Table 1 Study characteristics, methodological details, and risk of bias assessment for the 14 studies included in this review

Christensen, 2013 [37]

RCT

Australia

Callers to Lifeline, a charitable 24 h telephone counselling service providing crisis support and suicide prevention. Persons who were highly distressed and/or suicidal at intake were excluded.

MoodGYM: Internet self -guided psychoeducation program coupled with self-guided iCBT consis-ting of six modules delivered once a week over a six week period.

Participants were allocated to a wait-list control condition.

Suicidal ideation: scores on four suicidal ideation items from the 28-item General Health Questionnaire (GHQ–28).

Neither participants nor clinical personnel were blind to treatment allocation. Proportion completing treatment not stated.

Franklin, 2016 [50]

Recruited from online forums for self-injury. Participants had to self-report two or more episodes of self-cutting in past month to be eligible for participation.

Therapeutic Evaluative Conditioning: App-based gamified conditioning program in which self-harm related stimuli were sequentially paired with adverse stimuli. Participants were encouraged to interact with the program as often as necessary over a one month period.

Participants were allocated to an attentional control condition.

Suicidal ideation: scores on the Self-Injurious Thoughts and Behaviours Interview, indicating number of days in which the participant had had thoughts of suicide.

Self-harm: scores on the Self-Injurious Thoughts and Behaviours Interview, indicating the number of days in which the participant had engaged in self-cutting or NSSI.

Neither participants nor clinical personnel were blind to treatment allocation.

 Study 1

RCT

International

 Study 2

RCT

International

 Study 3

RCT

International

Guille, 2015 [41]

RCT

USA

Medical interns commencing their residency year recruited from one of two hospitals.

MoodGYM: Internet self -guided psychoeducation program coupled with self-guided iCBT consis-ting of four modules delivered once a week over a four week period.

Participants were allocated to an attentional control condition.

Suicidal ideation: proportion of participants self-reporting any thoughts of suicide according to scores on item nine of the Patient Health Questionnaire–9 (PHQ–9).

Neither participants nor clinical personnel were blind to treatment allocation. One-half (49%) did not complete treatment.

Hedman, 2014 [48]

Pre-test/Post-test

Sweden

Recruited from a university hospital psychia-tric clinic.

iCBT: consisting of 10 modules of behavioural activation and cognitive therapy delivered over a 12 week period.

N/A.

Suicidal ideation: scores on item 9 of the Montgomery Åsberg Depression Rating Scale (MADRS).

Appropriate adjustment was made for potential time-varying confounding. Neither participants nor clinical personnel were blind to treatment allocation. One-quarter (25%) did not complete treatment.

Hill, 2016 [44]

RCT

USA

Recruited following an advertising campaign. All participants had to self-report significant levels of perceived burdensomeness (i.e., score of 17 or greater on the Inter-personal Needs Question-naire Perceived Burden-someness sub-scale).

LEAP: Internet self-guided program informed by the Interpersonal Theory of Suicide (prin-cipally perceived burden-someness) consisting of two modules delivered over a two week period.

Participants were allocated to psychoeducation.

Suicidal ideation: scores on the Beck Scale for Suicidal Ideation.

Participants were not blind to treatment allocation. Control condition did not account for intervention time or engagement (i.e., did not satisfy attentional control requirements). Over one-third (39.0%) did not complete treatment.

Kordy, 2016 [46]

RCT

Germany

Recruited from one of six psychiatric departments. All patients had to meet diagnostic criteria for major depression accord-ing to the Structured Clinical Intervention for DSM-IV and to self-report a history of at least three prior episodes of depression to be eligible for participation.

SUMMIT: Internet self-guided emotion coping program. The number of treatment modules was not stated, but participants were encouraged to access the intervention as often as necessary over a 12 month period.

Participants were allocated to usual psychiatric care, including pharmaco-therapy and/or psychotherapy as necessary.

Self-harm and/or attempted suicide: re-hospitalisation following an episode of self-injury and/or a suicide attempt.

Neither participants nor clinical personnel were blind to treatment allocation. Proportion completing treatment not stated.

Mewton, 2015 [38]

Pre-test/Post-test

Australia

Patients diagnosed with depression who were prescribed a course of iCBT by their primary care provider.

iCBT: Consisting of six modules delivered over six week period.

N/A.

Suicidal ideation: scores on item 9 of the Patient Health Questionnaire–9 (PHQ–9).

Appropriate adjustment was made for presence of missing data and for potential time-varying confounding. Neither participants nor clinical personnel were blind to treatment allocation. Almost one-half (43.2%) did not complete treatment.

Mortiz, 2012 [45]

Pseudo-RCT

Germany

Recruited from websites providing support to persons with depression.

Deprexis: self-guided iCBT program consisting of 10 modules delivered over an eight week period.

Participants were allocated to a wait-list control condition.

Suicidal ideation: scores on the German language version of the Suicide Behaviors Questionnaire-Revised (SBQ–R).

Neither participants nor clinical personnel were blind to treatment allocation. Participants were alternately allocated to the intervention and control groups. One-fifth (21.9%) did not complete treatment.

Robinson, 2016 [39]

Pre-test/Post-test

Australia

Secondary school students presenting to school-based counsellors with suicidal ideation in the past month.

Reframe-IT: Internet self-guided iCBT program consisting of eight modules delivered over an eight week period.

N/A.

Suicidal ideation: scores on the Suicidal Ideation Questionnaire-Junior (SIQ-JR) or the Adult Suicidal Ideation Questionnaire (ASIQ) (as appropriate to the participant’s age).

Analyses based on treatment completers only (N = 21). Neither participants nor clinical personnel were blind to treatment allocation.

Tighe, 2017 [40]

RCT

Australia

Recruited though a community-based suicide prevention organisation. All participants had to identify as Aboriginal and /or Torres Strait Islander. All participants also had to report clinically sig-nificant levels of depression (i.e., scores of 10 or greater on the Patient Health Questionnaire-9), psychological distress (i.e., scores of 25 or greater on the Kessler Psychological Distress Scale), and had to self-report suicidal thoughts in the past two weeks.

iBobbly: App-based self-guided CBT program consisting of three modules completed over a six week period. Partici-pants were also encouraged to complete self-assessments on functioning, suicidal thinking, mind-fulness exercises, self-soothing activities, and cultural engagement activities.

Participants were allocated to a wait-list control condition.

Suicidal ideation: scores on the Depressive Symptoms Inventory-Suicidality Subscale (DSI-SS).

Due to changes in inclusion criteria after the commencement of the trial, one-quarter (26.2%) of the included participants did not meet the criterion for frequency of suicidal thoughts. Participants, clinical personnel, and outcome assessors were not blind to treatment allocation. Data on usage was available for 65.6% of the included participants. Of these, 15.0% did not complete treatment.

Van Spijker, 2014 [47]

RCT

Netherlands

Recruited following an advertising campaign. All participants had to report mild to moderate suicidal ideation to be eligible for participation. Participants self-reporting severe depression were excluded.

iCBT: Consisting of six modules delivered over a six week period with added activities from DBT, PST, and mindfulness-based cognitive therapy.

Participants were allocated to an attentional and online bibliotherapy.

Suicidal ideation: scores on the Beck Suicide Ideation Scale.

Neither participants nor clinical personnel were blind to treatment allocation. Around one-half (44.0%) did not com-plete at least three treatment modules.

Van Voorhees, 2009 [42]

Pre-test/Post-test^a

USA

Adolescents receiving treatment for depression at primary care facilities registered with one of five major Health Care Organisations (HMOs).

CATCH-IT: Internet self-guided program consisting of 14 modules of CBT, IPT, and comm-unity resiliency activities delivered over a 12 week period.

N/A.

Suicidal ideation: scores on item 9 of the Patient Health Questionnaire–9 (PHQ–9).

Almost all (96.4%) participants completed treatment. As data were analysed as pre-test/post-test compare-sons in this review, no appropriate adjustment could be made for potential time-varying confounding. Neither participants nor clinical personnel were blind to treatment allocation.

Wagner, 2014 [49]

RCT

Switzerland

Recruited from university websites, newspapers, poster adverts, and from referrals to a depression self-help group. Participants had to score at least 12 on the Beck Depression Inventory-II at the screening interview.

iCBT: Consisting of seven modules delivered over an eight week period.

Participants were allocated to eight weeks of manualised face-to-face CBT delivered by a registered psychologist.

Suicidal ideation: scores on the Beck Suicide Ideation Scale.

Neither participants nor clinical personnel were blind to treatment allocation. Despite rand-omisation, a significantly higher number of females were allocated to the intervention. One-fifth (22.0%) did not complete treatment.

Whiteside, 2014 [43]

Pre-test/Post-test

USA

Primary care patients receiving treatment for a new episode of depression and who were not already receiving either pharm-acotherapy or psycho-therapy. Those who had received treatment for a depressive episode in the last six months were not eligible to participate.

Thrive: iCBT program consisting of three modules delivered over a three week period. Partici-pants were also encouraged to access a secure messaging platform at least once a week for eight weeks to receive a moti-vational intervention from a qualified iCBT coach.

N/A.

Suicidal ideation: scores on item 13 of the Symptom Checklist-20.

No appropriate adjustment was made for potential time-varying confounding. Neither participants nor clinical personnel were blind to treatment allocation. One-fifth (22.0%) did not complete treatment.