogeneity was substantial for many effect sizes, with I2 values ranging from 0% to 91%.

3.2. AMSTAR-2 ratings

Table 3 presents the methodological quality of the five meta-analyses according to the AMSTAR-2 system. Collectively, the five meta-analyses were considered adequate for drawing reasonable conclusions about the efficacy of CBT for gambling harm. The Cowlishaw et al. (2012) was the only meta-analysis rated as high quality; the only noncritical weakness was the funding sources of studies included in their meta-analysis were not reported. The Pfund, Forman, et al. (2023) and Pfund, King, et al. (2023) meta-analyses were both rated “moderate” quality because neither meta-analysis explained the selection of the study designs (i.e., randomized controlled trials) for inclusion in the meta-analysis nor assessed the potential impact of risk of bias on the results. The Bergeron et al. (2022) and Gooding and Tarrier (2009) meta-analyses were both rated “critically low” quality because they did not include components of PICOTS in the research questions and inclusion criteria, did not use a comprehensive literature search strategy, did not use appropriate methods to statistically combine results, and did not provide a satisfactory explanation for or discussion of any heterogeneity observed in the results.

Table 3.

AMSTAR 2 ratings for five meta-analyses of cognitive-behavioral treatment for gambling harm.

Item Bergeron et al. (2022) Cowlishaw et al. (2012) Gooding and Tarrier (2009) Pfund, Forman, et al. (2023) Pfund, King, et al. (2023)
1. Did the research questions and inclusion criteria for the review include components of PICO? N Y N Y Y
2. Did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the review and did the report justify any significant deviations from the protocol? N Y N Y Y
3. Did the review authors explain their selection of the study designs for inclusion in the review? N Y N N N
4. Did the review authors use a comprehensive literature search strategy? PY Y N PY PY
5. Did the review authors perform study selection in duplicate? Y Y N Y Y
6. Did the review authors perform data extraction in duplicate? Y Y N Y Y
7. Did the review authors provide a list of excluded studies and justify the exclusions? N Y N Y Y
8. Did the review authors describe the included studies in adequate detail? N Y N PY PY
Did the review authors use a satisfactory technique for assessing the RoB in individual studies that were included in the review?
RCT NA Y N Y Y
NRSI Y NA N NA NA
Did the review authors report on the sources of funding for the studies included in the review? N N N Y Y
11. If meta-analysis was performed did the review authors use appropriate methods for statistical combination of results?
RCT NA Y Y Y Y
NRSI N NA Y NA NA
12. If meta-analysis was performed, did the review authors assess the potential impact of RoB in individual studies on the results of the meta-analysis or other evidence synthesis? NA Y Y N N
13. Did the review authors account for RoB in individual studies when interpreting/discussing the results of the review? NA Y Y Y Y
14. Did the review authors provide a satisfactory explanation for, and discussion of, any heterogeneity observed in the results of the review? N Y Y Y Y
15. If they performed quantitative synthesis did the review authors carry out an adequate investigation of publication bias (small study bias) and discuss its likely impact on the results of the review? N Y N Y Y
16. Did the review authors report any potential sources of conflict of interest, including any funding they received for conducting the review? N Y N Y Y
Overall rating Critically low High Critically low Moderate Moderate
Open in a new tab

Notes. Items in bold are considered critical weaknesses if coded “no.” All studies were double coded, and discrepancies were resolved by consensus to arrive at final ratings.

N = no; NA = not applicable; NRSI = nonrandomized studies of interventions; PICO = population, intervention, comparator group, and outcome; PY = partial yes; RCT = randomized controlled trial; RoB = risk of bias; Y = yes.

3.3. Meta-analysis

A robust variance meta-analysis was conducted to examine the effect of CBT, relative to control, with the two meta-analyses rated “critically low” quality removed from analyses (Table 4). The total number of studies comprising this meta-analysis was 31, and the total number of participants was 4023. At posttreatment, three outcomes, gambling disorder severity, gambling frequency, and gambling intensity, had sufficient degrees of freedom. The Hedges’s g effect sizes for all three outcomes were smaller than the average weighted Hedges’s g effect sizes with the “critically low” quality meta-analyses included. However, both analyses indicated that the effect of CBT was significantly better than controls on these three outcomes. (See Table 4.)

Table 4.

Robust variance estimation meta-analysis for all outcomes reported in the cowlishaw, pfund forman, and pfund king meta-analyses only.

Posttreatment
Outcome Meta-analyses included in Hedges’s g
calculation		 k comprising Hedges’s g
value		 n comprising Hedges’s g
value		 Hedges’s g 95% CI p-
value
Gambling disorder severity Cowlishaw, Pfund Forman 21 1643 −0.91 −1.33, −0.49 0.0002
Gambling frequency Cowlishaw, Pfund Forman 19 2468 −0.52 −0.77, −0.27 0.0004
Gambling intensity Cowlishaw, Pfund Forman 16 1549 −0.32 −0.48, −0.16 0.0009
Gambling duration Pfund Forman 5 263 – – –
Anxiety Cowlishaw, Pfund King 5 302 – – –
Depression Cowlishaw, Pfund King 5 302 – – –
Quality of life Pfund King 4 245 – – –
Substance use Pfund King 2 102 – – –
Follow-Up
Outcome Meta-analyses included in Hedges’s g
calculation		 k comprising Hedges’s g
value		 n comprising Hedges’s g
value		 Hedges’s g 95% CI p-
value
Gambling disorder severity Cowlishaw, Pfund Forman 7 848 −0.14 −0.40, 0.11 0.11
Gambling frequency Cowlishaw, Pfund Forman 7 1482 −0.10 −0.30, 0.10 0.26
Gambling intensity Cowlishaw, Pfund Forman 5 615 – – –
Open in a new tab

Notes. Omitted results (denoted with -) could not be estimated due to limited degrees of freedom.

CI = confidence interval; k = number of studies; n = number of participants.

At posttreatment, the effect of CBT relative to control on gambling disorder severity (k = 21, n = 1643) was g = −0.91, 95% CI [−1.33, −0.49]. There was considerable between-study heterogeneity (τ = 0.52), and 89% of the observed variability was attributable to true heterogeneity, suggesting considerable variation. This heterogeneity suggested that considerable variation in effects of CBT on gambling disorder severity at posttreatment could be expected in future trials (95% PI [−2.48, 0.66]). The effect of CBT relative to control on gambling frequency at posttreatment (k = 19, n = 2468) was g = −0.52, 95% CI [−0.77, −0.27]. There was also considerable between-study heterogeneity for this outcome (τ = 0.17) with 78% of the observed variability attributable to true heterogeneity. Similarly, this heterogeneity suggested that considerable variation in effects of CBT on gambling frequency at posttreatment could be expected in future trials (95% PI [−1.43, 0.39]). The effect of CBT relative to control on gambling intensity at posttreatment (k = 16, n = 1549) was g = −0.32, 95% CI [−0.48, −0.16]. Between-study heterogeneity was minimal (τ = 0.02), and 27% of the observed variability was attributable to true heterogeneity. However, this heterogeneity suggested that the effects of CBT on gambling intensity at posttreatment may not be reliable in future trials (95% PI [−0.67, 0.03]). The Cohens U3 values for gambling disorder severity, gambling frequency, and gambling severity at post-treatment indicated that between 65% and 82% of participants receiving CBT will show greater reductions in these outcomes than controls.

At follow-up, the effect of CBT on gambling disorder severity and gambling frequency was not significantly different from control. The effect of CBT relative to control on other outcomes could not be examined due to an insufficient number of degrees of freedom (Tipton, 2015) or because there were not follow up data available for these outcomes.

3.4. GRADE ratings

Three critical outcomes, gambling disorder severity, gambling frequency, and gambling frequency at posttreatment were rated moderate quality. None of these three posttreatment outcomes had serious indirectness or imprecision (Table 5). Serious inconsistency was found for gambling disorder severity and gambling frequency at posttreatment, but these concerns were dampened because the inconsistency may be due to differences in the number of sessions of treatment attended (Pfund et al., 2020), the format of CBT (i.e., therapist-assisted versus self-guided; Goslar, Leibetseder, Muench, Hofmann, & Laireiter, 2017), and grant funding (i.e., grant funded versus not; Pfund, Forman, et al., 2023). However, publication bias was detected for all three posttreatment outcomes based on asymmetry in funnel plots. Thus, all three critical outcomes were rated moderate quality based on GRADE. The consideration of contextual factors did not increase the confidence in GRADE ratings because there was an overall lack of evidence to make firm conclusions about most contextual factors (Table 6)

Table 5.

GRADE ratings of quality of the evidence for cognitive-behavioral treatment for gambling harm.

Posttreatment
 
 GRADE Quality Assessment
 Summary of Findings
  
Outcome Inconsistency Indirectness Imprecision Publication
Bias		 N Effect Size
(95% CI)		 I 2 Quality of
Evidence
Gambling disorder severity (critical) 21 RCTs Serious inconsistency No serious indirectness No serious imprecision Detected 1643 −0.91 (−1.33, −0.49) 89% ⊕⊕⊕Ο Moderate
Gambling frequency (critical) 19 RCTs Serious inconsistency No serious indirectness No serious imprecision Detected 2468 −0.52 (−0.77, −0.27) 78% ⊕⊕⊕Ο Moderate
Gambling intensity (critical) 16 RCTs No serious inconsistency No serious indirectness No serious imprecision Detected 1549 −0.32 (−0.48, −0.16) 27% ⊕⊕⊕Ο Moderate
Gambling duration (critical) 5 RCTs NA No serious indirectness NA NA 263 – – ⊕ΟΟΟ Very low
Anxiety (important) 5 RCTs NA No serious indirectness NA NA 302 – – ⊕ΟΟΟ Very low
Depression (important) 5 RCTs NA No serious indirectness NA NA 302 – – ⊕ΟΟΟ Very low
Quality of life (important) 4 RCTs NA No serious indirectness NA NA 245 – – ⊕ΟΟΟ Very low
Substance use (important) 2 RCTs NA No serious indirectness NA NA 102 – +OOO Very low
Follow-Up
Substance use (important) 2 RCTs GRADE Quality Assessment Summary of Findings
Outcome Inconsistency Indirectness Imprecision Publication
bias		 N Effect Size
(95% CI)		 I 2 Quality of
Evidence
Gambling disorder severity (critical) 7 RCTs No serious inconsistency No serious indirectness Serious imprecision Detected 955 −0.14 (−0.40, 0.11) 24% ⊕ ⊕ OO
Gambling frequency (critical) 7 RCTs No serious inconsistency No serious indirectness Serious imprecision Detected 1518 −0.10 (−0.30, 0.10) 17% ⊕ ⊕ oo
Gambling intensity (critical) 5 RCTs NA No serious indirectness NA NA 778 – – ⊕ΟΟΟ Very low
Open in a new tab

Notes. CI = confidence interval; N = no; NA = not assessed; RCT = randomized controlled trial; Y = yes.

Table 6.

Additional contextual factors considered in increasing or decreasing the GRADE recommendation.

Positive Negative
✓ Treatment appears superior to other well-studied treatment(s) □There are other psychological treatments that have well-documented and much larger effects
□ The treatment generates an effect that is similar to other well-studied treatments, but requires a very small number of sessions or length of time to generate the same effect at a much lower cost □ The treatment generates an effect that is similar to other well-studied treatments, but requires a very large number of sessions or length of time to generate the same effect at a much higher cost
□ Evidence supports the purported mechanism or active ingredient(s) of treatment □ Evidence fails to support the purported mechanism or active ingredient(s) of treatment
□ Treatment has demonstrated good effects with marginalized groups □ Treatment has demonstrated weak effects with marginalized groups
□ Treatment has been studied by a wide array of researchers without strong allegiance to the treatment □ Treatment has been studied by a narrow array of researchers with strong allegiance to the treatment
✓ Other: Treatment is flexibly adapted to Internet modules and self-help workbooks □Other:
Open in a new tab

Note. This table identifies additional positive contextual factors supported by the literature on cognitive-behavioral treatment for gambling harm and was adapted from Tolin et al. (2015). Lack of identification of a positive or negative assessment of a contextual factor indicates that there are not enough data to make a firm conclusion in this category.

One critical outcome, gambling duration at posttreatment, was rated very low quality. Although this outcome had no serious indirectness, a reliable effect size could not be estimated due to a lack of studies. All the other important posttreatment outcomes (i.e., anxiety, depression, quality of life, and substance use at posttreatment) were rated similarly for the same reason.

Of the critical outcomes at follow-up, gambling disorder severity and gambling frequency were rated low quality. Neither outcome had serious inconsistency or indirectness, but they both had serious imprecision and publication bias. Gambling intensity at follow-up was rated very low because there was a lack of studies to reliably estimate an effect size.

3.5. Overall treatment recommendation

Based on the criteria outlined by Tolin et al. (2015), the current literature merits a “strong” recommendation for CBT for gambling harm (see Table 7). There is at least moderate-quality evidence that CBT produces a statistically significant and clinically meaningful posttreatment effect on gambling disorder severity, gambling frequency, and gambling intensity relative to controls, as demonstrated by the studies comprising the Cowlishaw and Pfund meta-analyses. However, there is limited evidence that CBT affects functional outcomes and whether the effects of CBT endure after termination. Furthermore, there is limited evidence on the effectiveness of CBT in non-research settings.

Table 7.

Overall treatment recommendation for cognitive-behavioral treatment for gambling harm.

Recommendation Criteria
□ Very strong recommendation All the following:

  • There is high-quality evidence that the treatment produces a clinically meaningful effect on symptoms of the disorder being treated.

  • There is high-quality evidence that the treatment produces a clinically meaningful effect on functional outcomes.

  • There is high-quality evidence that the treatment produces a clinically meaningful effect on symptoms and/or functional outcomes at least three months after treatment discontinuation.

  • At least one well-conducted study has demonstrated effectiveness in nonresearch settings.
✓ Strong recommendation At least one of the following:

  • There is moderate- to high-quality evidence that the treatment produces a clinically meaningful effect on symptoms of the disorder being treated.

  • There is moderate- to high-quality evidence that the treatment produces a clinically meaningful effect on functional outcomes
□ Weak recommendation Any of the following:

  • There is only low- or very low-quality evidence that the treatment produces a clinically meaningful effect on symptoms of the disorder being treated.

  • There is only low- or very low-quality evidence that the treatment produces a clinically meaningful effect on functional outcomes.

  • There is moderate- to high-quality evidence that the effect of the treatment, although statistically significant, may not be of a magnitude that is clinically meaningful
Open in a new tab

Note. This table was adapted from Tolin et al. (2015).

4. Discussion

The present umbrella review synthesized the methodological rigor of five meta-analyses representing 56 unique studies and 5389 participants on the immediate and enduring effects of CBT for gambling harm. Collectively, these meta-analyses were deemed high-quality and resulted in a “strong” recommendation for CBT as an empirically supported treatment based on the Tolin et al. (2015) criteria. This recommendation is consistent with the findings of previous reviews of the gambling treatment literature (Cowlishaw et al., 2012; Petry et al., 2017), and the results of the umbrella review provide individuals, clinicians, managed care companies, and policymakers with rigorous empirical evidence that CBT produces immediate, clinically meaningful reductions in gambling disorder severity and gambling behavior.

Based only on the moderate- and high-quality meta-analyses (Cowlishaw et al., 2012; Pfund, Forman, et al., 2023; Pfund, King, et al., 2023), CBT significantly reduced gambling disorder severity, gambling frequency, and gambling intensity at posttreatment relative to minimal and inactive treatment controls. Compared to other well-studied treatments in the literature, these reductions across three different three outcomes (g = −0.32 to g = −0.91) are larger than the effects of antidepressants, opioid antagonists, and atypical antipsychotics on these same outcomes at posttreatment (Dowling et al., 2022).

The effects of CBT on all three outcomes at posttreatment are also larger than the reported effects of brief interventions on gambling behavior at posttreatment (g = −0.19) (Quilty, Wardell, Thiruchselvam, Keough, & Hendershot, 2019). Some researchers have argued that single-session, brief interventions produce similar effects on outcomes as multiple sessions of CBT (Toneatto, 2016). However, these researchers have not considered that an estimated 39% of individuals prematurely discontinue treatment (Pfund et al., 2021), and greater session attendance is significantly associated with more beneficial outcomes (Pfund et al., 2020). Most participants in randomized controlled trials of psychological treatment for gambling harm attend 1–2 sessions (Pfund et al., 2020), and the effect of attending one of six possible CBT sessions is comparable to the effect of attending a single-session, brief intervention (Larimer et al., 2012). Thus, it is likely that the effects of CBT are larger than the effects of brief interventions due to the number of sessions attended.

The effect of CBT on gambling frequency at posttreatment in the present meta-analysis (g = −0.52) is smaller than the effect of motivational interviewing on gambling frequency at posttreatment (g = −1.22) (Yakovenko, Quigley, Hemmelgarn, Hodgins, & Ronksley, 2015), and the effect of CBT on gambling intensity at posttreatment in the presentmeta-analysis (g = −0.32) is comparable to the effect of motivational interviewing on gambling intensity at posttreatment (g = −0.26) (Yakovenko et al., 2015). However, the meta-analysis on motivational interviewing comprises fewer studies (3–5 studies per estimate of treatment effects) than the present meta-analysis on CBT (16–21 per estimate of treatment effects). Furthermore, few head-to-head comparisons of CBT and motivational interviewing have been conducted, and those that have been conducted indicate no significant difference in outcomes between treatments (Carlbring, Jonsson, Josephson, & Forsberg, 2010). It is currently not possible to conclude that motivational interviewing produces larger reductions in gambling frequency and intensity than CBT based on the available evidence for motivational interviewing.

The body of literature identified in the present umbrella review serves as a strong foundation for researchers to understand how to optimize CBT for individuals experiencing gambling harm. However, there is considerable heterogeneity in the magnitude of effect size estimates whereby some individuals realized large reductions in gambling disorder severity and gambling behavior, some individuals realized no changes, and some individuals realized deterioration (Pfund, Forman, et al., 2023). More research is needed on the optimization of CBT using the longstanding quote “what treatment, by whom, is most effective for this individual with that specific problem, and under which specific set of circumstances” (Paul, 1967, p. 111).

Identifying what components of CBT are most effective will require researchers to dismantle the various treatment components and examine which components drive outcomes. In the present umbrella review, there were no limits on forms of CBT; included studies comprised CBT protocols that facilitated the identification of high-risk situations precipitating gambling, exposure to gambling cues, engagement in other value-consistent activities and adoption of new coping strategies, limit setting for time and amount of money gambled, cognitive restructuring, and/or relapse prevention. Attempts to dismantle various components of CBT for gambling harm have been limited in sample size and have not used rigorous methods to generate randomization sequences and randomly assign participants (Echeburúa, Baez, Fernandez-Montalvo, Báez, & Fernández-Montalvo, 1996). Dismantling studies on treatments for substance use disorder (SUD) are also limited and focus on the identification of statistical mediators of treatment outcomes (Magill et al., 2020; Witkiewitz, Pfund, & Tucker, 2022).

There is currently scant knowledge on who conducts CBT for gambling harm and how certain clinicians may facilitate better outcomes than other clinicians. No studies were located on this topic among individuals receiving CBT for gambling harm. However, the general psychotherapy literature suggests it would be advantageous to initiate research on differential effects of therapists based on clinical setting (Johns, Barkham, Kellett, & Saxon, 2019).

To maximize the effect of CBT with this individual experiencing gambling harm, researchers will need to recruit more racially and socioeconomically diverse samples for randomized controlled trials. The present umbrella review included studies from six of the seven continents, but a recent systematic review of randomized controlled trials of psychological treatments for gambling harm in the United States suggested that samples mostly represented participants who identified as White, were employed, and reported some level of college education (Peter, Pfund, & Ginley, 2021). The current body of knowledge may not generalize to individuals representing historically marginalized groups who disproportionately experience gambling harms (Welte, Barnes, Hoffman, & Wieczorek, 2015).

Current research is limited in that it often excludes samples of individuals with comorbid psychological conditions (Dowling, Merkouris, & Lorains, 2016; Pfund, King, et al., 2023). Yet, comorbid psychological conditions, especially SUDs, are common among individuals with gambling disorder (Dowling et al., 2015; Lorains, Cowlishaw, & Thomas, 2011). Given the considerable overlap in risk and maintenance factors between gambling disorder and SUDs, transdiagnostic theories of addictions propose addictive behaviors as representing the same underlying syndrome with different manifestations (Shaffer et al., 2004). More recently, Kim and Hodgins (2018) proposed a transdiagnostic treatment model for addiction that targets the overlapping maintaining mechanisms of both substance and behavioral addictions (e.g., emotion dysregulation, impulsivity). Beyond this transdiagnostic model, other models exist to inform the testing of specific hypotheses for personalization to individuals with specific symptom profiles (e.g., Boness & Witkiewitz, 2023; Ruggero et al., 2019). Given the high rates of comorbidity among individuals with gambling disorder, future randomized controlled trials that examine the efficacy of transdiagnostic treatments in reducing symptoms of gambling disorder and comorbid conditions compared to gambling-specific treatments would be highly informative.

Finally, maximizing the effects of CBT under which specific set of circumstances will require researchers to expand temporal units of analysis and contexts. Consistent with social cognitive theory (Bandura, 2001), most research focuses on individuals as the agents of change. This research samples individual behavior at discrete temporally contiguous time points, which neglects that behavior is continuously evolving and occurring in broader contexts of communities and neighborhoods. Future research on the effect of CBT among individuals with gambling harm needs to consider community- and neighborhood-level variables and social determinants of health (Witkiewitz et al., 2022). For example, what is the effect of CBT among individuals living in communities with legalized opportunities to gamble versus communities without legalized opportunities to gamble? Such questions will be important to understand the contexts and specific circumstances of individuals receiving CBT.

This present umbrella review is not without limitations. One limitation is that it was not possible to separate some posttreatment and follow-up effects. For example, the Cowlishaw et al. (2012) meta-analysis defined posttreatment as both immediately following treatment termination and up to 3 months after termination, and other meta-analyses defined posttreatment as only immediately following termination. A second limitation was that heterogeneity was considerable for many of the effect sizes. Several moderators have been identified in the literature on gambling treatment, including the number of sessions attended (Pfund et al., 2020), the presence of grant funding (Pfund, Forman, et al., 2023), and the format (e.g., in-person versus self-guided) of CBT (Goslar et al., 2017). However, future meta-analyses are needed to completely understand the heterogeneity in treatment effects. Third, the availability of long-term follow-up data was scarce – only seven studies reported the effect of CBT on some outcome after treatment was terminated. Thus, it was not possible to fully understand the effect of CBT on outcomes after termination and the association between the time elapsed since termination and the magnitude of effect sizes. Future longitudinal research will be needed to understand how treatment effects may deteriorate over time.

Despite the limitations of the current umbrella review and existing literature, the present evaluation is the first to evaluate the strength of CBT as an empirically supported treatment for gambling harm. Importantly, the evaluation of CBT in this manner provides individuals, clinicians, managed care companies, policymakers, and national organizations, like the SCP, with formal recommendations on treatment for gambling harm. The evaluation also provides further support for the application of the Tolin criteria in establishing recommendations for the field (Boness et al., 2023; Leichsenring et al., 2023; Pfund, Ginley, et al., 2023)

CBT for gambling harm can result in a meaningful reduction in gambling disorder severity and gambling behavior. High quality clinical trials that use comprehensive consideration of contextual factors, follow individuals over extended follow-ups, and purposively sample those individuals most likely to be experiencing gambling harm will allow for further clarification of the mechanisms of change. Research exploring, and ultimately tailoring CBT interventions in delivery format, content, and duration to an individual’s unique symptom profile, overall symptom severity of gambling disorder, and demographic risk factors can enhance uptake, improvement of comorbid symptomology, and increase treatment retention. Research to further improve the efficacy and effectiveness of CBT for gambling harm will require large, rigorous clinical trials that necessitate governments to provide funding (Weinstock, 2018).

Supplementary Material

supplemental material

Acknowledgments

This work was supported by the National Institutes of Health, award number K08AA030301. This work was also supported by the Tennessee Department of Mental Health and Substance Abuse Services. The content is the sole responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the state of Tennessee.

Footnotes

Declaration of Competing Interest

Authors RAP, MKG, and JPW were authors on two of the five meta-analyses included in the present study. To mitigate this conflict, RAP, MKG, and JPW did not contribute to the coding of the included meta-analyses. Author CLB was a member of the American Psychological Association’s Division 12 Committee on Science and Practice and aided in the development of the manual used to guide this evaluation. To mitigate this conflict, CLB was not involved in the evaluation or discussion of the evaluation report submitted to the Division 12 Committee on Science and Practice. No other conflicts of interest were declared.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cpr.2023.102336.

Data availability

Data used to examine the effects of CBT on multiple outcomes at the primary study level can be found and retrieved on www.metapsy.org/database/gambling.

References

  1. American Gaming Association. (2023). Interactive U.S. Map: Sports Betting. American Gaming Association. https://www.americangaming.org/research/state-gaming-map/. [Google Scholar]
  2. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed., p. 223). American Psychiatric Association. [Google Scholar]
  3. American Psychological Association. (2006). Evidence-based practice in psychology. American Psychologist, 61(4), 271–285. 10.1037/0003-066X.61.4.271 [DOI] [PubMed] [Google Scholar]
  4. Appelbaum M, Cooper H, Kline RB, Mayo-Wilson E, Nezu AM, & Rao SM (2018). Journal article reporting standards for quantitative research in psychology: The APA publications and communications board task force report. American Psychologist, 73(1), 947. 10.1037/amp0000389 [DOI] [PubMed] [Google Scholar]
  5. Balshem H, Helfand M, Schünemann HJ, Oxman AD, Kunz R, Brozek J, … Norris S (2011). GRADE guidelines: 3. Rating the quality of evidence. Journal of Clinical Epidemiology, 64(4), 401–406. 10.1016/j.jclinepi.2010.07.015 [DOI] [PubMed] [Google Scholar]
  6. Bandura A. (2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52, 1–26. [DOI] [PubMed] [Google Scholar]
  7. Bergeron P-Y, Giroux I, Chrétien M, & Bouchard S (2022). Exposure therapy for gambling disorder: Systematic review and meta-analysis. Current Addiction Reports, 9, 179–194. 10.1007/s40429-022-00428-5 [DOI] [Google Scholar]
  8. Boness CL, Hershenberg R, Grasso DJ, Kaye J, Mackintosh M-A, Nason E, … Raffa SD (2021). The Society of Clinical Psychology’s manual for the evaluation of psychological treatments using the Tolin Criteria. 10.31219/osf.io/8hcsz [DOI] [Google Scholar]
  9. Boness CL, Votaw VR, Schwebel FJ, Moniz-Lewis DIK, McHugh RK, & Witkiewitz K (2023). An evaluation of cognitive behavioral therapy for substance use disorders: A systematic review and application of the society of clinical psychology criteria for empirically supported treatments. Clinical Psychology: Science and Practice. 10.1037/cps0000131 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Boness CL, & Witkiewitz K (2023). Precision medicine in alcohol use disorder: Mapping etiologic and maintenance mechanisms to mechanisms of behavior change to improve patient outcomes. Experimental and Clinical Psychopharmacology. 10.1037/pha0000613 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Carlbring P, Jonsson J, Josephson H, & Forsberg L (2010). Motivational interviewing versus cognitive behavioral group therapy in the treatment of problem and pathological gambling: A randomized controlled trial. Cognitive Behaviour Therapy, 39, 92–103. 10.1080/16506070903190245 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Cohen J. (1988). Statistical power analysis for the behavioral sciences. Routledge. [Google Scholar]
  13. Cowlishaw S, Merkouris S, Dowling N, Anderson C, Jackson A, & Thomas S (2012). Psychological therapies for pathological and problem gambling. The Cochrane Database of Systematic Reviews, 11, CD008937. 10.1002/14651858.CD008937.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dowling N, Merkouris S, Lubman D, Thomas S, Bowden-Jones H, & Cowlishaw S (2022). Pharmacological interventions for the treatment of disordered and problem gambling. Cochrane Database of Systematic Reviews, 2022(9). 10.1002/14651858.CD008936.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dowling NA, Cowlishaw S, Jackson AC, Merkouris SS, Francis KL, & Christensen DR (2015). Prevalence of psychiatric co-morbidity in treatment-seeking problem gamblers: A systematic review and meta-analysis. The Australian and New Zealand Journal of Psychiatry, 49(6), 519–539. 10.1177/0004867415575774 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dowling NA, Merkouris SS, & Lorains FK (2016). Interventions for comorbid problem gambling and psychiatric disorders: Advancing a developing field of research. Addictive Behaviors, 58, 21–30. 10.1016/j.addbeh.2016.02.012 [DOI] [PubMed] [Google Scholar]
  17. Drevon D, Fursa SR, & Malcolm AL (2017). Intercoder reliability and validity of WebPlotDigitizer in extracting graphed data. Behavior Modification, 41(2), 323–339. 10.1177/0145445516673998 [DOI] [PubMed] [Google Scholar]
  18. Echeburúa E, Baez C, Fernandez-Montalvo J, Báez C, & Fernández-Montalvo J (1996). Comparative effectiveness of three therapeutic modalities in the psychological treatment of pathological gambling: Long-term outcome. Behavioural and Cognitive Psychotherapy, 24, 51–72. 10.1017/S1352465800016830 [DOI] [Google Scholar]
  19. Gooding P, & Tarrier N (2009). A systematic review and meta-analysis of cognitive-behavioural interventions to reduce problem gambling: Hedging our bets? Behaviour Research and Therapy, 47, 592–607. 10.1016/j.brat.2009.04.002 [DOI] [PubMed] [Google Scholar]
  20. Goodwin BC, Browne M, Rockloff M, & Rose J (2017). A typical problem gambler affects six others. International Gambling Studies, 17, 276–289. 10.1080/14459795.2017.1331252 [DOI] [Google Scholar]
  21. Goslar M, Leibetseder M, Muench HM, Hofmann SG, & Laireiter A-R (2017). Efficacy of face-to-face versus self-guided treatments for disordered gambling: A meta-analysis. Journal of Behavioral Addictions, 6, 142–162. 10.1556/2006.6.2017.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Grant JE, Donahue CB, Odlaug BL, Kim SW, Miller MJ, & Petry NM (2009). Imaginal desensitisation plus motivational interviewing for pathological gambling: Randomised controlled trial. The British Journal of Psychiatry, 195, 266–267. 10.1192/bjp.bp.108.062414 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Grinols EL (2011). The hidden social costs of gambling (pp. 19–28). [Google Scholar]
  24. Guyatt GH, Oxman AD, Kunz R, Atkins D, Brozek J, Vist G, … Schünemann HJ (2011). GRADE guidelines: 2. Framing the question and deciding on important outcomes. Journal of Clinical Epidemiology, 64(4), 395–400. 10.1016/j.jclinepi.2010.09.012 [DOI] [PubMed] [Google Scholar]
  25. Guyatt GH, Oxman AD, Kunz R, Brozek J, Alonso-Coello P, Rind D, … Schünemann HJ (2011). GRADE guidelines 6. Rating the quality of evidence—Imprecision. Journal of Clinical Epidemiology, 64(12), 1283–1293. 10.1016/j.jclinepi.2011.01.012 [DOI] [PubMed] [Google Scholar]
  26. Guyatt GH, Oxman AD, Kunz R, Woodcock J, Brozek J, Helfand M, … Schünemann HJ (2011). GRADE guidelines: 8. Rating the quality of evidence—Indirectness. Journal of Clinical Epidemiology, 64(12), 1303–1310. 10.1016/j.jclinepi.2011.04.014 [DOI] [PubMed] [Google Scholar]
  27. Guyatt GH, Oxman AD, Montori V, Vist G, Kunz R, Brozek J, … Schünemann HJ (2011). GRADE guidelines: 5. Rating the quality of evidence—Publication bias. Journal of Clinical Epidemiology, 64(12), 1277–1282. 10.1016/j.jclinepi.2011.01.011 [DOI] [PubMed] [Google Scholar]
  28. Guyatt GH, Oxman AD, Schünemann HJ, Tugwell P, & Knottnerus A (2011). GRADE guidelines: A new series of articles in the journal of clinical epidemiology. Journal of Clinical Epidemiology, 64(4), 380–382. 10.1016/j.jclinepi.2010.09.011 [DOI] [PubMed] [Google Scholar]
  29. Hedges LV, Tipton E, & Johnson MC (2010). Robust variance estimation in meta-regression with dependent effect size estimates. Research Synthesis Methods, 1(1), 39–65. 10.1002/jrsm.5 [DOI] [PubMed] [Google Scholar]
  30. Higgins JPT, Thompson SG, Deeks JJ, & Altman DG (2003). Measuring inconsistency in meta-analyses. BMJ, 327, 557–560. 10.1136/bmj.327.7414.557 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Hofmarcher T, Romild U, Spångberg J, Persson U, & Håkansson A (2020). The societal costs of problem gambling in Sweden. BMC Public Health, 20(1), 1921. 10.1186/s12889-020-10008-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Johns RG, Barkham M, Kellett S, & Saxon D (2019). A systematic review of therapist effects: A critical narrative update and refinement to Baldwin and Imel’s (2013) review. Clinical Psychology Review, 67, 78–93. 10.1016/j.cpr.2018.08.004 [DOI] [PubMed] [Google Scholar]
  33. Kim HS, & Hodgins DC (2018). Component model of addiction treatment: A pragmatic Transdiagnostic treatment model of behavioral and substance addictions. Frontiers in Psychiatry, 9, 406. 10.3389/fpsyt.2018.00406 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Langham E, Thorne H, Browne M, Donaldson P, Rose J, & Rockloff M (2015). Understanding gambling related harm: A proposed definition, conceptual framework, and taxonomy of harms. BMC Public Health, 16(1), 80. 10.1186/s12889-016-2747-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Larimer ME, Neighbors C, Lostutter TW, Whiteside U, Cronce JM, Kaysen D, & Walker DD (2012). Brief motivational feedback and cognitive behavioral interventions for prevention of disordered gambling: A randomized clinical trial. Addiction, 107, 1148–1158. 10.1111/j.1360-0443.2011.03776.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Latvala T, Lintonen T, & Konu A (2019). Public health effects of gambling – Debate on a conceptual model. BMC Public Health, 19(1), 1077. 10.1186/s12889-019-7391-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Leichsenring F, Abbass A, Heim N, Keefe JR, Kisely S, Luyten P, … Steinert C (2023). The status of psychodynamic psychotherapy as an empirically supported treatment for common mental disorders – An umbrella review based on updated criteria. World Psychiatry, 22(2), 286–304. 10.1002/wps.21104 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lorains FK, Cowlishaw S, & Thomas SA (2011). Prevalence of comorbid disorders in problem and pathological gambling: Systematic review and meta-analysis of population surveys. Addiction, 106, 490–498. 10.1111/j.1360-0443.2010.03300.x [DOI] [PubMed] [Google Scholar]
  39. Magill M, Tonigan JS, Kiluk B, Ray L, Walthers J, & Carroll K (2020). The search for mechanisms of cognitive behavioral therapy for alcohol or other drug use disorders: A systematic review. Behaviour Research and Therapy, 131, Article 103648. 10.1016/j.brat.2020.103648 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, … Moher D (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 372. 10.1136/bmj.n71 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Paul GL (1967). Strategy of outcome research in psychotherapy. Journal of Consulting Psychology, 31(2), 109–118. [DOI] [PubMed] [Google Scholar]
  42. Peter SC, Pfund RA, & Ginley MK (2021). Increased demographic representation in randomized control trials for gambling disorder in the United States is needed: A systematic review. Journal of Gambling Studies, 37(3), 1025–1041. 10.1007/s10899-021-10055-w [DOI] [PubMed] [Google Scholar]
  43. Petry NM, Ginley MK, & Rash CJ (2017). A systematic review of treatments for problem gambling. Psychology of Addictive Behaviors, 31, 951–961. 10.1037/adb0000290 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Pfund RA, Forman DP, Whalen SK, Zech JM, Ginley MK, Peter SC, … Whelan JP (2023). Effect of cognitive-behavioral techniques for problem gambling and gambling disorder: A systematic review and meta-analysis. Addiction. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Pfund RA, & Ginley MK (2019). Assessment and treatment of gambling behavior. Journal of Health Service Psychology, 45, 81–89. [Google Scholar]
  46. Pfund RA, Ginley MK, Boness CL, Rash CJ, Zajac K, & Witkiewitz K (2023). Contingency management for drug use disorders: Meta-analysis and application of Tolin’s criteria. Clinical Psychology: Science and Practice. 10.1037/cps0000121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Pfund RA, King SA, Forman DP, Zech JM, Ginley MK, Peter SC, … Whelan JP (2023). Effects of cognitive-behavioral techniques for gambling on recovery defined by gambling, psychological functioning, and quality of life: A systematic review and meta-analysis. Psychology of Addictive Behaviors. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Pfund RA, Peter SC, McAfee NW, Ginley MK, Whelan JP, & Meyers AW (2021). Dropout from face-to-face, multi-session psychological treatments for problem and disordered gambling: A systematic review and meta-analysis. Psychology of Addictive Behaviors, 35(8), 901–913. 10.1037/adb0000710 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Pfund RA, Peter SC, Whelan JP, Meyers AW, Ginley MK, & Relyea G (2020). Is more better? A meta-analysis of dose and efficacy in face-to-face psychological treatments for problem and disordered gambling. Psychology of Addictive Behaviors, 34(5), 557–568. 10.1037/adb0000560 [DOI] [PubMed] [Google Scholar]
  50. Quilty LC, Wardell JD, Thiruchselvam T, Keough MT, & Hendershot CS (2019). Brief interventions for problem gambling: A meta-analysis. PLoS One, 14(4), e0214502. 10.1371/journal.pone.0214502 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Ruggero CJ, Kotov R, Hopwood CJ, First M, Clark LA, Skodol AE, … Zimmermann J (2019). Integrating the hierarchical taxonomy of psychopathology (HiTOP) into clinical practice. Journal of Consulting and Clinical Psychology, 87(12), 1069–1084. 10.1037/ccp0000452 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Schardt C, Adams MB, Owens T, Keitz S, & Fontelo P (2007). Utilization of the PICO framework to improve searching PubMed for clinical questions. BMC Medical Informatics and Decision Making, 7, 1–6. 10.1186/1472-6947-7-16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Schünemann HJ, Brennan S, Akl EA, Hultcrantz M, Alonso-Coello P, Xia J, … Dahm P (2023). The development methods of official GRADE articles and requirements for claiming the use of GRADE – A statement by the GRADE guidance group. Journal of Clinical Epidemiology, 159, 79–84. 10.1016/j.jclinepi.2023.05.010 [DOI] [PubMed] [Google Scholar]
  54. Shaffer HJ, LaPlante DA, LaBrie RA, Kidman RC, Donato AN, & Stanton MV (2004). Toward a syndrome model of addiction: Multiple expressions, common etiology. Harvard Review of Psychiatry, 12(6), 367–374. 10.1080/10673220490905705 [DOI] [PubMed] [Google Scholar]
  55. Shea BJ, Reeves BC, Wells G, Thuku M, Hamel C, Moran J, … Henry DA (2017). AMSTAR 2: A critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ, 358, 1–9. 10.1136/bmj.j4008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Statista. (2023). Infographic: The largest online gambling markets. Statista Infographics. https://www.statista.com/chart/29048/largest-regulated-online-gambling-markets. [Google Scholar]
  57. Tipton E. (2015). Small sample adjustments for robust variance estimation with meta-regression. Psychological Methods, 20(3), 375–393. 10.1037/met0000011 [DOI] [PubMed] [Google Scholar]
  58. Tolin DF, McKay D, Forman EM, Klonsky ED, & Thombs BD (2015). Empirically supported treatment: Recommendations for a new model. Clinical Psychology: Science and Practice, 22, 317–338. 10.1111/cpsp.12122 [DOI] [Google Scholar]
  59. Toneatto T (2016). Single-session interventions for problem gambling may be as effective as longer treatments: Results of a randomized control trial. Addictive Behaviors, 52, 58–65. 10.1016/j.addbeh.2015.08.006 [DOI] [PubMed] [Google Scholar]
  60. Walker M, Toneatto T, Potenza MN, Petry N, Ladouceur R, Hodgins DC, … Blaszczynski A (2006). A framework for reporting outcomes in problem gambling treatment research: The Banff, Alberta Consensus. Addiction, 101(4), 504–511. 10.1111/j.1360-0443.2005.01341.x [DOI] [PubMed] [Google Scholar]
  61. Weinstock J. (2018). Call to action for gambling disorder in the United States. Addiction, 113, 1156–1158. [DOI] [PubMed] [Google Scholar]
  62. Welte JW, Barnes GM, Hoffman JH, & Wieczorek WF (2015). Gambling and problem gambling in the United States: Changes between 1999 and 2013. Journal of Gambling Studies, 31, 695–715. 10.1007/s10899-014-9471-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Winfree WR, Ginley MK, Whelan JP, & Meyers AW (2015). Psychometric evaluation of the Gamblers’ beliefs questionnaire with treatment-seeking disordered gamblers. Addictive Behaviors, 43, 97–102. 10.1016/j.addbeh.2014.12.016 [DOI] [PubMed] [Google Scholar]
  64. Witkiewitz K, Pfund RA, & Tucker JA (2022). Mechanisms of behavior change in substance use disorder with and without formal treatment. Annual Review of Clinical Psychology, 18, 1–29. 10.1146/annurev-clinpsy-072720-014802 [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Yakovenko I, Quigley L, Hemmelgarn BR, Hodgins DC, & Ronksley P (2015). The efficacy of motivational interviewing for disordered gambling: Systematic review and meta-analysis. Addictive Behaviors, 43, 72–82. 10.1016/j.addbeh.2014.12.011 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplemental material
NIHMS1986731-supplement-supplemental_material.pdf (232.6KB, pdf)

Data Availability Statement

Data used to examine the effects of CBT on multiple outcomes at the primary study level can be found and retrieved on www.metapsy.org/database/gambling.

RESOURCES