Table 2.

Description of the included studies

Author	Participants	Measures	Intervention Group	Comparison Group	Outcome
Bioulac et al. (2020)	N = 48 10 girls 38 boys, 7–11 years (M = 8.9, SD = 1.2 years), diagnosed with ADHD according to the DSM-IV Exclusion criteria: children with comorbid conditions and cognitive functioning < 85. No information concerning accompanying/previous treatment	ADHD-RS Symptom Inventory; Continuous Performance Test (CPT); Virtual Classroom Assessment (measuring the number of correct hits, and the number of commission errors)	Participants wore an HMD which displayed a virtual classroom. Participants completed a letter identification task whilst experiencing visual and audio distractors n = 16. No. of sessions = 12. Duration of each session = 30 min. Frequency of sessions = twice a week. Intervention period = 6 weeks. Drop-outs: 3	Psychotherapy Intervention: Participants underwent individual psychotherapy, focusing on primary ADHD, emotional symptoms of ADHD, and the impact of ADHD on daily life (notably on self-esteem) n = 16. No. of sessions = 12. Duration of each session = 30 min. Frequency of sessions = twice a week. Intervention period = 6 weeks. Drop-outs: 2	ADHD-RS Symptom Inventory There were significant differences between the groups on ADHD-RS total, ADHD-RS inattention, and ADHD-RS hyperactivity, from baseline to post-treatment. The psychostimulant treatment group exhibited significantly lower ADHD-RS total, ADHD-RS inattention, and ADHD-RS hyperactivity than the other two groups CPT There were significant differences for all groups on the number of omission errors and commission errors from baseline to post-treatment. Post-treatment there was no significant differences in the number of omission errors between the VR-based group and the psychostimulant group. Post-treatment the number of commission errors in the psychostimulant treatment group was significantly higher than in the VR-based group
Bioulac et al. (2020) cont				Psychostimulant Treatment: Participants were treated with long-acting methylphenidate (QUASYM®) and completed a clinical interview once every 2 weeks for 8 weeks with a posologist. The posology of the long-acting methylphenidate was adapted according to the clinical response n = 19. Drop-outs: 4	Virtual Classroom Assessment The number of correct hits post-treatment was significantly higher for the VR-based intervention than the psychotherapy treatment group, but not significantly higher than the psychostimulant treatment group. Post-treatment, the number of commissions in the VR-based intervention group was significantly lower than the psychostimulant group, but not significantly lower than the psychotherapy group
Cho et al. (2002)	N = 50, ages 14–18 years, with a history of learning difficulties, inattentiveness, impulsivity, and hyperactivity. Other demographic information not reported. No information concerning accompanying/previous treatment	Continuous Performance Test (CPT)	Experimental Group 1: Cognitive Training Participants underwent two cognitive training courses utilising an HMD: Virtual Reality Comparison Training (VRCT) and Virtual Reality Sustained Attention Training (VRST). During the VRCT course, participants viewed two 3D objects and decided whether the objects were identical or not. During the VRST course, participants viewed an Arabic numeral and were encouraged to respond by pressing a mouse button when the numeral ‘0’ was presented after any digit other than ‘8’	Group 1: Cognitive Training Participants performed the same task as Experimental Group 1 using only a computer monitor Group 2: Neurofeedback Training Participants performed the same task as Experimental Group 2 using only a computer monitor	Both the experimental and control groups showed a significant increase in the number of correct responses on the CPT post-treatment. Post-treatment the neurofeedback experimental and control groups scored high in the number of correct responses on the CPT than the cognitive training experimental and control groups, although this difference was not significant. The response sensitivity of both the experimental groups improved slightly more than the control groups, although this difference was not significant
Cho et al. (2002) cont			Experimental Group 2: Neurofeedback Training Participants underwent the same training outlines in Lee et al. (2001) n for experimental group 1 = 10. n for experimental group 2 = 10. No. of sessions = 8. Duration of each session = 20 min. Drop-outs: 0	Control Group 3: Participants underwent no training during the same period. No other information was reported n for Control group 1 = 10. n for Control group 2 = 10. n for control group 3 = 10. No. of sessions = 8. Duration of each session = 20 min. Drop-outs: 0
Cho et al. (2004)	N = 28, all were boys, ages 14–18, with no official ADHD diagnosis Subjects had a history of learning difficulties, inattentiveness, impulsivity, and hyperactivity. No information concerning accompanying/previous treatment	Continuous Performance Test (CPT)	Participants underwent the same intervention outlined in Lee et al. (2001) and Cho et al. (2002) (experimental group 2: neurofeedback training) n = 10. No. of sessions = 8. Duration of each session = 20 min. Intervention period = 2 weeks. Drop-outs: 0	Non-immersive-VR Group: Completed the same task as the VR-based intervention group, with only a computer monitor n = 9. Drop-outs: 0 Control Group: Participants received no intervention n = 9. Drop-outs: 0	At post-intervention, the VR-based intervention group had a significantly higher number of hits on the CPT than both control groups. At post-intervention, the VR-based intervention group was marginally significantly faster to react in the CPT than both control groups. All groups showed a decrease in perceptual sensitivity and a reduction in commission errors at post-intervention, and although the VR group demonstrated the lowest scores for both measures this difference was not significant. All groups demonstrated a reduction of omission errors at post-intervention; however, the omission errors of the VR group were significantly lower than both control groups
Kim et al. (2020)	N = 40, 5 girls and 35 boys, ages 8–10 years, with ADHD diagnosed by a psychiatrist. Included only children that were not using medication or other treatments during the intervention/game	Advanced test of attention (ATA); Interactive metronome (IM)	Participants wore a HoloLens (this is a self-contained mobile HMD), and then participate in a series of eye-contact training games. The first five sessions, the second five sessions, and the third five sessions had different levels of difficulty ranging from easy to difficult n = 20. No. of sessions = 15. Duration of sessions = 30 min. Intervention period = 6 weeks. Drop-outs: 0	No intervention was received. n = 20. Drop-outs: 0	ATA The omission and commission errors of the VR-based intervention group decreased significantly from pre- to post-intervention, but those of the control group exhibited a small decrease that was not significant. The omission errors of the VR-based intervention group at post-test were significantly lower than the control group, however, the commission errors of the VR-based intervention group were not significantly lower than the control group IM The mean response times decreased significantly for both groups from pre- to post-intervention. At post-intervention, the VR-based intervention group were significantly faster to respond than the control group in the hand, feet, both hands, both feet, right side, and bilateral cases
Lee et al. (2001)	N = 20, all participants were boys with impulsivity and attention problems. Other demographic information not reported. No information concerning accompanying/previous treatment	Continuous Performance Test (CPT)	Participants wore a head-mounted display (HMD) displaying a dinosaur egg on a desk. Electrodes behind the participant’s ears collected biofeedback data. The dinosaur egg fractured, and participants were instructed to reassemble the egg. The participants were also presented with information about the dinosaur and were asked multiple-choice questions about the dinosaur n = 10. No. of sessions = 10. Duration of each session = 10 min. Intervention period = 2 weeks. Drop-outs: 0	No intervention was received. No other information is reported. n = 10. Drop-outs: 0	The VR-based intervention group showed improvements in perceptual sensitivity from pre- to post-intervention compared to the control group. The authors do not indicate whether this improvement was statistically significant. Both the experimental and control groups showed reductions in omission and commission errors from pre- to post-intervention. The VR-based intervention group showed a larger decrease in commission and omission errors post-test in comparison with the control group, however, the authors do not indicate whether this difference was statistically significant
Skalski et al. (2021)	N = 87, 12 girls and74 boys, 9–15 years (M = 12.75, SD = 1.7 years), diagnosed with ADHD according to the DSM-V Exclusion criteria: neurological condition, intellectual disability, fast wave amplitude within the norm (< 30 μV in the 18–30 Hz band) in the quantitative EEG (QEEG)	The Short Form of the Mackworth Clock Task (SFMCT); The Visual Search Task (VST); The Multitasking Test (MT)	Participants were randomly distributed into two intervention groups: 3D VR hemoencephalographic biofeedback (HEG BFB) with and without distractors. In VR participants were transferred into a virtual room fitted with a Computer, furniture, paintings, plants and windows and observed the flash game image on a computer screen located in the virtual room. In the distractor condition participants were immersed directly into the environment, while in the no distractor condition the participants were immersed into the environment with limited objects available. HMD was a HTC VIVE n = 57. No. of sessions = 10. Length of sessions = 45 min. Intervention period = 10 weeks. Drop-outs: 3	Participants from the traditional 2D HEG BFB training Sessions (HEG BFB standard) group observed flash games on a 21- Inch TV screen. The room was additionally fitted with furniture, paintings, plants and windows. n = 30. No. of sessions = 10. Length of sessions = 45 min. Intervention period = 10 weeks. Drop-outs: 0	Significant improvements in favour of the VR with and distractor groups versus traditional 2D HEG BFB on all outcomes: omission and commission errors (measured with SFMCT), RT slope in visual search (VST), RT in single task and multi-task (MT)
Tabrizi et al. (2020)	N = 48, 16 girls and 32 boys, ages 7–12 years, with ADHD diagnosed by a psychiatrist. Participants had an IQ > 81. All participants were instructed to stop medication for 3 weeks before the study commenced. Included children that stopped medication at least 3 weeks before intervention	The digit span subscale, which is a subset of the verbal scale of Wechsler Children's Intelligence Test	Participants wore an HMD and completed a series of tasks within a virtual classroom with increasing difficulty (the number of targets and distractors increased as the session progressed) n = 16. No. of sessions = 10. Length of sessions = 3 min. Drop-outs: 0	Medication Group Participants received treatment with medicines, such as Ritalin, atomoxetine, and dexamphetamine. No. of participants = 16 Control group No intervention was received; participants did not take medication. n = 16. Drop-outs: 0	Participants in the VR-based intervention and medication group made significant improvements in memory from baseline to post-intervention. At post-intervention, the VR-based intervention scored significantly higher on the digit span subscale than both control groups. At post-intervention, the medication group scored significantly higher on the digit span subscale than the control group. At follow-up, the VR-based intervention scored significantly higher on the digit span subscale than both control groups. At follow-up, the medication group scored significantly higher on the digit span subscale than the control group

N the total number of participants, n number of participants in a fraction of the total sample, M average mean, SD standard deviation