Internet-based interventions (IBIs) offer numerous efficient prevention and treatment programs for a variety of psychological difficulties [1,2]. Generally based on methods originating from empirically supported face-to-face psychological interventions for people reporting complicated grief symptoms [3,4], IBIs for grief-related symptoms are also effective [5,6].

Offering guidance to participants is one of the most commonly cited means to improve IBI effectiveness [7], including grief-related symptoms [5]. However, recent evidence indicates that the benefits of guidance are lower in more interactive internet interventions [8]. Moreover, when given the choice, not all participants request guidance. Additionally, the efficacy of guidance on demand is similar to standard weekly guidance [9-11]. Finally, including specific and individualized principles, such as the Motive-Oriented Therapeutic Relationship (MOTR), in the guidance appears feasible and useful for IBIs [12,13]. Thus, MOTR-based guidance on demand appears as a cost-effective alternative to mandatory guidance. Other elements can also improve adherence rate and efficacy: (1) using automated reminders [14]; (2) providing interactivity in the tasks and exercises [15-18]; (3) promoting personal resources to cope when problematic experiences arise [19,20]; (4) tailoring the intervention to the participant’s characteristics and timing the content in accordance with the participant’s characteristics [21-24].

LIVIA 2.0 was developed as an alternative IBI to LIVIA 1, the original program, for grief-related symptoms after bereavement or separation [25], which was tested in German via a randomized controlled trial (RCT) [26] and in French via a noncontrolled trial [27]. Indeed, fundamental research has pointed to the many similarities between these kinds of losses [27,28]; moreover, studies on LIVIA 1 have proved that the same intervention can be provided to both populations [26,29]. LIVIA 2.0 implements different factors to improve patient adherence and program efficacy. Specifically, it includes automated emails, increased interactivity (quizzes, video files, and audio files), personal resource assessment and promotion, and the freedom to choose the session order combined with an individualized recommendation. The content is based on the Dual Process Model (DPM) of bereavement recovery and proposes an oscillation process mimicry [30,31] with an alternation of loss-focused and restauration-focused sessions. Finally, LIVIA 2.0 includes a module focused on identity and memory processes that play a key role in adapting to loss [32,33] as well as novel emotion regulation tools [34,35].

In detail, the development of LIVIA 2.0 and its innovations were based on the theoretical and empirical literature about grief and romantic dissolution. On the theoretical level, we relied on one of the most influential models of coping with loss, the DPM of Coping with Bereavement [30,31]. According to it, after loss, instead of going through consecutive phases, people oscillate between a focus on the loss and a focus on the restauration from the loss. This model postulates that oscillation is a natural and necessary movement to cope with loss. Moreover, DPM-based interventions are more efficient than classic ones [36]. Hence, LIVIA 2.0 imitates the oscillation process by alternating between loss- and restauration-focused sessions. Additionally, LIVIA 2.0 integrates recent loss-related empirical findings into its content and exercises. For example, the Emotion module proposes self-compassion exercises, as self-compassion has been shown to predict better grief recovery [37,38]. Moreover, LIVIA 2.0 includes a newly developed module based on empirical cognitive psychopathological knowledge and focused on identity processes, which play a key role in adaption to bereavement [32,33]. Addressing identity factors, such as fostering an independent sense of identity by focusing on adaptive specific autobiographical memories and future projections, could improve existing cognitive-behavioral programs for grief.

LIVIA 2.0 also integrates recent developments in IBIs [16,20]. First, a series of changes are designed to improve participant autonomy by sending automated emails [14], providing individualized recommendations about the order in which to complete the modules [22], promoting and encouraging the use of personal resources [20], and augmenting the interactivity of the website [15-18]. Finally, the emails exchanged with the participants within the guidance on demand framework will be based on the MOTR [12].

Our main hypotheses are the following: (1) LIVIA 1 and LIVIA 2.0 increase participants’ well-being and decrease their distress at posttest and follow-up. (2) LIVIA 2.0 is more efficient than LIVIA 1 for all outcomes. (3) LIVIA 2.0 has less dropouts than LIVIA 1.

Moreover, we will conduct the following exploratory analyses. First, we will compare the guidance requirements of the participants (ie, number of participants requiring guidance and number of exchanged emails) in LIVIA 2.0 with those of the LIVIA 1 participants and explore which session triggers more requests for guidance. Second, we will examine in LIVIA 2.0 the short-term effectiveness of each module on the participants' weekly moods, feelings of loneliness, and grief symptoms. Third, we will compare participant satisfaction in the 2 versions of LIVIA. Fourth, we will explore the role of multiple measures (attachment style, type of loss, interpersonal closeness to the lost person, and symptom severity) as moderators of the program’s effectiveness. Finally, we will investigate the semantic content of the responses to the LIVIA 2.0 exercises [39] to explore its relationship with improvement over the evaluation period.

This is a study of an IBI in French. We will recruit participants in Switzerland, but participation will be open to French-speaking people across the world.

This study is a monocentric, single-blinded, 2-arm RCT comparing the efficacy of 2 versions of an IBI—namely LIVIA 1 and LIVIA 2.0—aimed at relieving distress and augmenting the well-being of people suffering from prolonged grief symptoms. There will be three measurement points: a pretest (T0), posttest (T1), and follow-up (T2). The flowchart of the study design is presented in Figure 1.

This self-help intervention is embedded in a larger project on life-span vulnerabilities and strengths conducted by the LIVES Centre [40].

To assign participants to the LIVIA 2.0 (experimental) or LIVIA 1 (control) condition, we will use the randomization module in REDCap (Research Electronic Data Capture [41,42]), which will automatically generate a 1:1 randomization. We will apply a single-blinded randomization strategy stratified according to the gender and loss type (bereavement vs separation) of the participants. Due to the nature of the study, double-blinded randomization is not possible.

During the intervention, participants will receive a biweekly assessment of the occurrence of a serious adverse event. Additionally, they will be able to contact the investigation team at any time through a contact form available in both the IBIs. An automatic reply will be sent informing that their message will be processed within 3 working days and that they can contact the emergency numbers that will be provided. This will enable the participant to make contact and be attended to at any moment if necessary. We will then discuss with them the reason for contacting us and if a serious adverse event is occurring. If this is the case, the participants will be contacted, and their situation will be evaluated. If necessary, alternative treatments will be proposed.

The calendar for obtaining the measures from the participants can be found in Multimedia Appendix 1.

Study data will be collected and managed using REDCap electronic data capture tools hosted at the Lausanne University Hospital [41,42]. REDCap is a secure, web-based software platform designed to support data capture for research studies, providing (1) an intuitive interface for validated data assessment, (2) audit trails for tracking data manipulation and export procedures, (3) automated export procedures for data downloads to common statistical packages, and (4) procedures for data integration and interoperability with external sources. Data integrity is enforced through several mechanism (ie, referential data rules, valid values, range checks, and consistency checks).

Moreover, data on the use of the self-help sessions will be collected within the platform, as well as the entries of the participants in LIVIA 2.0 (exercises, quizzes, questions, etc). All data will be saved anonymously, identified only by a random code. The servers are protected by high-end firewall systems. The participant code list will be saved on an internal Network-Attached Storage. Only the researchers directly involved in the study will have access to the data.

Primary analyses will be performed using SPSS (IBM Corporation) and R (R Foundation for Statistical Computing) after the participants complete the study. Primary follow-up analyses will be conducted 3 months later. Intermediate analyses during data collection will be done. We will perform intention-to-treat analyses [70]. We will develop multilevel mixed effects models with repeated measures data in SPSS to evaluate the efficacy of LIVIA 2.0 compared to LIVIA 1 and the stability of the effects. These models have the advantages of considering the dependency of the data and accounting for the correlation of repeated measures within individuals. Moreover, they rely on all available data of every participant and estimate parameters of missing values [71].

We will explore the potential moderating effect of different variables on the intervention effects by conducting analyses of covariance for repeated measures. To explore the short-term efficacy of each module, we will use multilevel modeling whereby each participant’s monitoring data are nested within participants. These analyses will be conducted with Mplus [72] or R. Additionally, we will analyze qualitatively the satisfaction questionnaire using the thematic content analysis method [73]. Finally, we will analyze the semantic content of the exercises included in LIVIA 2.0 [39] and test if some categories (eg, “we-talk”) are associated with the efficacy of the program. We will rely on a significance level of a 2-sided α=.05 or smaller. We will use the Bonferroni correction to adjust for multiple testing.

To handle missing data and dropouts, we will conduct the analyses relying on the intention-to-treat paradigm. We will first analyze the magnitude of missing data, explore the missing data patterns, and determine the pattern of missing data (missing completely at random, missing at random, and not missing at random). If the missing mechanism is missing at random, we will use multilevel regression analyses, which allow for nonindependent observations and for different numbers of measurement points per participant and are thus less sensitive to missing data [71].

The Clinical Trial Unit of the Centre de Recherche Clinique (CTU/CRC) of the Lausanne University Hospital will monitor the study. This includes (1) a monitoring preparation meeting; (2) an initiation visit whereby before starting the study, the CRC team and the investigation team will go over the entire procedure of the clinical trial; (3) intermediary visits whereby the CTU/CRC will control the available data and write reports on a regular basis; and (4) a close-out visit whereby upon study completion, the CRC team and the investigation team will meet 1 last time and revise all the study materials.

The protocol (trial registration: NCT05219760) has been approved by a federally acknowledged ethics committee (Commission cantonale d'éthique de la recherche sur l'être humain, CER-VD, BASEC reference number: 2021-D0086) and by the Swiss Agency for Therapeutic Products (reference number: 102667545) in accordance with the Swiss Ordinance 810.306 on Clinical Trials with Medical Devices. We will obtain informed consent from all participants in the study (see Multimedia Appendix 2), codify their data, and ensure secure storage of the data.

The treatment gap in mental health care is a global reality [75], and we can expect the COVID-19 crisis to have aggravated it [76]. Moreover, because of the increased deaths and obstacles to social contact due to the distancing measures during the pandemic, it is particularly crucial to make IBIs available for general purposes [77] and for grief symptoms in particular [78]. To our knowledge, this study will be the first one to test an IBI targeting grief symptoms in French. As such, it will contribute toward facilitating access to high-quality IBIs for the French-speaking population. Additionally, research on IBIs using a guidance on demand framework is rare [10,11]. Hence, this study will contribute to knowledge about this method that may provide a promising outlook from human resource and economical perspectives. This project will thus contribute toward expanding the possibilities to offer accessible and cost-efficient interventions to people in need.

Some potential limitations can be anticipated. First, the self-selection of the sample may result in participants that have higher education levels and a higher proportion of women than the targeted population of people experiencing grief symptoms after bereavement or romantic separation [79]. However, the study will provide valuable information about the efficacy of LIVIA 2.0 for the targeted population. Second, because participants will have considerable freedom in the way they navigate through the program (ie, choosing the order of the modules and selecting the exercises they want to complete in each session), all the participants will not have taken the same path to complete LIVIA 2.0. However, this will provide them with more choices and higher individualization. Finally, we cannot confirm accurately if the participants respect the exclusion criteria of not undergoing face-to-face psychotherapy while doing the program. This would not be ethically nor practically feasible. However, as the participants are randomized to both programs, this potential bias should be equal in both study arms.

IBIs contribute to closing the so-called treatment gap, which refers to the difference between the number of people needing psychological treatment and those who actually undergo it [80,81]. This is particularly true in the case of interventions targeting grief symptoms, as a significant proportion of people reporting the need for support after bereavement do not get it [82,83]. Additionally, psychotherapists often lack the competences to deal optimally with this affection [84,85]. Finally, the COVID-19 pandemic crisis has contributed to the increased risk of suffering complicated grief symptoms [86,87]. Hence, developing, assessing, and offering IBIs for grief-related symptoms seems particularly crucial in the current times [78,86]. The results of this RCT will give insight into the relevance of the present developments in outcome improvement and dropout diminution for adults who experience grief symptoms. Besides, the study design will allow for conducting additional analyses that can provide a deeper and more fine-grained understanding of the mechanisms of change in IBIs. For example, we will be able to analyze the effect of specific modules on the weekly mood and symptom monitoring, assess the effect of some moderators (eg, attachment style or closeness to the lost person at pretest), or study the linguistic behaviors in the exercises where participants are required to describe a loss-related situation [39].