Table 1.

Characteristics of studies included in the review

References	Study design	Study population	Race, ethnicity, and SES	Description of intervention	Mode of delivery	Level of support	Outcome measures	Results (effect size if available)
Empirical studies testing virtual care interventions in youth with eating disorders
Aardoom et al. [12, 22]	RCT	N = 273 Mean age = 24.4 (range not reported, over 16) Sex = 99% females Diagnosis = approximated DSM-5 diagnosis using EDE-Q self-report; 97.6% above clinical significance cut-off Recruited = via Featback website of Dutch recovery e-community The Netherlands	NR	Featback, a fully automated Internet-based monitoring and feedback intervention was offered alone and with additional high- and low-intensity online therapist support in comparison to a waiting list control	Online: depending on condition, all online including for therapist support	2 of the 4 conditions include email, chat, and Skype support from therapists	SEED	Featback was superior to waitlist control for bulimic psychopathology (d = 0.16) as well as for ED-related quality of life at 3 month follow up (d = 0.22) No effect for anorexia nervosa psychopathology Additional support significantly enhanced patient experience of treatment but did not impact effectiveness
Aardoom et al. [23]	RCT	N = 273 Mean age = 24.4 (range not reported, over 16) Sex = 99% females Diagnosis = approximated DSM-5 diagnosis using EDE-Q self-report; 97.6% above clinical significance cut-off Recruited = via Featback website of Dutch recovery e-community The Netherlands	NR	Featback, a fully automated Internet-based monitoring and feedback intervention was offered alone and with additional high- and low-intensity online therapist support in comparison to a waiting list control	Online: depending on condition, all online including for therapist support	2 of the 4 conditions include email, chat, and Skype support from therapists	SEED	Featback was superior to usual care for mild/moderate bulimic but not for mild/moderate symptoms of anorexia nervosa Fully automated monitoring and feedback improved mild to moderate bulimic symptoms. Additional therapist support did not increase effectiveness for these patients
Anastasiadou et al. [24]	Multicentre RCT	N = 106 Mean age = 18.06 (SD 6.04) Diagnosis = any ED, according to DSM-5 criteria Recruited from public/private mental health services Spain	NR	Standard CBT + TCApp, a mobile health intervention including monitoring and chat with therapists (12 weeks) compared to standard CBT alone	Hybrid: included both in-person CBT and the online mobile app in between appointments	Online chat support by therapist (therapist responded minimally once per week)	EDE-Q and SEED	There were significant but negligible reductions in EDE-Q (d = 0.11) and SEED scores (AN severity index: d = -0.09; BN severity index d = 0.09) for participants overall No significant difference between the 2 groups on longitudinal EDE-Q total and subscales and in severity of SEED
Anderson et al. [25]	Case series	N = 10 Age 13–18 years (mean = 16.08, SD 1.99) Sex = 8/10 were females Diagnosis = AN or atypical AN, according to DSM-5 criteria Recruited from medical ads USA	Ethnicity: all Caucasian SES, race: NR	Videoconference delivered Family-Based Treatment consisting of weekly family meetings with therapist (20 sessions over 6 months)	Online: intervention delivered completely via telehealth platform	No additional support outside of primary treatment	%mBMI, EDE	Weight increased from baseline to end of treatment (d = 0.53) and to 6 months follow-up Changes in global EDE and subscales were significant at end of treatment (d = 1.06) and follow-up
Carrard et al. [26]	Uncontrolled trial	N = 127 Mean age = 24.7 (18–43 range, SD = 5.1) Sex = female Diagnosis = BN purging type or EDNOS bulimic type, according to DSM-IV criteria Recruited from multiple European treatment centers and ads Switzerland, Spain, Sweden, and Germany	NR	Internet self-help program based on classical CBT principles	Online: all modules and contact with coach via internet. 1 evaluation session done face to face	Weekly email contact with coach	EDI-2, SCL-90R	Severity of eating disorder symptoms and general psychopathology improved significantly. 45% of participants were considered clinically improved and 23% had no symptoms at end of self-treatment program
Fichter et al. [27]	RCT	N = 258 Mean age 23.8 (> 16 years old, SD: 6.5) Sex = female Diagnosis = AN or subthreshold AN without the requirement of amenorrhea according to DSM-IV criteria Recruited from 1 of 8 hospitals in Germany specialized for ED treatment Germany	Ethnicity: NR SES: middle/upper class	Internet-based relapse prevention for AN after inpatient treatment using CBT strategies (9 months) compared to no intervention	Online: all components were online (chapters, electronic messaging, online chat room, automatic electronic messages)	Electronic message board for peer support + monthly 1 h chat session with therapist + therapist available via email at any time within 24 h	BMI	Patients in relapse prevention gained significantly more weight (0.62 BMI points) than patients in TAU condition (0.03 BMI points) (small to medium effect sizes for BMI change in each group from T1 to T2) EDI-2 total score showed greater decrease in TAU group over time from T1 to T2
Fichter et al. [28]	RCT	N = 210 Mean age 23.95, > 16 years old Sex = female Diagnosis = AN or subthreshold AN without the requirement of amenorrhea according to DSM-IV criteria Germany	Ethnicity: NR SES: middle/upper class	Internet-based relapse prevention for AN after inpatient treatment using CBT strategies (9 months) compared to no intervention	Online: all components were online (chapters, electronic messaging, online chat room, automatic electronic messages)	Electronic message board for peer support + monthly 1 h chat session with therapist + therapist available via email at any time within 24 h	BMI	There was an overall increase in BMI after the intervention, with a greater increase in the “full completers” group compared to the “partial completers” and the control group. There was also a significant interaction of group by time
Giel et al. [29]	Uncontrolled trial	N = 16 Mean age 21.7 (SD 4.3) No details on gender Diagnosis = full or atypical AN according to ICD-10 Recruited from clinic after inpatient care when transitioning to outpatient care Germany	NR	Relapse prevention program based on Maudsley Model of Anorexia Nervosa Treatment delivered by videoconference (10 sessions over 4 months)	Hybrid: 2 sessions face to face, 8 other sessions via videoconferencing	No additional support outside of primary treatment	Feasibility/acceptability; BMI, EDE-Q	Large effect size for BMI increase after treatment (r = 0.50), small to medium effect sizes for reduction of ED symptoms and scores (r range = 0.06 to 0.57)
Kim et al. [30]	Uncontrolled trial	N = 22 Sex = not specified Mean age = 25.27 (SD 8.53) Diagnosis = AN confirmed using SCID-5 according to DSM-5 criteria Recruited = from outpatient clinic of Seoul Paik Hospital South Korea	NR	TAU in-person (either MANTRA, specialist supportive clinical management, or FBT) augmented by videos about ED recovery tips + daily text messages + 1 weekly face-to-face guidance meeting	Hybrid: vodcast + text-message support + face-to-face therapy	Daily text message from researcher + weekly face-to-face guidance with researcher	Feasibility/acceptability; EDE-Q	Significant reduction in global EDE-Q score (Δη2 = 0.59) and all subscales. No significant change in BMI
Lock et al. [31]	Multicenter RCT	N = 40 Sex = 34 female (85%) Mean age = 14.88 (range 12–18 years, SD 1.81) Diagnosis = DSM-5 criteria for AN Recruited = through clinics, hospitals, and online advertisements Canada, USA	Ethnicity:85% white, 5% Asian, 20% other SES: majority high income	Comparison of online guided self-help family-based treatment for parents of children with AN and family-based treatment delivered via videoconferencing	Online: self-help modules were completely internet-based. Family-based treatment was delivered fully via videoconferencing	Manualized guidance provided by a “coach” therapist (avg 3.90 h over 4–6 months). For videoconferencing group: no additional support outside of online therapy sessions	%EBW, EDE	In the videoconference arm, participants gained an average of 8.63 percentage points of EBW (Cohen's d = 1.46) and in the guided self-help arm, participants gained an average of 9.97 percentage points (Cohen's d = 1.51) The EDE Global score decreased on average from 2.94 to 1.55 (Cohen's d = 1.04), with roughly equivalent reductions noted across both treatment arms (videoconference: 2.89 to 1.56; Cohen's d = 1.03; guided self-help: 2.99 to 1.54; Cohen's d = 1.01)
Marco et al. [32]	RCT	N = 34 Mean age = 21.82 (range of 15–40, SD 5.75) Sex = females Diagnosis = BN, EDNOS, or AN, according to DSM-IV-TR Recruited from outpatient program for EDs Spain	NR	CBT for EDs (15 sessions) augmented by CBT for body image in EDs using virtual reality (8 psychotherapy sessions with VR techniques)	Hybrid: all components were technically in person, although the VR is a technology that could be done online now	No additional support outside of primary treatment	Body image (BAT, BIATQ, BASS, SIBID, BITE, EAT)	All participants significantly improved body image (η2 = 0.35, corresponding to a large effect size). Effect maintained during 1-year follow-up Those with VR showed more improvement than those without VR
Neumayr et al. [33]	Pilot RCT	N = 40 Sex = female Mean intervention group = 20.75 (range 15–36 years, SD 6.4), mean control group = 18.00 (range 15–30 years, SD 3.73) Diagnosis = primary AN diagnosis, according to ICD-10 criteria Recruited from inpatient treatment centers for EDs Germany	NR	Therapist-guided smartphone-based aftercare (Recovery Record) + TAU vs. TAU alone	Hybrid: therapist-guided use of smartphone app w/in-app therapist feedback + face-to-face therapy (TAU)	Once or twice weekly in-app feedback from therapist	EDE-Q	No significant difference in BMI or EDE-Q (intervention vs. control group) from baseline to postintervention and from baseline to follow-up From baseline to postintervention: nonsignificant between-group differences of moderate effect size on EDE-Q global score (mean difference, intervention group = -0.06; control group = 0.41, d = 0.56)
Sanchez-Ortiz et al. [34]	RCT	N = 76 Sex = 75 females, 1 male Mean age = 23.9 (SD 5.9) Diagnosis = DSM-IV criteria for BN or EDNOS Recruited = from higher education institutions through university email addresses, posters, and pamphlets United Kingdom	Ethnicity: 60% British, 40% other SES: NR	Internet-based CBT with email support vs. waiting-list control	Online: internet CBT + email support	Email support from therapists ever 1–2 weeks	EDE	Differences between the internet CBT and the waitlist/delayed treatment groups increased significantly from baseline to 3 months on two of the three primary outcome variables (i.e., the EDE Global score and binge eating) EDE Global score improved in internet CBT group (d = 1.28) from baseline to end of treatment and to a lesser extent in the waitlist/delayed treatment group (d = 0.52)
Shingleton et al. [35]	Replicated single-case alternating treatment design	N = 12 Sex 10 females (83%) Mean age = 21.5 (SD 2.35) Diagnosis = DSM-5 diagnostic criteria for AN, subclinical AN, or BN with high dietary restraint/restriction Recruited via community and clinician referrals USA	NR	Face-to-face therapy (motivational interview + CBT) augmented by motivational text-messages + smartphone self-monitoring. Each participant underwent a semi-randomized sequence of text message and no text message phases. The phases were 1 week in duration and summed to 4 weeks of receiving text messages and 4 weeks of not receiving text messages	Hybrid: in-person therapy + text messages + smartphone app	Participants received individualized motivational text message at each mealtime	RMQ; EDE-Q	No significant main effects of the text messages on self-reported kilocalorie intake or dietary restraint (operationalized as percentage of EDE-Q questions answered “yes”) when covarying for baseline BMI Significant main effect of the text messages to increase precontemplation scores (Wald chi-square = 17.64) and to increase dietary restraint action scores (Wald chi-square = 14.85)
Wagner et al. [36]	RCT	N = 126 Sex = female Mean age for adolescents = 19.31 (range 16–21, SD 1.77) and for adults mean = 26 (range 22–34, SD 3.34) Diagnosis = DSM-IV criteria for BN purging type, EDNOS with binge-eating or purging behavior Recruited via ads Austria	NR	Internet-based guided self-help program based on CBT strategies (4–7 months) compared to bibliotherapy	Online: programme via internet platform + weekly email support	Weekly emails from therapists guiding program use	EDI-2	No difference in outcome between conditions. Both groups had significant improvement over time with intervention for monthly binge eating, vomiting, and fasting in both groups
Wagner et al. [36]	RCT	N = 126 Sex = females Mean age = 24.55 Diagnosis = DSM-IV criteria for BN purging type, EDNOS with binge-eating or purging behavior Recruited via ads Austria	NR	Internet-based (email) guided self-help vs. guided bibliotherapy	Online: internet-based self-help + weekly email support	Weekly emails from therapist guiding program use	EDI-2	No differences in abstinence and remission rates between the two groups For all EDI-2 scale scores, the scores decreased over time but there were no differences between the two interventions and no interaction
Wagner et al. [37]	RCT	N = 126 Sex = females Mean age = 24.55 Diagnosis = DSM-IV criteria for BN purging type, EDNOS with binge-eating or purging behavior Recruited = via ads Austria	NR	Internet-based (email) guided self-help vs. guided bibliotherapy	Online: internet-based self-help + weekly email support	Weekly emails from therapist guiding program use	BDI; GenPsy SIAB-EX; TCI-R; EDI-2; BNSOCQ	At the end of treatment, higher motivation to change, higher harm-avoidance, lower baseline frequency of binge eating and lower body dissatisfaction EDI-2 scores predicted good outcomes At follow-up, less baseline frequency of binge eating and higher EDI-drive-for-thinness scores were associated with good outcome at follow-up Higher BDI total scores and lower self-directedness scores at baseline predicted drop-out from therapy
Wilksch et al. [38]	Pragmatic RCT	N = 316 (220 met criteria for ED at baseline); Sex = female Mean age = 20.80 (range = 18–25 years, SD 2.26) Diagnosis = AN, BN, BED, or OSFED assessed using EDE-Q Recruited via self-referral by advertisements Australia and New Zealand	NR	Media Smart-Targeted (MS-T), an internet self-help program vs. receiving tips for positive body image	Online: internet-based self-help program without support	No support provided via app	EDE-Q	MS-T participants were 75% less likely than controls to meet ED criteria at 12-month follow-up This was statistically significant amongst both nontreatment seekers and treatment seekers who were 71% and 86% less likely than controls to meet diagnosis at 12-month follow-up, respectively
Observational and mixed-methods studies
Barney et al. [18]	Observational	Set at Adolescent and Young Adult Medicine clinic Age range = 12–26 years Sex = mostly females 1715 unique patients during this time Diagnosis = any adolescent/young adult health concern, including all EDs Info also included providers, clinical support staff, clerical support staff and other health professionals USA	NR	Within 1 week of pandemic, all providers used Zoom for telemedicine. By week 2, daily telemedicine sessions began for urgent care, mental health, ED, and addiction treatments	Online: telemedicine clinics, daily 2 h in-person sessions for special cases only	All providers (e.g., nurses, fellows, physicians) continued their assessments and sessions online	Measured healthcare service usage and qualitative user/provider experience throughout pandemic	Telemedicine visits increased from 0 to 97% in one month, number of visits comparable to year prior. Providers identified several barriers and solutions to providing different types of care: using earphones for privacy, full training on Zoom, patients submitting results to their patient portal, training families to collect weight at home, having weight taken by nurses, etc. These were identified to help with barriers such as patients not having appropriate devices, technology literacy gap with families, lack of connection etc
Carretier et al. [39]	Case study and description of implementation of telemedicine during COVID-19 pandemic	N = 1 (case study) 16-year-old female teenager with restrictive AN Recruited from the Maison des Adolescents where being treated pre-COVID-19 France	NR	Virtual day hospital care, including individual group, and family therapy was provided by videoconference by a multidisciplinary adolescent medicine clinic	Online: all sessions moved online. for critical cases, some in-person appts possible	No additional support outside of multidisciplinary videoconference therapy	Qualitative case description and discussion	The center was able to move online quickly and efficiently. Benefits of telemedicine are discussed: teleconsultations allow to identify potential risk factors (ex. see how Anna deteriorated during the confinement), telemedicine reduces isolation of patients and allows for clinical support at home, the technology involved is mastered by adolescents and they feel comfortable using it, telemedicine can facilitate the discussion of difficult questions and help work on relationships (ex. helped Anna's family create a dialogue on ED)
Criquillion et al. [40]	Proposal for a non-randomized clinical trial	Goal is N = 50 patients Sex not specified Age range = between 12 and 35 years old Recruited from a clinic for EDs, patients (42 patients currently involved in the testing of the app) France	n/a	A psychoeducational smartphone application including real-time follow-up for use in clinical settings is proposed	Hybrid: partially virtual; followed at the medical clinic regularly, the app on the side for continuity	No support via app, although the patients are followed for the rest of their treatment by doctors at the clinic	Evolution of baseline characteristics	n/a
Datta et al. [15]	Observational	Description of patients seen at the clinic: N = 1582 82% female, 17% male, 1% non-binary Mean age = 16.5 (age range up to 25 years, SD 1.72) Diagnosis = mostly AN but included all EDs, as diagnosed by physicians in the specialized ED clinic Recruited from a comprehensive care unit for EDs USA	NR	Several parts of treatment were delivered via videoconference: 1) psychotherapy and medication management 2) therapeutic groups based on CBT, cognitive remediation therapy (CRT), process group (PG) and parent/patient tele-groups 3) new admission evaluations	Online: mostly videoconferencing for therapy session and group work. When necessary, some in-person	All online aspects of treatments were provided by clinicians: psychologists, psychiatrist, and dietitians. No support outside of regular treatment	Qualitative descriptions of the changes made at the treatment center	All aspects of new admissions, psychotherapy, medication management, and therapeutic groups were moved online where they used videoconferencing. benefits/takeaways are these: 1) useful for parents to take part in initial evaluations remotely 2) remote work generates exposures for social anxiety triggers and complaints 3) videoconferencing useful for bed restriction patients 4) setting group rules is possible and helps when conducting remote group sessions
Davis et al. [16]	Observational	No sample population Description of Singapore hospital multidisciplinary ED service for patients 9 to 16 (mean 13.9, SD 1.5) Singapore	NR	For inpatient: individual meal supervision by nurses. For outpatient: telehealth instituted for physicians and psychologists to manage selected cases and for ongoing engagement with families by psychologists	Hybrid: mix of in person and telehealth	No additional support described outside of videoconference therapy	n/a	The paper describes several considerations for use of telemedicine: for patient/family: willing to use telemedicine, weight stable or good consistent progress, parent willing to weigh patient, no safety concerns, no concerns of medical stability. for health care provider: develop patient eligibility criteria for telemedicine, undergo training in telemedicine
Kasson et al. [41]	Mixed method	(a) N = 14 Sex = 100% female Age range = 14- 17 years Diagnosis = clinical or subclinical ED according to DSM-5 criteria, based on self-report questionnaire (excluding clinical AN) Recruited via Instagram & Facebook (b) N = 30 Sex = 13% gender minority, 74% female, 13% male Age range = 14 – 17 years Diagnosis = same as above Recruited via Instagram, Facebook, Snapchat, YouTube, & TikTok USA	(a) 92% White; 8% Other Hispanic 14%; Non-Hispanic 86% 71% sexual minority (b) 74% White; 26% Other 13% Hispanic; 87% Non-Hispanic 57% sexual minority; 63% LGBTQ + group	Mobile phone and desktop application, Space from Body and Eating Concerns Program, based on CBT for EDs	n/a	n/a	Stanford-Washington Eating Disorder Screen	Using social media for recruitment of teens with EDs is feasible and may help in capturing more diverse samples The youth who participated in the survey were highly interested in a mobile app to help with ED recovery
Rajankar et al. [42]	Intervention description	India	n/a	Proposal for a new smartphone application combining awareness of etiology/informational section, health tracker, calorie tracker, virtual intervention (CBT, DBT, Self-help, support groups), and alert function notifying nearby NGOs/helplines	Smartphone app	n/a	n/a	n/a
Shaw [43]	Observational	N = 3 Sex = female Age range = 12–17 years old Diagnosis = AN, as diagnosed by treating physicians in a specialized ED clinic Recruited from an eating disorder treatment service United Kingdom	NR	7-week video-conference art therapy sessions held on Microsoft Teams platform	Online: videoconference	No additional support outside of videoconference therapy	Qualitative description of experience	Challenges when working with this client group online include increased body image concerns changed experience of gaze
Stewart et al. [44]	Mixed methods	N = 53 Age range =  < 18 years old Diagnosis = AN, AAN, BN, ARFID, or other specified feeding and eating disorder, as diagnosed by treating physicians in the specialized ED clinic Recruitment = all clinicians, young people, and parents working or receiving treatment at an outpatient community ED program for youth during data collection were eligible United Kingdom	NR	Experience in response to COVID-19 pandemic changes: all but essential face-to-face contact was ceased, all outpatient treatment was delivered by video/phone calls	Online only (besides essential services)	No additional support outside primary treatment from family therapists, psychiatrists, nurse therapists, clinical psychologists	Questions about clinician, young people, and parent experience of providing/receiving online treatment	Young people (YP) and parents had overall positive experience with online therapy, low level of impact of technology on treatment experience; Qualitative Themes: Gratitude/something lost/something gained; home as a therapeutic space; changes in therapeutic relationship; implications for the future; trade-off between increasing and limiting access by moving online
Raykos et al. [45]	Uncontrolled trial	N = 25 Sex = 93% female Mean age = 24.4 (SD 7.6) Diagnosis = any ED, according to DSM-5 criteria, using EDE & SCID-5 Recruited from patients being followed in an outpatient ED clinic Australia	70% Anglo/European-Australian	Rapid transition from face to face to telehealth care in an outpatient clinic for EDs, in the context of COVID-19. FBT or CBT was delivered via telehealth, based on age and diagnosis	Online: telehealth	No additional support provided outside regular treatment	EDE-Q, ED-15, CIA, BMI,	Throughout the study period, patients experienced a large decrease in ED symptoms, comparable to historic benchmarks at the clinic (from pre- to during COVID-19, EDE-Q d = 1.62) Patients rated the quality of treatment and therapeutic alliance highly
Tregarthen et al. [46]	Observational	N = 108,996 Sex = 57,940 provided gender (87.2% female) Mean age = 22 (reported by 48,830, range = 13–77 years) Diagnosis = symptom severity assessed by EDE-Q, 87.2% were ≥ 1 SDs above community norms Recruitment = app was made available for download in app stores	NR	CBT-based self-monitoring smartphone application	Online: smartphone app	No support provided via app	Acceptability; EDE-Q, BMI at baseline	Acceptability: 2,503 anonymous ratings: 84% were 5/5, 13.3% were 4/5, 1.7% were 3/5, and 1% were 2/5 BMI (calculated for 13,784 users): 10.6% met the BMI suggested weight criterion (< 17.5) for a diagnosis of anorexia nervosa Any occurrence of objective binge episodes 78.8%, subjective binge episodes 76.79%. self-induced vomiting 41.9%, fasting 63.3%, excessive exercise 59.7%, laxative misuse 20.6%
Yaffa et al. [47]	Case series	N = 4 Sex = female Age range = 13–17 years Diagnosis = any ED Recruitment = patients were receiving services at a specialized pediatric ED treatment center in a children’s hospital Israel	NR	Management of pts in the context of COVID-19 pandemic and lockdown	Online: treatment via videoconferencing; in-person visits occurred only in specific conditions	No additional support outside videoconference treatment provided by psychiatrist, clinical nutritionists, psychotherapists	-	The condition of the four adolescents with AN was compromised at the start of the COVID-19 quarantine. The use of multi-disciplinary long- distance telemedicine treatment resulted in an improvement in the condition in three of the four adolescents, living in well-organized families, with the motivation and ability to adjust to the new conditions, but not in one adolescent whose family experienced more problems
Qualitative studies
Anastasiadou et al. [48]	Qualitative study	1) N = 8, health care providers and mobile health experts, mean age 34 (SD 7.21), psychiatry, psychology and nursing, males and females 2) N = 9 participants of RCT Mean age = 15 (SD 0.50) Sex = all females Diagnosis = any ED, according to DSM-5 criteria Recruited from RCT participants [24] Spain	NR	Standard CBT + TCApp, a mobile health intervention including monitoring and chat with therapists (12 weeks)	Hybrid: included both in-person CBT and the online mobile app in between appointments	Online chat support by therapist	Questionnaire assessing services based on effectiveness, user experiences, economic aspects, organizational aspects, sociocultural aspects	Key themes identified by specialists: lack of time, lack of strategic plan, lack of budget, insufficient training, lack of tech support, lack of familiarity and mobile health skills, lack of agreement, insufficient interaction, security/privacy issues. Perceived advantages by patients: ease of use, useful, liked the app design, content appropriate. Perceived disadvantages: problems with design, lack of satisfaction with personalization, some say limited interaction
Brothwood et al. [49]	Qualitative study	14 patients and 19 parents Age range = 12 -18 years Mostly females, one male Diagnosis = AN-R, according to DSM-5 criteria Recruited via day program at Maudsley center for child and adolescent ED in London United Kingdom	NR	During the COVID-19 pandemic, day treatment program based on principles of family therapy was delivered online. Online activities included therapy, groups, meal support, and education support. All kinds of therapy, exercises, worksheets, and coaching for meal support, were provided online	Online: treatment online. Urgent appts offered in person if increased physical or psychiatric risks	Mix of clinician support for groups, therapy, and meals. Some exercises were self-help online	User experience questionnaires	Overall experience was positive (slightly more so for parents), quality partly impacted, parents felt more comfortable, young people reported issues with technology. But majority thought could not have done anything differently. Each component rated by majority as somewhat or very helpful. Compared to face to face, varied response but found less helpful (specifically meal support, family sessions and individual sessions). Young people rated that they want those to be in-person in future but would keep medication reviews or dietetic appts virtually. Qualitative data themes: 1) new discoveries 2) lost in translation 3) best of bad situation
Lindgreen et al. [50]	Qualitative study	N = 41 Mean age = 24.0 (age range = 15–41 years, SD 5.9) Sex: 3 male, 38 female Diagnoses = anorexia nervosa and bulimia, as diagnosed by clinicians at the specialized ED treatment facility Recruited = from a specialized ED treatment facility Denmark	NR	Treatment as usual (for AN: family-based treatment or weekly group/individual sessions; for BN: 10 weekly manualized group sessions followed by an additional group or individual treatment) augmented using Recovery Record (RR) self-monitoring app for eating disorders	Hybrid: in-person sessions augmented by guided self-monitoring smartphone app use	Therapists provided guidance on app usage during in-person therapy sessions	User experience with Recovery Record app	Patient experience with the app depended on its app features, the impact of these features on patients, and their specific app usage. This patient-app interaction affected and was affected by changeable contexts making patients' experiences dynamic
Naccache et al. [51]	Qualitative study	N = 8 Mean age = 15.5 (age range = 12–18, SD 1.07), Sex = all females Diagnosis = AN, according to DSM-5 criteria All enrolled in treatment at Toulouse Teaching Hospital France	NR	Unguided self-help program app: based on psychoeducation, CBT, and motivational interview strategies. Additional emotional management feature and gamification elements	Online: testing only user experience of the smartphone app	No additional support in relation to the app	UEQ and semi-structured interviews	Patient experience of the app was demonstrated with key themes, including finding coping strategies, information, motivational content helpful. Overall, the app was rated positively, although opinions were mixed regarding novelty, dependability and efficiency. Qualitative data showed that patients and clinicians think the psychoeducation portion is important
Sanchez-Ortiz et al. [52]	Qualitative study (recruited as part of an RCT)	Interview: N = 9 Sex = female Mean age = 23.2 (SD 3.5) Diagnoses of BN or EDNOS-BN Questionnaire: N = 31 All participants were drawn from those participating in RCT on internet CBT and were students recruited from university networks [34] United Kingdom	NR	Internet-based CBT program called "Overcoming bulimia online" comprised of immediate iCBT with email support from therapist (4 to 8 sessions)	Online: intervention provided completely online via platform + email	Therapist support via weekly emails	Semi-structured and questionnaires on iCBT treatment experience	Experiences of treatment were summarised in five categories: confidentiality/privacy, flexibility, ease of use, feeling supported, and content of programme. As for impact of treatment, overall positive impression, categorized into: (1) expectations about outcome, (2) effectiveness – reported change in ED symptoms, (3) effectiveness—other changes, and (4) tools for coping in future. As for all RCT participants impressions: struggled to find motivation to continue treatment, knowledge and confidence grew, most useful aspects = weekly emails, problems with access to treatment/technical difficulties, most common suggestion = more support and some face to face*
Shaw et al. [53]	Qualitative	N = 43 participants; 19 parents/carers, 12 members of staff, 12 patients, all < 18 years of age Diagnosis = typical or non-typical AN, BN, or BED recruited from a hospital ED service United Kingdom	NR	Describes experience of adaptations made in response to COVID-19 pandemic	Hybrid: in-person physical examinations with all other appointments online	No additional support outside of primary treatment	Healthcare service usage (hospitalizations, referrals, admissions); user satisfaction, narrative analysis	More referrals accepted in 2020 than 2019, higher % increase for urgent vs. routine consultations (150% vs 129%, respectively); Qualitative results: Using virtual platforms improved ease of access but altered relational experiences
Case studies
Hellner et al. [54]	Case series	N = 2 Female Age = 20 and 15, respectively Diagnosis = AN or atypical AN Recruited by Equip providers through postings USA	NR	Videoconference delivered multidisciplinary treatment based on FBT, CBT, and DBT. Patient met with each provider weekly or more	Online: all done through telehealth and unlimited messages on the telehealth platform	Team support (family therapist, dietitian, peer mentor and family mentor) via chat	Weight gain, EDE-Q short form	Both patients gained at least one pound per week. Scores for EDE-Q decreased by 7–12 point
Duncan et al. [ 55]	Case study	N = 1 Sex = female Age = 14 years Diagnosis = EDNOS Recruited from rural tele-psychological clinic in her community USA	NR	Videoconference delivered therapy, based on cognitive behavioral strategies	Online: all sessions done over videoconferencing platform. 1 appointment was done in person	No additional support outside of primary treatment	n/a	At the end of the 12 sessions of therapy, Mila increased food intake, fell within normal range on growth chart, and scores on CDI no longer elevated
Study protocols*
Anderson et al. [56]	Case series	N = 10 (target) Age range = 13–18 years Sex = no specification Diagnosis = DSM-5 criteria for AN or atypical AN Recruited = by referrals and through community USA	NR	Videoconference delivered Family-Based Treatment consisting of weekly family meetings with therapist (20 sessions over 6 months)	Online: intervention delivered completely via telehealth platform	No additional support outside of primary treatment	Feasibility/Acceptability of FBT via Telemedicine	Protocol only
Barakat et al. [57]	Multi-site three-arm RCT	N = 110 (target) Gender will be reported Age range ≥ 16 years Diagnosis = DSM-5 criteria for BN or OSFED with bulimic behaviors Recruited = from general population via advertisement and referrals from health professional Australia	Income, and cultural background/ethnicity will be reported	Online CBT-based self-help program completed independently vs. completed with therapist guidance (= weekly telemedicine sessions) vs. waitlist control	Online: either online self-help platform or secure videoconference platform	For self-help group with guidance only: weekly 30 min videoconference session with therapist	Frequency of objective binge episodes (EDE-Q)	Protocol only
Bulik et al. [58]	RCT	N = 180 (target) Sex: no specification Age range ≥ 18 years old Diagnosis = criteria for DSM-IV BN Recruited from within local health services and via ads USA	NR	Group CBT delivered via the internet versus face-to-face	Online: online modules with text-based chat group (without audio/video)	Therapist-led chat groups or in-person therapist support in groups	EDE	Protocol only
de Zwaan et al. [59]	Multicenter RCT	N = 178 (target) Age range ≥ 18 years old Diagnosis = DSM-IV criteria for BED or subsyndromal BED Recruited = via ads Germany and Switzerland	NR	Internet-based guided self-help vs. individual in-person CBT	Hybrid: 2 in-person meetings with therapist + internet-based email-guided self help	Therapist support by email	Difference in number of days with OBEs over the past 28 days (assessed by EDE)	Protocol only
Hambleton et al. [60]	Uncontrolled trial	Goal is 41 families Age range = 12–18 years Gender will be reported Diagnosis = criteria for AN according to DSM-5 Recruited from regional or rural health district via referral Australia	SES will be reported	Family-based treatment delivered via videoconference (18 session over 9 months) via telemedicine	Online: via videoconferencing platform (or via telephone if technical difficulties)	No additional support outside of primary treatment	Remission (increase in %mBMI to ≥ 85%)	Protocol only
Jenkins et al. [61]	RCT	N = 51 (target) Sex not specified Age range ≥ 17.5 years Diagnosis = regular binge eating in the context of an ED Recruited = clinical referrals United Kingdom	NR	Self-help delivered face-to-face versus via e-mail versus delayed treatment control condition	Hybrid: first guidance session provided in-person	Therapist support via email, up to twice weekly	Frequency of objective bulimic episodes assessed by EDE-Q	Protocol only
Kolar et al. [62]	RCT	N = 30 (target) Sex = female Age range = 12–19 years Diagnosis = AN Germany	NR	In-person consultations with smartphone app vs. in-person consultations alone	Hybrid: in-person therapy + smartphone app use	Therapist support in-person	Weight gain	Protocol only
Kolar et al. [63]	Multi-center RCT	N = 30 (target) Sex = female Age range = 12–19 years Diagnosis = AN or atypical AN confirmed by EDE structured interview Recruited = from waiting list from 3 child & adolescent psychiatry centers Germany	NR	Face-to-face supportive psychiatric follow-up augmented by therapist-guided use of a DBT-informed smartphone application, (Jourvie) vs. TAU (supportive therapy by psychiatrist)	Hybrid: face-to-face therapy + guided smartphone app use	In-person therapy sessions will include guidance on use of the smartphone app	BMI-SDS, EDI-2	Protocol only
Rohrbach et al. [64]	RCT (two-by-two factorial design with repeated measures)	N = 352 (target) Sex not specified Age range ≥ 16 years Diagnosis = at least mild self-reported ED symptoms on standardized questionnaire Recruitment = community Dutch e-community ED website Netherlands	NR	(1) Featback, a fully automated self-guided Internet-based monitoring and feedback intervention augmented by weekly chat/email support from an expert patient was compared to (2) Featback without expert patient support, (3) expert patient support only, and (4) waiting list	Online: unguided internet psychoeducation with or without chat/email support	Weekly expert patient support via chat and email (in 2/4 conditions), no additional support in 2/4 conditions	EDE-Q	Protocol only
Schlegl et al. [65]	RCT	N = 186 participants (target) Sex = female Age range = 12–60 years Diagnosis = diagnosis of AN at hospital admission, according to DSM-5 criteria Recruited = at an ED clinic Germany	NR	Therapist-guided smartphone-based aftercare (Recovery Record) + TAU vs. TAU alone	Hybrid: therapist-guided use of smartphone app w/in-app therapist feedback + face-to-face therapy (TAU)	Psychotherapists provide individual feedback via-in app messages	EDE	Protocol only
ter Huurne et al. [66]	RCT	N = 252 (target) Age range ≥ 18 year Sex = female Diagnosis = DSM-IV criteria for BN, BED, or EDNOS Recruited = self-selection via targeted ads The Netherlands	NR	CBT-based internet self-help compared to waiting list control group	Online: intervention and therapist support via online treatment platform	Therapist support by email twice weekly; Forum for peer support	EDE-Q	Protocol only

NR not reported, n/a not applicable, ED eating disorder, AN anorexia nervosa, AAN atypical anorexia nervosa, AN-R anorexia nervosa restricting type, ARFID avoidant/restrictive food intake disorder, BN bulimia nervosa, BED binge-eating disorder, EDNOS eating disorder not otherwise specified OSFED other specified feeding or eating disorders, ICD-10 international classification of diseases 10th revision, DSM-IV diagnostic and statistical manual of mental disorders fourth edition, DSM-5 diagnostic and statistical manual of mental disorders fifth edition, SES socioeconomic status, CBT cognitive behavioral therapy, DBT dialectical behavior therapy, RCT randomized controlled trial, TAU treatment as usual, SCID-5: SEED Short evaluation of eating disorders, %mBMI Percent median BMI, EDE-Q eating disorder examination-questionnaire, EDE eating disorder examination, EDI-2 eating disorder inventory-2, SCL-90R symptom checklist-90-revised instrument, BMI body mass index, %EBW percent expected mean body weight, BAT body attitude test, BIATQ body image automatic thoughts questionnaire, BASS body areas satisfaction scale, SIBID situational inventory of body image dysphoria, BITE bulimic investigatory test, EAT eating attitudes test, BDI Beck depression inventory, GenPsy of SIAB-EX general psychopathology of the structured interview for anorexic and bulimic disorders, TCI-R temperament and character inventory revised, BNSOCQ bulimia nervosa stages of change questionnaire, UEQ user experience questionnaire, BMI body mass index, USA United States of America

^*Note that we included these study protocols on the basis that they had the potential of having a sample mean age that fit our inclusion criteria considering their target population