Data were collected in the EDs of two hospitals in Madrid, Spain, between September 2022 and August 2023. Participants were adults (aged ≥18 years) presenting to the ED for suicidal behavior. Suicidal behavior was defined as any action or self-injurious act with the intent to die, encompassing both suicidal ideation and suicide attempts. Once potential participants received standard emergency care and were deemed medically stable, they were offered the opportunity to participate in the study prior to discharge. Those who consented to participate met with a psychiatrist for 30‐40 minutes to receive a detailed explanation of the study, install the study app on their smartphones, and complete the digital safety plan. The inclusion criteria were (1) presentation to the ED for suicidal behavior; (2) ownership of a smartphone with internet connection; (3) speaking and understanding Spanish; and (4) being able to give informed consent. Exclusion criteria included unwillingness to install the app or limited smartphone access.

The study was approved by the Research Ethics Committee of the Fundacion Jimenez Diaz Hospital (Approval number: EC005-21_FJD). All participants provided informed consent. Participants were free to withdraw from the study at any time or decline further participation during follow-up without affecting their medical care. All collected data were anonymized to protect participant identity. No compensation was provided for study participation.

A total of 78 participants were included in the study, with a mean age of 39.08 (SD 13.45) years; most were female (n=56, 72%). According to the baseline suicidal behavior characteristics (ie, C-SSRS), 72 (90%) participants exhibited high-intensity suicidal ideation, and 63 (80%) participants reported prior suicide attempts (Table 1). Over the one-year follow-up, 24 (31%) participants returned to the ED services for suicidal behavior (ie, suicidal ideation or attempt).

Subsequently, the frequency of return to the ED and digital safety plan usability by the participants during the follow-up period were explored. A total of 2.768 observations were collected for the digital safety plan usage. Of the overall enrolled participants, 31 (40%) activated the digital safety plan during the follow-up period.

Given the non-normal distribution and positive skew of these variables, we reported the median (IQR) and 90th percentile (P90). Within the one-year follow-up, the median time for return to the ED postindex event was 55 (IQR 211, range 1‐359; P90=301) days, with a higher frequency observed within the first 60 days, followed by a more uniform distribution throughout the remaining year (Figure 2A). Excluding the initial app installation and collaborative safety plan installation with the psychiatrist, the median overall digital safety plan usage occurred at 14 (IQR 47, range 2‐359; P90=105) days (Figure 2B). When examining the frequency of digital safety plan configuration (ie, personalizing the safety plan), the median usage occurred at 11 (IQR 28, range 2‐359; P90=107) days, with a decreased frequency throughout the rest of the year (Figure 2C). For digital safety plan activation (ie, accessing the safety plan during crisis moments), the median usage occurred at 15 (IQR 67, range 2‐260; P90=112) days, with a stable frequency in the first 90 days and a significantly reduced frequency in the remaining follow-up period (Figure 2D). Overall, 90% (62/69) of participants returned to the ED due to suicidal behavior within 300 days, while 90% of digital safety plan configurations (1392/1547) and activations (529/588) occurred within the first 100 days of follow-up.

When examining the usage of the digital safety plan configuration, “external coping strategies” (n=362, 23.30%), “professional ask for help” (n=328, 21.10%), and “warning signals” (n=234, 15%) were the most frequently accessed functions, followed by “personal contacts” (n=188, 12.10%) and “internal coping strategies” (n=178, 11.40%); “creating a safe environment” (n=141, 9.10%) and “reasons to live” (n=124, 8%) were less frequently used. For digital safety plan activation, the most frequently activated tabs were “external coping strategies” (n=140, 22.90%), “warning signals” (n =125, 20.85%), “professional ask for help” (n= 120, 19.48%), “internal coping strategies” (n =74, 12.30%), and “personal contacts” (n =64, 10.42%); “creating a safe environment” (n= 50, 7.69%) and “reasons to live” (n= 41, 6.32%) were less frequently used. No significant differences were found in age, sex, suicidal behavior characteristics, or severity of suicidal ideation between those who activated the digital safety plan and those who did not (all P values >.05).

In general, while all safety plan function or sections were used to some extent, we observed a usage pattern that corresponded with the tab sequence in the app’s interface. Specifically, “creating a safe environment” and “reasons to live,” which were positioned at the end of the app’s interface, were least frequently used. This finding suggests that the arrangement of the functions influenced the usability of the digital safety plan.

Participants who activated the digital safety plan had a lower frequency of ED visits compared to those who did not activate the plan (n=102, 16.6% vs n=313, 29.1%; χ²₁=37.48, P<.001).

Subsequent analyses examined the predictive utility of digital safety plan activation for ED visits over the one-year follow-up period. Using the coxme function, we identified the best-fitting model with participant ID and days of safety plan usage as random effects. Receiver operating characteristic analysis demonstrated good discriminatory capacity of the model (area under the curve [AUC]=0.82) for predicting ED visits for suicidal behavior based on digital safety plan activation (Figure 3).

The Cox proportional hazards model, adjusted for participant demographics, revealed a significant association between digital safety plan activation and a reduced likelihood of subsequent ED visits. Participants who activated the safety plan had a 50% lower likelihood of returning to the ED for suicidal behavior (ie, suicidal ideation or attempt) compared to those who did not (Figure 4). Furthermore, male participants and those aged ≥30 showed a reduced likelihood of returning for subsequent ED visits (Table 2). These results suggest that engaging with the digital safety plan and consequently accessing preconfigured suicide-related coping strategies within the app, may enable individuals to effectively manage suicidal thoughts and prevent escalation during a suicidal crisis, potentially reducing the frequency of ED readmissions. Additionally, sex and age were important clinical factors and should be considered in predicting ED readmissions for suicidality.

Prior studies have established the relevance of safety plan usage in preventing or managing suicidal crises. However, to date, the effectiveness of a mobile digital safety plan in reducing ED visits for suicidal behavior has not been evaluated. This study found that participants who chose to activate a digital safety plan showed a 50% reduction in the likelihood of subsequent suicide-related ED visits, compared to those who did not activate it.

Additionally, male participants and those aged ≥30 were less likely to return to emergency services for suicidal behavior during the one-year follow-up after an index event, compared to female and younger participants. Generally, men are less inclined to seek assistance for suicide-related issues, even though they tend to use more hazardous methods and exhibit elevated fatal-suicide rates globally [3], including in European countries [33]. Nonetheless, no association was observed between digital safety plan activation and participants’ sex, age, or severity of suicidal ideation.

Our analysis also revealed that 30.8% of participants returned to the ED due to suicidal behavior following the index event, predominantly within the first 60 days. The initial three-month period post discharge represents a time of heightened risk for reattempt, thus highlighting the relevance of prevention and follow-up during this phase [6]. Similarly, the highest frequency of digital safety plan activation occurred within the first few weeks, with a median of 15 days. Consequently, digital safety plan usage was brief and strongly associated with the index event. Prior research has indicated that the usability of the digital safety plan is sporadic and concentrated in the early follow-up weeks [24]. Safety plans are designed for use during suicidal crises [12]; therefore, it is expected that users will engage with them only when necessary, particularly at the onset of suicidal thoughts.

The most frequently activated sections of the digital safety plan were “warning signs,” “external coping strategies,” and “professional contacts.” In contrast, less frequently used sections included “reasons to live,” “safe environments,” “personal contacts,” and “internal coping strategies.” These findings suggest a potential preference for seeking external support during suicidal behavior, such as contacting professionals, rather than relying solely on internal resources. This observation is consistent with findings that indicate social support serves as a protective factor against suicidal behavior [34], particularly among individuals with a history of suicide attempts [35]. It is also possible that during periods of intense or severe suicidal ideation, internal coping strategies become more difficult to activate and manage independently, whereas external suicide-related coping are more readily accessible [36]. Additionally, the design and interface of the digital safety plan app may impact which features are prioritized by users. For example, the interface’s tab order—with “warning signs” positioned first and “reasons to live” last—may impact usability. Future research should further explore the usability of the app and identify which sections are most useful to users. Suggested improvements could include in-app digital scales to assess satisfaction with app use and the perceived effectiveness of the functions included in the digital safety plan.

The findings of this study have important clinical and practical implications. The observed association between safety plan activation and reduced ED readmission for suicidal behavior underscores the potential of digital safety plans in secondary suicide prevention. Integrating such tools into routine postsuicide attempt care could provide an accessible, scalable, and potentially cost-effective intervention to mitigate future crises. Moreover, digital safety plans offer an ecologically valid approach by enabling real-time usage monitoring and empowering individuals with digital tools and suicide-related coping strategies to manage future suicidal crises [24].

Digital mental health interventions have the potential to leverage technology to increase user engagement in suicide prevention and ensure continuity of treatment between emergency or hospital services and outpatient care [37]. Furthermore, digital interventions could be cost-effective, given their promising cost-benefit ratio in suicide prevention within the framework of current mental health interventions, which could support increased public health policy investment in digital strategies [38,39]. In particular, digital safety plans for suicide prevention present a compelling alternative to traditional methods, offering enhanced usability and reduced access barriers. These plans can integrate multimedia elements such as personal photos, prerecorded messages, relaxation videos, and links to health resources [19]. Their immediate availability, customizable nature, and dynamic functionality—allowing for real-time updates—further distinguish them [26].

However, despite these advantages, the practical usability of digital safety plan outside clinical environments remains underexplored. Enhancing accessibility and monitoring use are crucial for understanding their preventive impact. Research indicates that many individuals seeking suicide-related support are unfamiliar with safety plans or lack these in their treatment plans [17]. Furthermore, depressive moods and a perceived lack of collaborative planning with clinicians may hinder effective use [18]. Future studies should prioritize the sustainability of these tools, focusing on identifying usage barriers and implementing user-driven improvements through rigorous follow-up. Integrating specialized telephone follow-up could also provide timely support and reinforce user engagement by offering guidance on coping strategies [14].

First, the generalizability of our findings is limited by the small sample size, which was largely composed of women and did not include adolescents. Second, this study lacked detailed data on the duration of the use for each digital safety plan function and its perceived utility. Such data could provide valuable insights into which specific functions, or their combinations, are associated with the reduction of suicidal ideation and ED visits. Third, we did not collect information on participants who did not use or activate the digital safety plan. Understanding why some individuals did not engage with the tool (eg, due to perceived lack of need, usefulness, or technological barriers) would be valuable. Fourth, the absence of a control group limited our ability to directly compare the efficacy of digital versus traditional safety plan in suicide prevention. Finally, data on suicidal behavior features were based on baseline and self-reported measures. Future research could enhance these findings by incorporating real-time, in-app monitoring of suicidal ideation (eg, intensity and duration) and correlating it with digital safety plan use.

To our knowledge, this is the first study to demonstrate that activating a digital safety plan can reduce the likelihood of repeat suicide-related ED visits by 50%. While these results do not directly demonstrate a decrease in suicidal behavior among participants, the reduction in ED visits following an index suicidal event suggests that individuals may find supportive elements to manage a crisis by readily accessing suicide-related coping strategies through the activation of digital safety plan. This suggests that digital safety plans may serve as a valuable resource, as a complement to interventions delivered in the ED. Future research should investigate the dynamics of digital safety plan use among participants, focusing on which functions (ie, tabs) are most frequently used and their specific associations with reductions in suicidal behavior and ED visits.