Diagnosis of GD and IGD

The DSM-5-IGD criteria include preoccupation with gaming, withdrawal symptoms, tolerance, unsuccessful attempts to quit gaming, excessive gaming despite negative consequences, loss of interest, deceiving, escape, and functional impairment (American Psychiatric Association, 2013). The diagnosis is established when individuals fulfill five or more of these criteria during 12 months; a previous clinical study reported the satisfactory validity of the DSM-5-IGD (Ko et al., 2014). However, two criteria—escape and deceiving—were determined to have controversial validity (Ko et al., 2014). Furthermore, two criteria—tolerance and withdrawal—are not defined biologically in the DSM-5 (Billieux, Schimmenti, Khazaal, Maurage, & Heeren, 2015; Kaptsis, King, Delfabbro, & Gradisar, 2016). Moreover, preoccupation and tolerance could be experienced by engaged gamers with a harmonious passion that usually did not impair their function (Billieux et al., 2017). Thus, functional impairment should be a required feature in diagnosing GD.

In the ICD-11-GD criteria, GD is defined as a dysfunctional pattern of gaming characterized by impaired control, an increased priority given to gaming over other interests and daily activities, excessive gaming despite its negative consequences, and functional impairment (World Health Organization, 2019). Despite differences in the criteria between the DSM-5-IGD and ICD-11-GD, their diagnostic accuracy (DA) for GD was 91.6% in a clinical study (Higuchi, Nakayama, Matsuzaki, Mihara, & Kitayuguchi, 2021).

The ICD-11 describes HG as a pattern of gaming behavior that increases the risks of harmful physical or mental health consequences. Ko, Lin, Lin, and Yen (2020) reported that 63.8% and 36.2% of individuals with DSM-5-IGD fulfilled the ICD-11-GD and HG criteria, respectively (Ko et al., 2020). However, no study has evaluated the DA of the criteria for HG and compared it with those for DSM-5-IGD based on their severity.

Functional impairment and health problems in GD

Kardefelt-Winther et al. (2017) suggested functional impairment with the repeated, and chronic course should be revealed in diagnosing GD. Krossbakken, Pallesen, Molde, Mentzoni, and Finserås (2017) reported that functional impairment is a key criterion for differentiating addictive gamers from engaged gamers. However, the relationship between functional impairment and GD remains unclear (Przybylski, Weinstein, & Murayama, 2017). Ko et al. (2020) reported that 89.9% of patients with DSM-5-IGD diagnosis had a functional impairment, and most of them experienced health problems. Higuchi et al. (2021) demonstrated impairment in the daily activities and academic and job performance of students with GD. The authors suggested examining the gaming pattern of individuals whose GD persisted in their adulthood. Thus, a tool must be developed to evaluate functional impairment and health problems and their severity in adults with GD.

Insights into GD

Kardefelt-Winther et al. (2017) reported that functional impairment could be a consequence of a deliberate choice. However, the perception of functional impairment might depend on the individual being aware of their disorder (i.e., the insight stage) or having the motivation to change. The transtheoretical model posits that an addictive behavior change involves progress through the following six stages: pre-contemplation, contemplation, determination, action, maintenance, and relapse or termination (Vilela, Jungerman, Laranjeira, & Callaghan, 2009). According to a study on smoking behavior, individuals with advanced stages take more process of changes to quit their smoking. Further, the effective intervention should be determined based on their insight stage (Prochaska & DiClemente, 1983). Thus, it is essential to understand the stage of individuals with GD when implementing a treatment. However, no diagnostic interview study has explored the insights of individuals with GD.

We hypothesized that: (1) there is a consistency between DSM-5-IGD and ICD-11-GD criteria; (2) HG could compensate for the high threshold of DSM-5-IGD and ICD-11-GD criteria; (3) Individuals with GD experience significant functional impairment and is at an early stage of change with lower insight. Thus, this study investigated the consistency between the ICD-11-GD and the DSM-5- IGD criteria, evaluated the functional impairment and health problems of individuals with GD or HG, and determined the awareness of individuals regarding their GD.

Participants

We recruited participants by posting an advertisement on the bulletin board system of a university. Adults aged between 20 and 40 years with completed upper secondary education were eligible for inclusion in this study. An experienced psychiatrist, the corresponding author, diagnoses GD or HG using a semi-structural interview based on ICD-11 criteria (Appendix 1). The IGD diagnosis was determined by interviewing schedule based on DSM-5- IGD criteria (Ko et al., 2020). In addition, the Chinese version of the Mini International Neuropsychiatric Interview (Sheehan et al., 1998) was used to evaluate the psychiatric comorbidity and exclude prospective participants with psychotic disorders, bipolar I disorder, or substance use disorder. Furthermore, an interview was conducted to exclude participants with intellectual disabilities or brain injury based on tracing their medical and schooling history.

The participants were divided into a GD, HG, or control group according to a psychiatric semi-structural diagnostic interviewing (Appendix 1) based on the ICD-11 criteria. Each participant with GD was matched with two sex-matched and age-matched (±3 years) controls. Participants who engaged in excessive gaming but did not fulfill the diagnostic criteria for GD were included in the HG group.

The GD, HG, and control groups had 60, 45, and 120 participants; all gave their written informed consent. This study was approved by the Institutional Review Board of Kaohsiung Medical University Hospital, Taiwan.

Measures

Procedure

We examined the participants' functional impairments, health problems, and stage of insight. In addition, the participants were asked to complete the CGI, BGNCS, and CIAS-gaming version (CIAS-G).

Statistical analysis

We performed a chi-square test to analyze the association of diagnostic criteria for GD between the ICD-11 and DSM-5. An analysis of variance with a post hoc test was performed to evaluate differences in age, educational levels, and scores in the CGI, CIAS-G, and BGNCS among the IGD, HG, and control groups and the differences in functional impairment and the stage of change between the GD and HG groups. A Pearson correlation analysis was performed to examine the correlation between the scores of the CGI, CIAS-G, and BGNCS and the dimensions of functional impairment. The receiver operating characteristic (ROC) curve was created by plotting the true positive rate (TPR) against the false positive rate (FPR) at various threshold settings of BGNCS to determine its diagnostic ability. A P-value of < 0.05 was considered statistically significant. All statistical analyses were performed using SPSS version 26 (IBM Corp. Released, 2019).

Ethics

All relevant ethical safeguards have been met in relation to subject protection and we have complied with the Declaration of Helsinki. All participants had completed the informed consent. This study was approved by the Institutional Review Board of Kaohsiung Medical University Hospital, Taiwan.

Association between the ICD-11-GD and DSM-5-IGD criteria

The DA (True positive + True negative/All) of the DSM-5-IGD criteria for GD was 91.5% (The lower part of Table 2). A total of 57 participants fulfilled both the ICD-11-GD (95.0%) and DSM-5-IGD (78.1%) criteria. Furthermore, three (5.0%) participants fulfilling ICD-11-GD criteria were included in the control group (2.0%) by DSM-5-IGD criteria. On the other hand, sixteen (21.9%) participants fulfilling DSM-5-IGD criteria were included in the HG group (35.6%) by ICD-11-GD criteria. The kappa value between the ICD-11-GD and DSM-5-IGD criteria was 0.80 (P < 0.001).

Sensitivity (Sen.) and Specificity (Spe.) of each criterion for ICD-11-GD diagnosis

The diagnostic criteria of ICD-11-GD, namely impaired control over gaming habits, increased priority given to gaming to the extent that gaming tasks take precedence over other activities, continued gaming despite negative consequences, and functional impairment, had a Sen. of 100% (first part of Table 2). They also have adequate specificity.

The results of the chi-square analysis indicated a significant association between the criteria item of DSM-5-IGD and ICD-11-GD diagnosis. We observed adequate Sen., Spe., and DA of the DSM-5-IGD criteria items for ICD-11-GD diagnosis (second part of Table 2), namely preoccupation (Sen.: 78.3%, Spe.: 90.9%, and DA: 87.6%), withdrawal (Sen.: 80.0%, Spe.: 83.6%, and DA: 82.6%), loss of interest (Sen.: 76.7%, Spe.: 84.2%, and DA: 82.2%), and escapism (Sen.: 73.3%, Spe.: 84.8%, and DA: 81.8%). Furthermore, we noted a lower Sen. of 61.7% with adequate Spe. (93.3%) and DA (84.8%) for tolerance. The criterion of deceiving had a lower Sen. of 50.0% and a lower DA of 76.9% for GD diagnosis, despite its adequate Spe. of 86.7%.

Sen. and Spe. of each criterion for DSM-5-IGD (Supplement data 1)

The results indicated adequate Sen., Spe., and DA of the DSM-5 diagnostic criteria for IGD: preoccupation, withdrawal, loss of interest, escapism, loss of control, continued gaming despite negative consequences, and functional impairment. A lower Sen. was noted in criteria of Tolerance (Sen.: 61.6%, Spe.: 98.0%, and DA: 86.2%) and deceiving (Sen.: 52.1%, Spe.: 90.8%, and DA: 78.2%).

Functional impairment and health problems of GD and HG in the ICD-11

The analysis of variance results demonstrated that the GD group had higher scores in the CIAS-G (F = 180.15, P < 0.001), CGI (F = 761.27, P < 0.001), and BGNCS (F = 80.49, P < 0.001) than did the HG group, whereas the HG group had a higher score in Tukey's post hoc analysis than did the control group (Second part of Table 1). We noted that GD group experienced health problems (96.7%), functional impairment in the dimensions of career (73.3%), social interaction (61.6%), academic performance (36.7%), and job performance (35%; First part in Table 3). Furthermore, 73.3% of the participants in the GD group experienced impairment in three or more dimensions of impairments or health problems. The HG group also experienced health problems (86.7%), functional impairment in career (24.4%), social interaction (22.2%), academic performance (17.8%), and job performance (11.1%; First part in Table 3). Furthermore, 11.1% of the HG group participants experienced three or more dimensions of impairments or health problems. Lastly, the BGNCS was correlated with the severity of IGD and the dimensions of impairment in the GD group (Table 4).

Insight of GD and HG group

The self-reported data indicated that a proportion of the GD group considered reducing or controlling their gaming (83.3%), considered quitting their gaming (33.3%), had a discussion with their friends regarding leaving gaming (25.0%), decided to continue gaming as before even if they could turn back time (70.0%; the second part of Table 3), and had the motivation to undergo treatment or change their excessive gaming behavior (71.6%; the third part of Table 3). The participants in the GD group were in their precontemplation (25.0%), contemplation (61.7%), preparation (10%), and action stages (3.3%) based on interviewing (second part of Table 3).

We also evaluate the self-reported insight among the HG group. The proportion of them considered reducing or controlling their gaming (80.0%), considered quitting their gaming (28.9%), had a discussion with their friends regarding leaving gaming (24.5%), or decided to continue gaming as before even if they could turn back time (64.4%) are similar to GD group.

Cut-off point of the BGNCS for differentiating HG from IGD

In the receiver operating characteristic (ROC) analysis of the BGNCS scores, the areas under the ROC curve were 91.8% for diagnosing GD (GD group vs. control group; Fig. 1A) and 87.3% for diagnosing HG (HG and GD groups vs. control group; Fig. 1B). According to the results, a score of ≥11 had the highest DA (True positive and True negative/All) of 87.8% for GD with a kappa value of 0.72. Furthermore, a score of ≥9 was the best cut-off point for screening those with functional impairment in both the HG and GD groups (80% Sen. and 75.8% Spe.). This score can be used as a threshold for identifying individuals with gaming-related problems.

View Full Size

The ROC analysis for diagnostic accuracy of Brief Gaming Negative Consequence Scale (BGNCS) in diagnosing gaming disorder or hazardous gaming from controls

Citation: Journal of Behavioral Addictions 11, 4; 10.1556/2006.2022.00079

• Download Figure
• Download figure as PowerPoint slide

Comparison between the ICD-11-GD and DSM-5-IGD criteria

This study compared the DSM-5-IGD and ICD-11-GD criteria. The kappa value of 80% indicated adequate consistency between the DSM-5-IGD and ICD-11-GD criteria. This result agrees with that reported by Higuchi et al. (2021). In addition, this study evaluated the ICD-11 diagnostic criteria for HG. A total of 16 participants who fulfilled the DSM-5-IGD criteria were diagnosed as having HG. Among them, 10 participants did not meet the criterion of increased priority, 5 did not meet the criterion of impaired control, 1 did not fulfill the negative consequence, and three did not meet the criterion of functional impairment. Thus, the ICD-11-GD had a higher threshold, resulting in excluded individuals without functional impairment. Furthermore, because all the requirements should be fulfilled, those with no typical presentations, such as increased priority given to gaming and impaired control, could be excluded. However, those with no specific presentations experience gaming-associated academic, occupational, career, social, and health problems will be also excluded. Thus, they are diagnosed with HG even if they do not fulfill the ICD-11-GD criteria This finding supports the hypothesis of a previous study (Ko et al., 2020) that HG compensates for the false negative results of the ICD-11-GD criteria.

In our study, three participants with ICD-11-GD and 29 with HG did not fulfill DSM-5-IGD criteria as inadequate criteria (<5). However, all of them experienced at least one gaming-related problem, such as health problems (86.7%). Thus, individuals with gaming-related problems could lose the chance to get help under DSM-5-IGD criteria. Determining gaming-related problems and implementing interventions, such as improving sleep hygiene and performing the exercise, are necessary to prevent HG progression to GD. This result indicates that the ICD-11 criteria for HG can be used to identify individuals with a negative consequence or functional impairment whom the DSM-5 criteria for IGD could miss.

The ICD-11-GD criteria have relatively fewer criteria than the DSM-5-IGD criteria. This might save time, but limit the assessment for GD. Some IGD criteria, such as preoccupation with gaming, withdrawal symptoms such as irritability when Internet gaming is taken away, and loss of interest exhibited high DA for GD. The withdrawal symptoms might contribute to impaired control in gaming. The presentation of "preoccupation with gaming" and "loss of interest" can be similar to "increased priority given to gaming over other life interests and daily activities." Although they are not included as diagnostic criteria for GD, these symptoms can be used to identify individuals engaging in uncontrolled gaming. When a diagnosis cannot be confirmed based on the ICD-11-GD criteria, these symptoms can be assessed to contribute to a final diagnosis.

Addiction and engagement in gaming are distinct from each other but are correlated (Deleuze, Long, Liu, Maurage, & Billieux, 2018). Vallerand et al. revealed that obsessive passion contributes to an internal pressure to engage in activity and causes persistent negative consequences. On the other hand, harmonious passion leads to choosing to engage it and promote healthy adaption (Vallerand et al., 2003). However, it is a challenge to differentiate GD from engagement in gaming with passion. The Delphi study of Castro-Calvo et al. (2021) suggest tolerance, deceiving, and escapism in DSM-5-IGD criteria are limited in validity in diagnosing GD from high engagement (but non-problematic) gaming. In line with this suggestion, this presenting study demonstrated that tolerance and deceiving had lower sensitivity. These two criteria need to be modified or excluded in revising criteria in the DSM system. On the other hand, functional impairment is required for ICD-11-GD criteria. It also emphasized “high engagement for developing skills” in "boundaries with normality" to prevent overdiagnosis based on only heavy gaming. Thus, experts in a Delphi study agree that ICD-11-GD criteria are likely to diagnose GD without pathologizing healthy gaming (Castro-Calvo et al., 2021).

Functional impairment and health problems

Health problems were the most frequently experienced problems in both the GD (96.7%) and HG (86.7%) groups. We observed that 51.7% of the participants in the GD group engaged in excessive gaming for more than ten years and were thus more likely to experience physical complications. A disrupted sleep cycle (65%) and immobilization (65%, for ≥4 h in ≥3 days per week) were the most frequently experienced health-related problems, followed by inadequate sleep (36.7%; <4 h in ≥3 days per week). This result is compatible with a previous study demonstrating that obesity, delayed sleep phase disorder, and insomnia were the common complications of IGD (Ko et al., 2020). Thus, sleep problems should be routinely evaluated among individuals with GD. Moreover, sleep hygiene, behavior, or medical intervention should be provided. Sleep hygiene can prevent such individuals from gaming all night, helping them get adequate sleep and improving their everyday functioning.

To investigate the chronic effect of GD, this study recruited participants aged between 20 and 40 years. In this study, 51.6% of the participants in the GD group engaged in excessive gaming for more than ten years. Thus, only 31.7% of the participants were school students. Furthermore, 26.7% and 13.3% of the nonstudent participants in the GD group had regular and irregular jobs, respectively. However, 35% of the participants in the GD group experienced functional impairment in job performance due to excessive gaming. They experienced fatigue at work because of inadequate sleep or used a smartphone for gaming at work. In addition, 73.3% of the participants in the GD group experienced functional impairment in their careers due to excessive gaming. They did not have a job or lacked the motivation to get a job because of being occupied with gaming (38.3%). Thus, 45% perceived their life as falling below their expectations because of gaming. Therefore, GD group had a higher self-reported BGNCS score than did the HG and control groups. This study provided empirical evidence for functional impairment under chronic GD.

Assessment of negative consequences

In this study, the negative consequences of gaming were subjectively assessed using the BGNCS. This is the first scale developed to examine the negative consequences of GD. This scale did not assess the symptoms of addiction, which might exaggerate the resistance of clinical cases. The correlation of the BGNCS with the CIAS-G and functional impairment dimensions demonstrated the BGNCS' clinical utility for assessing the severity of GD. Evaluating these gaming-related problems using brief self-reported questionnaires, such as the BGNCS, could provide information on their functional impairment. The BGNCS had a satisfactory DA of 91.8% and 87.3% in the ROC analysis for differentiating GD or HG from controls. Individuals with a score of ≥11 should be further evaluated for GD. A score of ≥9 can be used as a threshold for identifying those with gaming-related problems. The BGNCS can be used to identify the negative consequences of gaming and to further arrive at a diagnosis of GD and HG, and design appropriate interventions.

Realization of HG and GD in the insight stage

Based on the stages involved in changing an individual's smoking behavior, increasing the individual's consciousness is the first step to enhancing their motivation to change (Prochaska & DiClemente, 1983). According to self-reported functional impairment, most of the participants in the GD group exhibited concern toward their gaming behavior. The second stage of behavior change is self-reevaluation which involves determining how one thinks about oneself concerning the problem behavior (Prochaska & DiClemente, 1983). In our study, 83.3% of the participants in the GD group demonstrated the self-reevaluation stage and considered reducing or controlling their gaming time. Self-liberation is the next step that involves making a commitment to act (Andersen, 2007). However, only one-third of the participants considered quitting gaming. Building a helpful relationship is critical for initial action (Prochaska & DiClemente, 1983). Only one-fourth of the participants had ever discussed quitting gaming with their friends. Approximately two in three participants in the GD group preferred continuing gaming even if they could turn back time. This result demonstrated their ambivalence in making a decision or taking action to change. The HG group demonstrated a similar insight in self-reported data. This result might indicate the same ambivalence among gamers with a risk of getting gaming-related problems.

In the interview, only 13.3% of the participants in the GD group had prepared or implemented measures to change their gaming behavior. Our result suggested that most chronic cases of IGD had not prepared to change their gaming behavior and need help from mental health professionals.

Because most of the participants in the GD group were at the early stage of change and had not prepared for change in their gaming behavior, the first step of intervention for a chronic case of GD should be a motivational interview that can help them move from the pre-contemplation to the preparation stage. Discussing the individual's life vision, the harmful consequences of gaming, the barriers impeding change, and the benefits of change are essential in this stage. The BGNCS can be utilized to help them appreciate the negative consequences of gaming. Assessing functional impairment and health problems can help identify individuals with GD requiring intervention. A problem-oriented intervention for negative consequences can be an initial step in helping patients. The awareness of these problems helps them transition to the preparation stage where they develop a commitment to controlling their gaming.

Limitations

This study has several limitations that should be addressed. First, the evaluation was solely based on the participants' responses to a psychiatric interview. Additional information should be gathered from other individuals, such as parents or partners. Second, recall bias could not be excluded without the direct observation of some unhealthy behaviors. Third, the difficulty in recruiting community cases limited the study's sample size, which is relatively small for comprehensively assessing the validity of the criteria or scale. Further studies to recruit a larger sample are necessary to support the results.