The COVID-19 pandemic, declared as a public health emergency by the World Health Organization [World Health Organization (WHO), 2020], has affected more than 200 countries around the world since the first case appeared in Wuhan (China) in December 2019 (Yuki et al., 2020). Spain has been considered one of the most affected countries by this pandemic since the early stages. Its first case of infection was reported within the first months of 2021, and since then, the number of cases has continued to increase. Along with other European countries, such as Italy and France, Spain has presented one of the highest mortality rates due to the pandemic (Ceylan, 2020).

To control pandemic spread, confinement and other restrictive measures have been stated everywhere (Hale et al., 2020), being social distancing, disruption of daily routines, and economic and employment uncertainty reported as some of the most frequent COVID-19-related concerns (Chew et al., 2020). This context alongside with health worries due to the pandemic has led to an increase in anxiety and depressive symptoms among general population (Islam et al., 2020; Wang et al., 2020). In this vein, people with a mental illness, such as gambling disorder (GD), have been considered at particularly high risk of suffering the psychological impact of the COVID-19 pandemic (González-Sanguino et al., 2020; Özdin and Bayrak Özdin, 2020; Rajkumar, 2020).

GD is defined as engaging in a repeated compulsive gambling behavior, accompanied by unsuccessful efforts to stop the behavior and uncontrollable urges to keep gambling, which leads to considerable distress and impairment. This gambling behavior tends to persist over time despite negative consequences on the person’s physical, mental, social, or financial well-being [American Psychiatry Association (APA), 2013]. GD was included within the “substance-related and addictive disorders” category in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders [DSM-5; American Psychiatry Association (APA), 2013]. It has been considered the most prevalent behavioral addiction, with a lifetime prevalence ranging from 1.1 to 10.6% (Calado and Griffiths, 2016). In Europe, epidemiological studies have shown that gambling problems range from 0.3 to 3.3% (Dirección General de Ordenación del Juego (DGOJ) 2015).

In the face of environmental changes, maladaptive coping strategies and behaviors, such as those related to gambling activity, could be used to cope with psychological distress (Asmundson and Taylor, 2020; Rajkumar, 2020; Fernández-Aranda et al., 2020a; Avena et al., 2021), specifically in individuals with higher vulnerability (e.g., those with a mental illness; Fernández-Aranda et al., 2020a; Baenas et al., 2020).

During confinement, the use of technologies has extended to more aspects of daily life. Hence, excessive engagement in specific online activities, such as gambling, may lead to severe problems and increases the risk of a disordered or addictive use (King et al., 2020; Mestre-Bach et al., 2020). These maladaptive behaviors are usually presented in combination with social isolation and boredom, factors which have been also linked to confinement and previously associated with the pathogenesis of GD (Ledgerwood and Petry, 2006; Mercer and Eastwood, 2010). On the other hand, the use of gambling has been also stated as a mechanism to deal with adverse economic circumstances in prior financial crises (Olason et al., 2015; Economou et al., 2019). Similarly, economic difficulties due to pandemic and confinement might represent a risk factor that could favor gambling activity. Moreover, some works have described a risk profile for engaging in gambling behavior during confinement (being young male regular gambler with high alcohol use prior to confinement; Håkansson, 2020a; Emond et al., 2021; Hodgins and Stevens, 2021). In the non-clinical study by Price (2020), online gambling modality was associated with the subgroup of gamblers at a higher risk of developing GD, characterized by greater psychological impact due to pandemic and substance use while gambling. In this line, gambling behavior prior to the pandemic predicted gambling activity during confinement, mainly at the expense of online gambling (Gainsbury et al., 2020; Håkansson, 2020b; Emond et al., 2021; Hodgins and Stevens, 2021).

With reference to the potential of the pandemic worsening gambling problems (King et al., 2020; Marsden et al., 2020), legal measures have been adopted by governments in an attempt to minimize these consequences. For instance, in Spain, advertising related to gambling has been limited (SBC News, 2020). Among other useful tools, a consensus guidance on how to avoid problematic use of the internet during the pandemic has been also published (Király et al., 2020). In this guidance, healthy general practices, such as those related to exercise or sleep hygiene were recommended, as well as specific tips regarding Internet use. Monitoring and regulating screen time exposure in children and adults, using digital well-being apps and analogue technical tools, keeping social contact with relatives and friends, or seeking professional help when needed were highly advised (Király et al., 2020).

Among people with GD, the challenges represented by the pandemic and confinement (Håkansson et al., 2020; King et al., 2020) were based on different vulnerability aspects (Allami et al., 2021; Hodgins and Stevens, 2021). For instance, higher levels of anxiety, depressive symptoms and hostility have been described in subjects with GD during confinement, in comparison with the general population (Salerno and Pallanti, 2021). In this context, gambling behavior could be conducted (and reinforced) to regulate emotions (Barrault et al., 2019), being a trigger for relapse in those patients who remained abstinent (Campbell, 2020). However, restrictive strategies and some policies adopted during confinement seemed to reduce gambling opportunities, specifically at offline locations, promoting abstinence (Donati et al., 2021). In this line, patients already in treatment for GD reported a positive attitude about restrictions due to pandemic and related to gambling activity, with improved social and familiar environments. This positive social reinforcement also contributed to maintain abstinence. Interestingly, large-scale migration from offline to online gambling has not been observed so far among patients with GD during confinement (Donati et al., 2021).

Works regarding GD treatment during confinement did not report an increase in treatment-seeking for GD, at least in the short term (Gunstone et al., 2020; Donati et al., 2021; Håkansson et al., 2021). When talking about GD treatment, cognitive-behavioral therapy (CBT) and group-based modalities have been widely demonstrated as effective options for the management of this disorder (Gooding and Tarrier, 2009; Cowlishaw et al., 2012; Jiménez-Murcia et al., 2015). Besides, internet-based treatments for GD demonstrated similar efficacy to traditional treatments (Hedman et al., 2012; Chakrabarti, 2015). Trying more flexible approaches to patient care in the face of pandemic situation (Marsden et al., 2020), tele-psychiatry has shown to be helpful (Hollander and Carr, 2020) and internet-based tools for GD have been used in some treatment facilities during confinement (Columb et al., 2020).

Within the treatment outcome in GD, dropout has been defined as leaving before the completion of a predetermined program (Milton et al., 2002; Robson et al., 2002). Despite a broad dropout range (14–50%) reported by previous national and international studies (Melville et al., 2007; Smith et al., 2010; Jimenez-Murcia et al., 2012), an average of around 26% has been agreed (Melville et al., 2007). Prior to confinement, a recent national study (Jiménez-Murcia et al., 2019) reported dropout rates of 32.4% among patients with GD regarding clinical trajectories based on GD severity among patients following a 12months manualized CBT program. Features such as patients’ medical or family issues, length of therapy, personality traits (e.g., perseverance, reward sensitivity, sensation-seeking), younger age, lower educational level, and neurocognitive variables, among others, have been studied as potential predictors of treatment dropout (Melville et al., 2007; Álvarez-Moya et al., 2011; Jiménez-Murcia et al., 2015; Mallorquí-Bagué et al., 2018, 2019; Mestre-Bach et al., 2019).

So far, most published studies concerning gambling activity during confinement focused on non-clinical populations (Auer et al., 2020; Brown and Hickman, 2020; Gunstone et al., 2020; Lindner et al., 2020; Zamboni et al., 2021). Scarce research on the effects of pandemic on GD treatment outcome and possible associated factors has been reported (Price, 2020; Donati et al., 2021). Furthermore, information related to initial stages of confinement provided a limited knowledge not only regarding its longer-term impact, but also in distinguishing its effects during time periods characterized by different levels of restrictive measures, such as the lockdown and the following months.

To the best of our knowledge, this is the first observational clinical study regarding the effect of the pandemic on GD treatment outcome related to dropout rates among clinical population. The longitudinal nature of the present work favors reporting data comparing two well-established periods of time across confinement (i.e., during and after the lockdown). Finally, the already existing involvement of these patients with the treatment Unit and a suitable observational period allow for evaluating clinical features possibly associated with dropout. Thus, the main aims of this study were to analyze treatment dropout rates in patients with GD in the two established periods, and assess clinical features that would be linked to treatment dropout in both periods.

The present study was designed to explore the impact of the COVID-19 pandemic and confinement on treatment adherence among patients with GD. For this purpose, we evaluated dropout rates both during and after the lockdown. The relationship between clinical variables and dropout was also explored in both periods.

Most patients of the sample were male and employed, and their main gambling problem involved offline gambling. This socio-demographic and clinical profile is similar to the characterization reported in previous studies of our Unit (Jiménez-Murcia et al., 2007, 2015). Regarding treatment outcome, the risk of dropout over the nine-month observational period reached similar values to those mentioned in the literature (Melville et al., 2007) and described before confinement started (Jiménez-Murcia et al., 2019). A slightly lower mean incidence of dropout was found during the lockdown in comparison with the following months although without reaching statistical significance. The restrictive characteristics of the lockdown has allowed for greater stimulus control, especially related to offline gambling activity. Consequently, the reduction of gambling opportunities, improved social/familiar environments (Donati et al., 2021), and the maintenance of an active treatment adapted to the context might be considered as protective factors (e.g., reassessing the situation and seeking alternative methods of treatment and prevention; Lee and Rovers, 2008; Côté et al., 2020).

Lower levels of anxiety and depressive symptoms due to pandemic and higher socio-economic positions at baseline were stated among the individuals who dropped out during the lockdown. These findings possibly suggest that those patients with better socio-economic background prior to confinement and lower referred psychological impact due to pandemic (in terms of anxiety and depressive symptoms) may have been more likely to discontinue treatment, already from the initial stages of confinement. Also, a lower use of adaptive coping strategies during the lockdown was also described in patients who dropped out in comparison to those who remained in treatment. In contrast, patients with greater insight about their gambling problem and more urgent needs for professional help may have been more likely to remain in treatment, even if the need was not directly related to gambling behavior (i.e., social support, anxiety, and depressive symptoms). Several individuals who remained in treatment reported using non-adaptive strategies during the lockdown (such as concerns related to COVID-19 and gambling behavior). In this sense, we hypothesize that treatment attendance may represent an adaptive mechanism to deal with emotional distress. Hence, treatment should focus on coping with emotions as a helpful approach to tackle gambling problems in the face of similar future adverse circumstances (Estévez et al., 2021).

During the lockdown, a more flexible approach to treatment was employed (Marsden et al., 2020), using Internet-based tools therapeutic strategies (Columb et al., 2020). This approach may have influenced treatment adherence in patients who were already in treatment for GD, from two different perspectives. On the one hand, online therapy was more accessible to people who had previously found it difficult to attend (American Psychological Association, 2015). Moreover, the relationship between therapist and patient in the case of virtual therapy has been described as good as for in-person therapy (Bergman and Kelly, 2021), allowing the therapeutic alliance to be preserved in these exceptional circumstances. On the other hand, the absence of in-person activities, a lack of peer-to-peer social and emotional connections, inequities in Internet access, and online distractions could represent challenges for adapting to shifting from face-to-face to online therapy. These issues have been previously reported in virtual therapeutic approaches in other mental disorders during confinement, such as in substance use disorders and eating disorders (Fernández-Aranda et al., 2020b; Bergman and Kelly, 2021). In the present study, although changes in the treatment approach did not seem to have detrimental effects on treatment adherence, some of the patients who dropped out may have presented some of the described difficulties. Pandemic situation highlights the need to efficiently adapt the available treatment strategies to these exceptional circumstances to ensure therapeutic continuity. In this vein, customizing proven treatment strategies to virtual formats has been a useful tool, preserving the initial therapeutic objectives.

After the lockdown, restrictive measures were progressively reduced (e.g., related to “stay-at-home” measures, social distancing, and diminished mobility) and the treatment program in its original format was reinstated. Dropout was associated with more employment concerns due to the pandemic. In this line, the financial pressure of job insecurity and unemployment could motivate individuals to use gambling as a mechanism to earn money and manage debts (van Schalkwyk et al., 2020; Salerno and Pallanti, 2021). Also, previous financial crises have shown the possibility of engaging in gambling behaviors to attempt economic restitution (Olason et al., 2015; Economou et al., 2019). One plausible explanation for our results could be that those patients reporting work instability and difficult financial situations due to the lockdown could have relapsed and abandoned treatment. Another possible rationale could be that, coinciding with the easing of restrictions after the lockdown, some patients may find it difficult to juggle work and treatment, prioritizing the search and maintenance of a job over treatment. While many of these notions are currently speculative, they warrant direct examination in future studies.

The survival analysis showed that most dropouts were concentrated in the first months of this period, especially between May and July 2020. These months in Spain could be considered as a transition period between the measures taken during the lockdown and the beginning of the gradual phasing out of these measures. It was characterized by an increased uncertainty about the return to daily routines (e.g., work routines), the ubiquity of social contact and health concerns related to COVID-19. The general instability of this phase, along with the reopening of the gambling establishments, made these first months after lockdown a particular period of high vulnerability for relapse and for treatment dropout. Furthermore, some of the patients who abandoned treatment during the first months of confinement were in the early stages of the treatment phase. These results are in line with previous studies which described a predominant tendency to dropout at the start of the treatment or after the first few sessions (Jiménez-Murcia et al., 2015).

Going one step further, mean incidence of relapse for gambling behavior was significantly decreased during and after lockdown in comparison with mean rates prior to the confinement started. However, there were not differences in relapse rates between during and after the lockdown. Most patients reported mainly gambling offline and living with others during the nine-month observational period. A high abstinence from gambling of around 75% has previously been described in patients who were already in CBT treatment and/or in the follow-up phase (Jiménez-Murcia et al., 2007; Cowlishaw et al., 2012). In the context of pandemic, our results also agreed with recent literature that pointed out higher abstinence rate during confinement (Avanzi et al., 2020). In line with Donati et al. (2021), restrictive measures due to pandemic, social support and some policies related to gambling activity adopted during confinement, have promoted gambling abstinence by reducing offline gambling opportunities.

Finally, an association between treatment duration and dropout was found during and after lockdown. Specifically, the shorter the duration of treatment, the higher the risk of dropout during and after the lockdown. Furthermore, and according to the predictive analysis model, the duration of the treatment would be a predictive factor of dropout. Length of therapy has been positively related to treatment outcome in previous studies (Orlinsky et al., 1994). In this sense, achieving a strong therapeutic alliance with the patient early in treatment is important for improving adherence to therapy (Gómez-Peña et al., 2012). Therefore, patients with weaker therapeutic links and lower durations of treatment may have found it more challenging to preserve treatment adherence in the face of similar adverse situations.

The strengths of the study included a clinical sample of patients already engaged with our treatment Unit. Moreover, we used clinical records and assessments to evaluate patients rather than just self-report measures. The longitudinal nature of this work allowed reporting treatment outcomes related to data over the course of 9months. Limitations included the modest sample size, limited geographic location, and the absence of a control group for comparison and multiple qualitative assessments. Some clinical and socio-demographic variables were collected prior to the onset of the lockdown (i.e., at the baseline evaluation prior to treatment) and their possible changes due to the pandemic were updated in the short-term during confinement. However, the influence of new changes in these variables in the middle-long term should be contemplated in future research. Moreover, upcoming studies would benefit from including quantitative validated scales and evaluating changes over longer time periods in GD symptomatology and associated factors, throughout different stages of the COVID-19 pandemic in multiple jurisdictions and cultures.

In conclusion, the present findings provide data on the treatment dropout in clinical population diagnosed with GD and potential contributing factors during confinement. In this line, duration of treatment could be a possible predictor of dropout in the face of adverse external situations. These results have relevance for identifying potentially high-risk subjects and optimizing individualized early interventions in the setting of future similar adverse circumstances.

Individuals may inquire with Dr. Jiménez-Murcia regarding availability of the data as there is ongoing studies using the data. To avoid overlapping research efforts, Dr. Jiménez-Murcia will consider a request on a case-by-case basis.

The present study was conducted in accordance with the latest version of the Declaration of Helsinki. The Clinical Research Ethics Committee of the first author’s Public University Hospital gave approval before confinement, as most of the data used in this study was part of the baseline 308 assessment protocol prior to the start of treatment and previously used in the Unit’s studies related to treatment (Ref: PR329/19). We added the brief telephone survey to this protocol at the beginning of confinement. Signed informed consent was obtained from all participants.

IB, EC, ME, and SJ-M contributed to the development of the study concept and design and aided with interpretation of data and the writing of the manuscript. RG performed the statistical analysis. IB, MG-P, LM, SR, AP-G, BM-M, EV-M, and SJ-M aided with data collection. FF-A, SJ-M, MP, AH, and JM revised the manuscript and provided substantial comments. FF-A and SJ-M obtained funding. All authors contributed to the article and approved the submitted version.

CERCA Programme/Generalitat de Catalunya gave institutional support. This work was additionally supported by a grant from the Ministerio de Ciencia, Innovación y Universidades (grant RTI2018-101837-B-100). The research was funded by the Delegación del Gobierno para el Plan Nacional sobre Drogas (2017I067 and 2019I47), Instituto de Salud Carlos III (ISCIII) (PI17/01167) and co-funded by FEDER funds/European Regional Development Fund (ERDF), a way to build Europe. CIBEROBN and CIBERSAM are both initiatives of ISCIII. IB was partially supported by a Post-Residency Grant from Research Committee of the University Hospital of Bellvitge (HUB; Barcelona, Spain) 2020–2021. The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.

AH holds a position at Lund University, Sweden, sponsored by the state-owned Swedish gambling operator AB Svenska Spel and has funding from the research council of the same organization and from the research councils of the Swedish state-owned alcohol monopoly Systembolaget AB and the Swedish sports federation. MP notes the following disclosures. He has: consulted for and advised Game Day Data, the Addiction Policy Forum, AXA, Idorsia, and Opiant/Lakelight Therapeutics; received research support from the Veteran’s Administration, Mohegan Sun Casino, and the National Center for Responsible Gaming (on the International Center for Responsible Gambling); participated in surveys, mailings, or telephone consultations related to addictions, impulse-control disorders or other health topics; consulted for law offices and the federal public defender’s office in issues related to impulse-control and addictive disorders; provided clinical care in the Connecticut Department of Mental Health and Addiction Services Problem Gambling Services Program; performed grant reviews for the National Institutes of Health and other agencies; edited journals and journal sections; given academic lectures in grand rounds, CME events and other clinical/scientific venues; and generated books or chapters for publishers of mental health texts.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.