Suicide is a severe public health problem. Globally, more than 800,000 people die because of suicide each year, and it is the second leading cause of death among those aged 15 to 29 years [1]. Suicide attempts are estimated to be 20 times more prevalent than completed suicides [2]. Globally, suicide rates have increased by 60% within the last 45 years [2].

Although there are effective suicide prevention strategies [3], many people at risk of suicide do not seek treatment, which may limit the effectiveness or impact of these strategies [4]. Surveys conducted worldwide (N=55,302 participants) by the World Health Organization show that only 39% of people with suicidal behavior had received any kind of therapeutic intervention for emotional difficulties in the preceding 12 months [5]. Treatment, including treatment of suicidal thoughts, was most prevalent in high-income countries (56% received treatment within the past 12 months) and less frequent in middle-income (28%) and low-income (17%) countries. Common barriers to treatment-seeking have been identified as (1) the wish to solve the problem by oneself, (2) the belief that one would get better without treatment, (3) the belief that the problem was not that severe, (4) stigma, (5) structural problems, for example, financial effort and low availability of treatment, and (6) low perceived need [5].

Over the past two decades, the development and evaluation of internet-and mobile-based interventions has been an emerging focus of mental health research. Previous meta-analyses have verified the effectiveness of digital interventions for a variety of mental disorders and health issues, including depression [6-10], anxiety [6] and post-traumatic disorder [11]. Internet-based interventions have now been integrated in clinical practice in several countries, including Australia, the Netherlands, Sweden, Norway, and England [12,13], demonstrating their value as a part of health care for at-risk individuals. Within the diverse field of digital interventions, internet-based self-help interventions (ISIs) are the most commonly developed and used. ISIs are stand-alone interventions that provide participants with evidence-based therapeutic material, which can be used self-reliantly [12]. ISIs can involve different levels of human support [14]. In guided ISIs, a clinician accompanies the intervention by providing feedback or guidance on the tasks and progress, often on a weekly basis [14]. Human support is typically limited to positive reinforcement, giving feedback and clarification of content instead of delivering additional therapeutic techniques [12].

The use of ISIs might address several of the barriers mentioned above: (1) The desire to “solve the problem by oneself” can be appropriately addressed by using guided or unguided self-help interventions [12,14]. (2) Individuals who assume that they will get better without treatment might still look for information or social support online [15], and psychoeducational content can be readily delivered through internet-based interventions. It has also been shown that suicidal individuals spend more time online than nonsuicidal users [16-18], which indicates that ISIs may be very appealing to these individuals. (3) The perception of the problem as not that severe can be addressed by providing a low-threshold program within a stepped-care approach. As a first step, ISIs can be offered, and if the patient does not respond, further programs with more intensive therapeutic support can be provided [12]. (4) Individuals who face stigmatization might benefit from the anonymity that ISIs offer [12], (5) whereas structural barriers can be addressed by the accessibility and flexibility of ISIs. In addition, ISIs can be provided at low costs [19,20]. (6) Although offering self-help interventions does not address the barrier of low perceived need, online self-screening programs may increase the perceived need for treatment by providing feedback to participants [21]. In sum, ISIs may be an appropriate, low-threshold intervention for individuals at risk of suicide.

ISIs for suicide prevention have been developed in recent years. In response to the growth of electronic health (eHealth) interventions in mental health, several reviews and meta-analyses have been published which summarize the evidence of internet-and-mobile-based interventions for suicidality [21-25]. The umbrella term internet-and- mobile-based suicide prevention comprises widely divergent approaches, including social networking sites, videos, podcasts, email support programs, mobile apps, gaming interventions [26], self-screenings, text analyses [21], and self-help interventions based on psychological treatment approaches [12]. As a result, reviews have included a variety of interventions.

Overall, those reviews found mixed results and pointed out a paucity of high-quality evidence for the effectiveness of the reviewed interventions. Some reviews reported promising effects of ISIs on suicide-related outcomes [22,23], and one review concluded that there is no evidence for their effectiveness [25]. The most recent of these meta-analyses, by Witt et al [24], found that internet-and-mobile-based interventions (mainly developed for depression treatment) reduced suicide ideation at postintervention. Christensen et al [21] reported that self-help interventions for depression have specific effects on depressiveness but not on suicidal ideation, as there seem to be reductions in suicidal ideation in both depression and control conditions. They included only 1 study with an intervention directly addressing suicidality. It showed a significant effect on suicidal thoughts compared with the control group. The authors concluded that online programs that directly target suicidal ideation and behavior might be more effective than those programs that are designed for mental health more broadly. This is in line with the meta-analytic finding that interventions directly addressing suicidal thoughts and behavior show immediate post-treatment and long-term effects, whereas programs that solely address associated symptoms seem to be only effective in the long term [27].

However, there are some limitations concerning previous reviews in this field. First, ISIs were often not differentiated from other Web-based strategies or mobile-based interventions [22-24]. Differential effects are highly plausible owing to variations in treatment approaches, treatment dose, and application context. Second, several reviews did not restrict study inclusion to controlled trials [21-24] and, third, did not assess risk of bias [21,23]. Fourth, none of the reviews assessed publication bias [21-23,25]. Finally, most reviews did not differentiate between studies investigating interventions that directly targeted suicide versus those that focused on other conditions such as depression or anxiety [22,24,25].

Therefore, this review and meta-analysis will (1) focus on the effectiveness of ISIs directly targeting suicidal thoughts or behavior, (2) exclude Web-based interventions for other health conditions and mobile interventions, and (3) only include randomized controlled trials (RCTs), (4) perform the Cochrane Risk of Bias tool, (5) check for publication bias and (6) search a clinical trial register to give an overview of ongoing trials. As Web-based suicide prevention is a fast-growing field, this review will provide readers with a valuable up-to-date overview of the current state of research and identify gaps in the literature to benefit the design of future research investigations.

The systematic literature search will be conducted in the following databases: PsycINFO, MEDLINE, Cochrane Central Register of Controlled Trials (CENTRAL) and the Centre for Research Excellence of Suicide Prevention (CRESP) Database. Search strings enabling a sensitive search (incorporating numerous Medical Subject Headings, subject terms, keywords, and publication types associated with internet, eHealth, suicide, or RCT) were developed for PsycINFO, MEDLINE, and CENTRAL (see Multimedia Appendix 1). As CRESP contains a manageable number of trials, we will screen all studies included in the database. We performed a pilot testing of the outlined search strategy. Hand searches identified 5 eligible trials [33-37]; 100% of these trials were identified by searching the databases using the search strings. In addition, a search in the clinical trial register, ClinicalTrials.gov (provided by the US National Library), will be performed to identify ongoing trials. Hence, this review will provide not only the current state but also the outline of emerging developments in the field. We will screen the reference lists of all included studies and relevant reviews articles for additional studies (backward search), and we will screen studies that cited the included studies and relevant reviews (forward search). In addition, we will perform hand searches. We plan to conduct the searches until April 30, 2019. Gray literature will not be included. Registered trials that have not been published will be used to evaluate possible publication bias.

If it remains unclear whether a study meets the eligibility criteria or if relevant data or analyses have not been reported, we will contact the authors for clarification. We will also contact authors to ask for unpublished results when study protocols without a subsequent publication are identified.