How to prevent technostress at the digital workplace: a Delphi study

Technostress is a rising issue in the changing world of digital work. Technostress can cause severe adverse outcomes for individuals and organizations. Thus, organizations face the moral, legal, and economic responsibility to prevent employees’ excessive technostress. As technostress develops over time, it is crucial to prevent it throughout the process of its emergence instead of only reacting after adverse outcomes occur. Contextualizing the Theory of Preventive Stress management to technostress, we synthesize and advance existing knowledge on inhibiting technostress. We develop a set of 24 technostress prevention measures from technostress inhibitor literature, other technostress literature, and based on qualitative and quantitative contributions from a Delphi study. Based on expert feedback, we characterize each measure and, where possible, assess its relevance in addressing specific technostressors. Our paper contributes to research by transferring the Theory of Preventive Stress Management into the context of technostress and presenting specific measures to prevent technostress. This offers a complementary view to technostress inhibitors by expanding the theoretical grounding and adding a time perspective through the implementation of primary, secondary, and tertiary prevention measures. For practice, we offer a comprehensive and applicable overview of measures organizations can implement to prevent technostress.


Introduction
Information and communication technologies (ICTs) are ubiquitous in our private and business lives. Digitalization and ICTs generally facilitate work activities and enable new ways of work ). However, the use of ICTs may also create technostress, contributing to individuals' overall experience of stress. Thereby, technostress describes any stress from using ICTs (Ragu-Nathan et al. 2008). It has become a severe issue with societal, economic, and personal consequences such as impaired individuals' health or decreased work productivity (Ragu-Nathan et al. 2008;Srivastava et al. 2015). While technostress can also have positive effects-techno-eustress (Benlian 2020; Califf et al. 2020;Tarafdar et al. 2019)-we focus on the negative side, techno-distress (in this paper referred to as technostress), which is prevailing in literature (e.g., Pirkkalainen et al. 2019;Weinert et al. 2020).
Technostress is mostly described as a process, starting with technology environmental conditions that refer to attributes of specific ICT (e.g., push notifications) and represent a demand to the individual. The individual next appraises if the demand demonstrates a threat or a challenge, indicating a technostressor. This then leads to a technostress response (i.e., physiological, psychological, and coping responses). In the long term, the experienced technostress response can lead to serious adverse technostress outcomes including severe health impairments, such as burnout, depression, or exhaustion (Maier et al. 2015). Technology environmental conditions do not necessarily lead to an adverse course of the technostress process. There exists a variety of actions that can mitigate the adverse technostress outcomes throughout the technostress process. It is, therefore, crucial to not only react after negative technostress outcomes already arose.
There is a stream of literature on technostress inhibitors (Ragu-Nathan et al. 2008;Tarafdar et al. 2010Tarafdar et al. , 2011Tarafdar et al. , 2015Fuglseth and Sørebø 2014;Chandra et al. 2019). Technostress inhibitors are "organizational mechanisms that have the potential to reduce the effects of technostress" (Ragu-Nathan et al. 2008: 422). Examples of technostress inhibitors include technical support provision or literacy facilitation (Ragu-Nathan et al. 2008). We build on the literature on technostress inhibitors, combine it with preventive stress management, and develop the notion of preventive technostress management.
The Theory of Preventive Stress Management can help to understand how throughout the technostress process, the experience of technostress can be reduced (Hargrove et al. 2011;Quick et al. 1997;Quick and Quick 1979). The theory includes a time perspective on different time windows for preventing stress. The theory recommends intervening throughout the process and divides possible measures into three categories: primary, secondary, and tertiary preventive measures (Quick et al. 1997;Quick and Quick 1979). Transferring the Theory of Preventive Stress Management into the technostress world, preventive technostress management includes primary prevention targeting the technostressors, thus focusing on the cause of technostress (Hargrove et al. 2011), and secondary prevention targeting the technostress response, thus including actions designed to improve the individual's response (e.g., improving coping skills). Tertiary prevention aims to treat the adverse technostress outcomes and targets the very end of the process. By contextualizing the Theory of Preventive Stress Management to the narrower context of technostress following guidelines of Hong et al. (2014), we transfer previously generated insights in prevention theory to the technostress domain as a novel, complementary viewpoint to technostress inhibition.
Prior research in the context of preventive stress management emphasizes the role of organizations in preventing individuals' stress and promoting employees' and organizations' well-being (Hargrove et al. 2011). Since many technostressors stem from ICT use at work, organizations face the moral and legal responsibility to improve employee health by preventing excessive work-related technostress. Some countries, such as Germany, have even imposed legal requirements for organizations to assess and reduce employees' negative psychological responses (e.g., caused by technostress) at work. 1 Not only does the prevention of technostress reduce negative consequences for employees' health, but it saves organizations the costs of substituting employees on sick leave, among others. An analysis and comprehensive knowledge base of technostress prevention measures an organization can introduce are needed to enable organizations to reduce technostress among the employees proactively. So far, prior research has analyzed the influence of single measures (Day et al. 2012;Valta et al. 2021) or technology characteristics (Ayyagari et al. 2011;Becker et al. 2020) on selected technostressors. Given the severity of technostress' adverse outcomes, however, research needs to provide organizations with guidance on what actionable measures exist, which they can implement to prevent technostress for their employees (Brivio et al. 2018). Hence, our research aim is to (1) identify technostress prevention measures by bringing together different strands of research, and (2) characterize them in terms of their basic approach to preventive technostress management, their applicability in practice, and their relevance in targeting technostressors.
To achieve this aim, we synthesize and advance a knowledge base of technostress prevention measures by bringing together the research strands of technostress inhibition, further technostress literature, and stress management. We use the Theory of Preventive Stress Management as a theoretical basis that we apply to technostress. We assess characteristics for the applicability of the technostress prevention measures in practice. To do so, we conduct a structured literature review on organizational measures that can prevent technostress and reframe them into actionable technostress prevention measures. We enrich the resulting list by conducting multiple focus group workshops followed by a Delphi study with industry experts, yielding 24 validated prevention measures. Based on the experts' assessments, we produce a description for each prevention measure, a classification of several characteristics, and, in the case of primary prevention measures, an indication of the technostressors they are expected to target.
Our study advances both technostress theory and practice. We contribute to technostress literature by embedding technostress inhibitors in the larger context of preventive technostress management and by transferring the Theory of Preventive Stress Management into the specific context of technostress. Second, we provide an overview of technostress prevention measures in the research strands of technostress inhibition and related technostress literature, and develop new technostress prevention measures based on experts' insights. Third, we enrich the understanding of all 24 measures through the characterization of their basic approach to technostress prevention (primary vs. secondary technostress prevention) and their applicability in practice. Therein, we offer a basis for understanding the different roles of prevention and provide an overview of actionable measures based on insights of related research strands. The characterization in terms of the measure's applicability highlights the need to carefully select measures that fit the specific organizational context. Finally, our study sheds light on the dynamics underlying technostress prevention and links specific primary measures to specific technostressors, revealing that single prevention measures are no one-fits-all solutions. Overall, by uniting different perspectives on technostress prevention, we contribute towards substantial benefits for organizations, individuals, and societies by potentially reducing healthcare costs and preventing adverse personal and organizational outcomes (Maier et al. 2015;Ragu-Nathan et al. 2008;Srivastava et al. 2015).

Theoretical background on technostress as a specific form of human stress
In their daily life, people use a large variety of ICTs, including devices such as smartphones or laptops and applications that facilitate business processes by providing tools for inter-and intra-organizational communication and collaboration (Dittes and Smolnik 2019;Zuppo 2012). ICTs shape modern work life and contribute to many positive facets of work, such as the potential to work from home or seamless collaboration across countries. For example, during the COVID-19 pandemic, ICTs allow many companies to uphold their operation despite stay-at-home and social distancing orders (Ketter et al. 2020). However, the intensive use of ICTs also risks employees' health and performance. One of these risks is technostress. Technostress is a specific form of human stress that is triggered by the use of IS . Human stress has been extensively studied and is often explained through Lazarus and Folkman's (1984) Transactional Theory of Stress (Ayyagari et al. 2011;Ragu-Nathan et al. 2008;Tarafdar et al. 2019). The theory conceptualizes stress as a process that includes the existence of internal and external demands (e.g., time pressure, social conflicts), which the individual assesses in two appraisal steps. First, the individual subconsciously evaluates if the demand falls into the category of positive, irrelevant, or stressful. If categorized as stressful, the demand may threaten the individual, indicating a stressor. Next, the individual subconsciously examines if the available resources are sufficient to cope with the stressor (Lazarus and Folkman 1984). The process is therefore followed by an interdependent cycle of physiological and psychological responses (e.g., negative emotional states) and coping (i.e., behavior in response to stressors) that run in parallel, repeatedly, and cannot be separated in time (Califf et al. 2020;Lazarus and Folkman 1984;Tarafdar et al. 2019). This process, in turn, may lead to adverse outcomes such as decreased health or lower productivity. Transferring the Transactional Theory of Stress into the digital world, Tarafdar et al. (2019) describe technostress as the "stress process activated due to the use of IS" (p. 8). Like stress in general, technostress can be positive (techno-eustress) or negative (techno-distress) ). In the following, we focus on the negative side of the technostress process and specifically the prevention of that negative side of technostress. The technostress process is sketched in Fig. 1 based on Califf et al. (2020) and Tarafdar et al. (2019): Technology environmental conditions include potential sources of a ICT-related stressful situation. The individual exposed to such a demand from the environment appraises if the demand is harmful. If the demand is appraised as harmful it is a technostressor. Technostressors are "conditions or factors that can create stress because of ICT use" (Tarafdar et al. 2015: 106) and "are appraised by the individual as damaging" (Tarafdar et al. 2019: 9). When confronted with a technostressor, this leads to a multifaceted technostress response. The technostress response includes short-term physiological (e.g., the release of cortisol, adrenaline, and noradrenaline) and psychological (e.g., negative affect) responses that lead to coping responses (e.g., avoiding or stopping IS use). 2 Coping response describes "actions or emotions to overcome or deal with the threat or hindrance the individual perceives from the" technostressor (Tarafdar et al. 2019: 20). Examples include seeking social or technical support (Pirkkalainen et al. 2019;Schmidt et al. 2021;Weinert et al. 2020). Coping is undoubtedly crucial to mitigate technostress but is performed entirely by the affected individual only after the technostressor has emerged. In parallel, the individuals experience a psychological or physiological response like negative emotions (e.g., anxiety, hostility) or an increase in cortisol, or a fast heart beating (Califf et al. 2020;Riedl et al. 2012). The sequence of the psychological or physiological response and the coping response cannot be clearly separated in time and can run in parallel and repetitively. Therefore, we  Califf et al. (2020) summarize this interdependent cycle as a technostress response, consisting of psychological, physiological, and coping responses. We define technostress response as an interdependent, repeatedly, and potentially parallel process of negative psychological and/or physiological states caused by the technostressor (i.e., psychological and physiological response) and the application of actions or emotions to deal with the threat of the technostressor (i.e., coping response) [based on Califf et al. (2020), Tarafdar et al. (2019), Riedl et al. (2012), and Lazarus and Folkman (1984)]. In the long-term, the technostress response can lead to adverse technostress outcomes, especially when the technostress response is intense and long-lasting or frequently repeated. These include adverse outcomes for the individual, such as sleeping problems or emotional exhaustion -as well as adverse organizational outcomes like decreasing productivity, lower levels of job satisfaction, and less organizational commitment (e.g., Gimpel et al. 2018;Maier et al. 2015;Riedl et al. 2012;Srivastava et al. 2015).
Several technostressors have been discussed in the literature. Tarafdar et al. (2007) and Ragu-Nathan et al. (2008) were the first to develop and empirically validate scales for five technostressors: overload, invasion, complexity, insecurity, and uncertainty. 3 In subsequent research, these technostressors have been applied in many studies in different contexts and are well-established today (e.g., Becker et al. 2020;Fuglseth and Sørebø 2014;Maier et al. 2019;Pirkkalainen et al. 2019). In another seminal technostress paper, Ayyagari et al. (2011) identify unreliability, role ambiguity, and invasion of privacy as additional technostressors. Further, Galluch et al. (2015) and Tams et al. (2018) propose ICT-enabled interruptions as another technostressor in work environments. While various other technostressors have been suggested in the literature (e.g., Fischer et al. 2021;Maier et al. 2012Maier et al. , 2015Riedl et al. 2012), we did not consider them in our study because they are either specific to a particular technology, can be subsumed under another technostressor, or were published after the start of the Delphi study. The nine mentioned technostressors included in this research are presented in Table 1.

Applying the theory of preventive stress management to technostress
The Theory of Preventive Stress Management, originally developed by Quick and Quick (1979), has roots in preventive medicine and public health. Nowadays, it is well established (Quick et al. 1997;Hargrove et al. 2011). The theory provides general principles of preventive stress management and tells us that there exists a variety of specific actions that can reduce stress throughout the stress process (Quick and Quick 1979). We contextualize the Theory of Preventive Stress Management to technostress (Hong et al. 2014). Hong et al. (2014) suggest different approaches to incorporating context in theorizing. According to their typology, we engage in single-context theory contextualization with the domain of interest being technostress at work and the Theory of Preventive Stress Management as a well-established general theory. We consider concepts from both technostress theories and the Theory of Preventive Stress Management and explain their interrelations based on the underlying theory [level 2a contextualization from Hong et al. (2014)]. This leads us to the conceptual model of preventive technostress management presented in this section. In the subsequent sections, we decompose core constructs (specifically "primary technostress prevention measures" and "secondary technostress prevention measures") into contextualized constructs (i.e., 24 specific technostress-prevention measures; level 2c contextualization from Hong et al. (2014)). In that, we follow guidelines suggested by Hong et al. (2014): We perform a thorough literature-based and expert-based evaluation of the technostress context to identify specific factors. In that, we adapt the principles Table 1 Considered technostressors   Techno-stressor  Description Complexity "Situations where the complexity associated with ICTs makes users feel inadequate as far as their skills are concerned and forces them to spend time and effort in learning and understanding various aspects of ICTs" (Tarafdar et al. 2007: 315) Insecurity "Situations where users feel threatened about losing their jobs as a result of new ICT replacing them, or to people who have a better understanding of the ICT" (Tarafdar et al. 2007: 315) Interruptions Situations in which ICTs or ICT-based sources cause the user to shift their attention away from the task they are working on at that moment (Galluch et al. 2015) Invasion "The invasive effect of ICTs in terms of creating situations where users can potentially be reached anytime, employees feel the need to be constantly 'connected,' and there is a blurring between work-related and personal contexts" (Tarafdar et al. 2007: 315) Invasion of Privacy Situations in which users "are becoming increasingly concerned that their privacy could be invaded by computer technologies. The problem is acerbated due to the present work pressures, which create an unspoken value that appreciates individuals who are constantly available" (Ayyagari et al. 2011: 841) Overload "Situations where ICTs force users to work faster and longer" (Tarafdar et al. 2007: 315) Role Ambiguity Situations in which "there is uncertainty as to whether an individual should expend his or her resources to perform the task requirements at work or to acquire new skills. These competing demands between the job and learning new skills constrain individual abilities" (Ayyagari et al. 2011: 842) Uncertainty "Contexts where continuing changes and upgrades in an ICT unsettle users and create uncertainty for them, in that they have to constantly learn and educate themselves about the new ICTs." (Tarafdar et al. 2007: 315) Unreliability "System malfunctions and other IT-hassles" Riedl 2015: 1462) that are caused by ICTs that are perceived to increase workload due to the necessity to repeat tasks (Ayyagari et al. 2011) of preventive stress management to the technostress context and decompose core high-level constructs of the Theory of Preventive Stress Management. As our principles of preventive technostress management and the prevention measures are manifestations of general constructs from the Theory of Preventive Stress Management, their effects can be explained based on the theoretical rationale of the general theory. Quick and Quick (1984b) defined preventive stress management as "an organizational philosophy and set of principles that employ specific methods for promoting individual and organizational health while preventing individual and organizational distress." It is important to note that the Theory of Preventive Stress Management takes a broader view of stress than the technostress theory. As the definition of preventive stress management shows, stress is seen as a phenomenon at the individual and organizational levels. The Theory of Preventive Stress Management considers stress an overarching rubric for how individuals and organizations react and adjust to their environments. It becomes specific for the concepts of stressor, stress response, and stress. Quick et al. (1997, p. 6) distinguish between individual stress and organizational stress, with the latter being "the degree of deviation an organization experiences from a healthy, productive level of functioning." Stress is not bad per se -it can have positive and constructive outcomes that support performance. However, "excessive, prolonged, intense, or mismanaged stress at the workplace" result in physiological, psychological, and behavioral deviations from an individual's healthy functioning (Quick et al. 1997, p. 18). The aggregate of this individuallevel stress becomes organizational stress. This conceptualization of stress and, specifically, stress outcomes on the organizational level is uncommon for technostress research. However, it is compatible with technostress research suggesting that the organizational environment, including the technological environment, is an important determinant of technostress (e.g., Tarafdar et al. 2019) and it is in line with research on technostress inhibitors as organizational mechanisms to reduce technostress (e.g., Ragu-Nathan et al. 2008).
The Theory of Preventive Stress Management builds on five guiding principles which are applicable to technostress management (Quick et al. 1997 This paper's practical contribution relates to the second and third guiding principles in supporting leaders in selecting adequate technostress prevention measures. 4. Each individual and organization reacts uniquely to stress. This is well in line with technostress theories and suggests that preventive technostress management needs to be tailored to specific organizations and needs to allow flexibility for individuals. 5. Organizations are ever-changing, dynamic entities. This implies that preventive technostress management cannot be a one-time effort but needs to be constantly evaluated and developed to meet the interests of the organization and the employees. Beyond these principles, the Theory of Preventive Stress Management considers the stress process and a translated overlay which is composed of preventive interventions (Quick and Quick 1979;Quick et al. 1997;Hargrove et al. 2011). One fundamental premise of preventive medicine is that preventive measures can target each stage in the life history of a disease to slow, stop, or revert the progression of it. Applied to stress, this means preventive measures can address various points in the stress process: "Primary prevention is aimed at modifying the organizational stressors that may eventually lead to stress. Secondary prevention aims at changing individual stress responses to necessary demands. Tertiary prevention attempts to minimize the amount of individual and organizational stress that results when organizational stressors and resulting stress responses have not been adequately controlled" (Quick et al. 1997, p. 154).
Building on Quick and Quick's (1984b) definition of preventive stress management, we define preventive technostress management as an organizational philosophy and set of principles that employ specific measures to inhibit technostress to promote individual and organizational health. The set of five principles is described above. The measures are technostress prevention measures. Figure 2 positions the prevention measures along the technostress process. It is important to note that secondary and tertiary prevention are not alternatives but complements of primary technostress prevention (Quick et al. 1998). In general, excessive, prolonged, and intense technostress that will eventually lead to adverse outcomes at the individual level and, consequently, on the organizational level should be prevented as early as possible.  Califf et al. (2020), and Quick et al. (1997) Technostress inhibitors and technostress prevention are related. Prevention takes measures to inhibit technostressors, technostress responses, or adverse technostress outcomes. Comparing research on technostress inhibitors to the previously presented insights on preventive stress management shows that the terms are based on the same theorization in the technostress process and refer to avoiding technostress to promote individual and organizational health. As summarized by Sarabadani (2018), technostress inhibitors are theorized to either (1) act as antecedents to technostressors, hence reducing the technostressor itself (Jena 2015;Tarafdar et al. 2011Tarafdar et al. , 2015, (2) moderate the relationship between technostressor and outcomes (Ragu-Nathan et al. 2008), or (3) decrease the adverse outcomes (e.g., job satisfaction, sales performance, organizational commitment) directly (Jena 2015;Ragu-Nathan et al. 2008;Tarafdar et al. 2011). These three mechanisms of technostress inhibitors align well with the three stages of prevention (primary, secondary, and tertiary). The terminology of prevention has not been widely used in the context of technostress. Only Salo et al. (2017) used the foundation of LaMontagne et al. (2007), which is based on the Theory of Preventive Stress Management, to structure individuals' ways of reducing technostress without introducing the term technostress prevention. Salo et al. (2017) categorized five different individual measures into stressor reduction (e.g., modification of IT features), stressor toleration (e.g., modification of personal reactions to IT stressors), and recovery from strain (e.g., online/offline venting). Building on the work of LaMontagne et al. (2007), this classification is equivalent to the classification of primary, secondary, and tertiary technostress prevention measures. However, while Salo et al. (2017) study individuals' coping with technostress, preventive stress management theory emphasizes the duty of organizations to promote individual and organizational health and to minimize individual and organizational stress to create an organization in which their employees can thrive and produce (Hargrove et al. 2011;Quick et al. 1997). We proceed to use the term technostress prevention measure to build on the existing knowledge of preventive stress management. We integrate both existing knowledge on technostress inhibitors and the differentiation between primary, secondary, and tertiary preventive measures in the conceptual model of preventive technostress management (Fig. 2). The model is based on the technostress process of Tarafdar et al. (2019) and Califf et al. (2020) (Fig. 1), synthesized with the Theory of Preventive Stress Management in organizations (Quick and Quick 1979;Quick et al. 1997;Hargrove et al. 2011) (Fig. 2).
In analogy with general stress prevention (Quick et al. 1997), we characterize the three technostress prevention stages as follows: Primary technostress prevention is technostressor-directed. It targets reducing the frequency, duration, and/or intensity of one or multiple technostressors. The Theory of Preventive Stress Management refers to organizational stressors as it focuses on work stress. These organizational stressors result from task, role, physical, and interpersonal demands. Specifically for technostress, the organizational stressors are activated due to the use of IS (technostressors). These (techno-)stressor-directed interventions are located at the onset of the (techno-) stress process directly related to the (techno-)stressors and represent the most efficient and effective means of prevention (Quick and Quick 1979). One example of primary technostress prevention is the development of team norms or reachability rules to avoid work-related communication when not appreciated.
As primary prevention might not be effective for all individuals and in every setting, secondary stress prevention aims at improving how individuals respond to the respective stressor (Quick et al. 1997). Hence, secondary prevention is responsedirected. Secondary technostress prevention aims at improving the individuals' technostress responses consisting of coping responses and psychological and physiological responses. An example of a secondary technostress prevention measure is providing an ICT helpdesk. At this point, the differentiation between secondary technostress prevention measures and technostress coping becomes apparent: technostress coping refers to the behavior an individual adopts when affected by a technostressor (e.g., calling the ICT helpdesk or seeking social support). Secondary technostress prevention measures are organizational-level measures that aim at improving the individual coping skills or resources and positively influence the psychological or physiological responses, for example, by offering platforms to exchange experiences on ICT use, training to improve individual coping skills, or the setup of an ICT helpdesk. Secondary technostress prevention measures do not concern the actual execution of the coping response but aim to optimize the possibility of coping.
Tertiary technostress prevention concerns the treatment, compensation, and rehabilitation of adverse individual or organizational technostress outcomes (Hargrove et al. 2011;Quick et al. 1997). These interventions are outcome-directed. As they intervene after the onset of adverse outcomes (e.g., burnout or depression regarding individual health outcomes), they are at the individual level mainly related to medical or psychiatric treatments. This is why we focus on primary and secondary but not tertiary technostress prevention. Table 2 provides an overview of the key constructs of our conceptual model of work-related technostress prevention and their definitions.

Research process for identifying and characterizing primary and secondary technostress prevention measures
We conducted a Delphi study with 13 experts from research and practice ( Fig. 3) to achieve our research aim of (1) identifying actionable technostress prevention measures and (2) characterizing them in terms of their basic approach to technostress prevention, their applicability in practice, and in case of primary technostress prevention measures, their relevance in targeting technostressors.
To provide an appropriate starting point for the Delphi study based on theoretical and practical knowledge, we included a preliminary phase 0 to prepare a sound set of potential technostress prevention measures from literature on technostress prevention, inhibition and related technostress literature, that investigate measures to reduce the effect of technostress. Thereby, theoretical knowledge was drawn from a structured literature review (Webster and Watson 2002) on measures that may be qualified to prevent work-related technostress. To validate and expand the list of technostress prevention measures, we conducted two focus group workshops (Kitzinger 1995) with experts from research and practice before conducting phases 1-3 of the Delphi study. The central part of the Delphi study builds on these measure candidates to, again, validate, but especially to detail the technostress prevention measures. Since literature has not yet produced and documented extensive knowledge on the characteristics of technostress prevention measures, an exploratory qualitative research method is appropriate. Unlike other exploratory qualitative research methods, Delphi studies support explorative, consensus-seeking research such as problem identification, concept development, and prioritization and have become a popular research method in IS research (Okoli and Pawlowski 2004;Skinner et al. 2015). The

Phase 0: structured literature review and focus group workshops
We conducted a structured literature review to identify known measures for reducing and eliminating technostress, which can be interpreted and converted into technostress prevention measures (Webster and Watson 2002). In theory, coping is not part of primary and secondary prevention. We nevertheless included it as a search term because some scholars provide examples of coping measures that can be used as input to develop technostress prevention measures. One reason for the lack of clear concepts could be the novelty of technostress prevention within the research stream. As suggested by vom Brocke et al. (2015) and Webster and Watson (2002), the literature review comprised three phases: (1) literature search, (2) selection of relevant literature, and (3) analysis of the results. We defined a search string, databases, and inclusion and exclusion criteria (vom Brocke et al. 2015). As the literature on technostress prevention is highly dispersed and often described under different terminology, we chose a broad search string that considers adjacent concepts and literature from various disciplines (information systems, organizational science, health, and psychology). Therefore, we searched for topics, abstracts, titles, and keywords that matched the search string (("technostress" OR "techno stress" OR "digital stress") AND ("prevent*" OR "reduc*" OR "mitigat*" OR "overcome" OR "cop*" OR "inhibit*")) in the three databases Web of Science, PubMed, and AISeL as well as in the Journal of Business Economics. In total, we found 204 articles. After filtering for English, peer-reviewed, full research articles (n = 192), removing duplicates (n = 167), and adding previously known articles according to Larsen et al. (2019) our set consisted of 171 articles. Based on our selection criteria, we conducted a three-step selection process, including title, abstract, and full-text review (Levy and Ellis 2006;Okoli and Schabram 2010). The selection criteria included: (1) the article is within the domain of technostress, and (2) the article includes at least one recommendation for preventing technostress (at work/in organizations) following our definition of technostress prevention measures. After the title screening, 102 articles remained. The abstract screening excluded an additional 42 articles, resulting in 60 articles that went into detailed analysis. For details, please see Fig. 5 in the appendix. Finally, the full-text screening resulted in a list of 40 relevant articles out of which seven articles belong to the specific domain of technostress inhibition, while the remaining 33 articles result from the additional terms in the search string (Table 5 and Table 6 in the appendix).
Independently from the literature review, we conducted two focus group workshops with technostress and occupational health experts from research and practice ( Table 8 in the appendix) to potentially expand the list of technostress prevention measures. The workshops aimed to identify recent or modern concepts for dealing with technostress that are being applied in practice but are not (yet) embedded in literature. Therefore, in the first focus group workshop, eight experts from practice and four from research developed a set of technostress prevention measures guided by a moderator and a minute taker (Conklin and Hayhoe 2010). The participants' areas of expertise include information systems, psychology, and occupational health and safety. The workshop's procedure was inspired by Then et al. (2015). The participants collected and discussed possible technostress prevention measures that organizations can take by making targeted changes to technologies, organizations, or individuals. A second moderated workshop with five experts (four from practice, one from research) expanded on the first workshop's results. The participants were introduced to the technostressors to develop adequate technostress prevention measures. To guarantee the privacy and foster an open atmosphere, we refrained from recording and transcribing the workshops. Instead, we created a photo protocol by taking photographs of the final whiteboards and collected the focus groups' notes as field notes (Miles and Huberman 1994)-a valid qualitative data source in workshops (Ørngreen and Levinsen 2017). Based on this information, the moderators of the two workshops summarized the results and prepared them for the participants. The participants then had the opportunity to add to or adjust the summaries until all participants were satisfied.
With the two workshops, we collected 94 technostress prevention measure candidates on different levels of detail. From the 40 relevant literature articles, we extracted 34 distinct recommendations for preventing technostress (Tables 5 and 6 in the appendix). Two researchers jointly categorized the measure candidates, grouped them on the same level of detail, and fitted them to our definition of technostress prevention. This procedure yielded an initial list of 24 technostress prevention measures that went into the Delphi study. We used the detailed technostress prevention measure candidates to describe and exemplify the aggregated technostress prevention measures. We formulated the underlying inhibiting effect and a measure's description based on both input sources during this process.

Phases 1, 2, and 3: The Delphi study
Our study is a slightly modified ranking-type Delphi study (Schmidt 1997), a common Delphi study approach (Paré et al. 2013). The ranking-type Delphi study proposed by Schmidt (1997) consists of three phases: (1) brainstorming to discover the issues, (2) narrowing down to determine the most important issues, and (3) ranking the issues. Similar to other published Delphi studies (Paré et al. 2013), we merged the first and second phases because the literature review and the focus group workshops in phase 0 already structured the topic. Instead, we added a phase for characterizing each technostress prevention measure. Each phase consisted of two rounds to validate the previous round's result, which is appropriate for producing credible results (Skinner et al. 2015). Consequently, our Delphi study consisted of three phases: (

1) Validation and extension of technostress prevention measures (2) Characterization of technostress prevention measures (3) Identification and relevance assessment of primary technostress prevention measures that target specific technostressors
To ensure that the participants of our Delphi study are knowledgeable experts in technostress and are aware of the issue in the larger world around them (Delbecq et al. 1975;Keeney et al. 2006;Skinner et al. 2015), we applied the following selection criteria (Okoli and Pawlowski 2004): experts should (1) be responsible for, for example, occupational safety/medicine, psychological risk assessment/ operational health management, or human resources, (2) have experience in the field of technostress, stress management, or occupational health and safety, (3) have at least three years of work experience, and (4) be frequent users of ICTs themselves. These requirements ensured that all experts were familiar with technostress and possible technostress prevention measures. We ensured that all experts professionally deal with stress management programs (all work on general occupational stress management; some focus on technostress management). Not every expert had experience with all measures presented but at least with multiple technostress prevention measures. Thus, they can assess the measures, evaluate them, and estimate the implementation. Experts had the choice not to provide assessments on individual measures when they felt they had insufficient information or experience with the measure. But this option was never used in the Delphi study. We have no indication for systematic differences in the assessment of measures by the experts depending on the expert's first-hand experience with a measure as compared to only abstract knowledge about the measure. Such differences could have occurred in phases 2 to 3. In the progression of the Delphi study, the experts' assessments tended to converge, given even less basis to differentiate by experience. Hence, we report aggregate results from the entire panel.
To ensure broad topic coverage, we recruited experts from different industries and company sizes via our industrial network and several occupational health and safety events on the topic of technostress. We reached out to a total of 50 possible experts. Of those, 15 experts fulfilled the selection criteria and agreed to participate in our Delphi study. None of them participated in the focus group workshops in Phase 0. One participant missed the first round, and another dropped out after the first round. Thus, 13 experts completed each round, an appropriate panel size (Linstone and Turoff 2002;Paré et al. 2013). Table 8 in the appendix provides additional information on the panel.
The Delphi study took place via the online survey tool LimeSurvey and lasted five months (August to December 2020). In each round, we invited all experts to participate via e-mail and provided detailed instructions. To provide the experts with sufficient guidance, we e-mailed them their answers and suggestions from the previous round, an overview of the changes made based on all experts' suggestions, and descriptive information on quantitative evaluations after each round. Thereby, the experts could reflect on their opinion based on the aggregated results from all experts. Also, experts could provide free text feedback in addition to quantitative assessments. Through the free text feedback, we also provide the opportunity to state that one has had too little experience with the measure for assessing it. Following the suggestion of Strasser (2016), Okoli and Pawlowski (2004), and Skinner et al. (2015), participants were not known to each other by name, and we did not show the individual results of other panelists to ensure anonymity and avoid bias.
Phase 1 introduced the experts to the 24 technostress prevention measure candidates and suggested a description of each measure and their inhibiting effects. We asked the experts to rate each technostress prevention measure's relevance for preventing technostress on a scale ranging from 0 ("not relevant") to 6 ("highly relevant"). Literature on the Delphi method suggests to identify the key issues (Paré et al. 2013) or most important issues (Schmidt 1997) in the respective context. In our study, this means identifying the most relevant prevention measures. Relevance in general refers to (1) a relation between two entities in a context, (2) to an intention, and (3) to an assessment of the effectiveness of the relation regarding the intention (Cosijn and Ingwersen 2000;Saracevic 2007). In our study, the context is technostress at work, and the relation is between technostress prevention measures and parts of the technostress process. Specifically, primary prevention measures relate to technostressors, while secondary prevention measures relate to technostress responses (also see Fig. 2 for the underlying model). The intention is to prevent technostress. Effectiveness relates to the ability of a measure to reduce technostress, i.e., mitigate technostressors (primary prevention) or mitigate negative stress responses (secondary prevention). Hence, we define the relevance of a technostress prevention measure as the assessment of the measure's effectiveness in preventing the negative effects of technostress. The experts' judgment on relevance thus inherently relates to the expected effectiveness of a prevention measure. Directly assessing effectiveness is not possible in a Delphi study. Assessing expected or perceived effectiveness would in principle be possible in a Delphi study. However, using the term effectiveness might be perceived to suggest a level of precision of measurement that our Delphi study could not deliver. Hence, we asked the experts to assess the broader concept of relevance.
We queried qualitative feedback regarding each measure's evaluation and overall comprehensibility by providing optional text fields. Additionally, the experts were asked to review the entire list, emphasizing its completeness regarding the most relevant technostress prevention measures and adding further measures if necessary. For evaluating this phase, we examined the written qualitative feedback and the relevance assessment. We defined consensus on the technostress prevention measures' relevance as a state in which (1) the experts mention no further wording changes or new technostress prevention measures and (2) more than 75% of the experts assess each measure's relevance with a score of 3 (middle option) or higher. To avoid bias by individual experts, recommendations for description changes or additions of technostress prevention measures were only implemented if at least two experts made suggestions in that direction. The experts did not yet reach a consensus in the first round. Based on their feedback, two measures of the initial set were merged, and one new measure was added. In the second round, we presented the aggregated results of the first round to the experts. This round resulted in a consensus, yielding the final set of technostress prevention measures. To take a first step toward the technostress prevention measures' categorization, three authors separately grouped the technostress prevention measures into primary and secondary preventions based on the measures' descriptions. For each measure with different categorizations of the three researchers, they discussed them together until they agreed. Figure 4 summaries key aspects of this first phase and the following phases.
Phase 2 introduced the experts to different characteristics of technostress prevention measures through short explanatory descriptions and figures. The characteristics were developed by the research team with the help of practitioners from organizational health and safety and represent a valuable step to better understand technostress prevention. Each expert could indicate one manifestation for each characteristic. The six characteristics are: (1) Entity of change describes whether the technostress prevention measure alters an organization's I) technology, II) organizational structures and procedures, or III) the individuals' skills and abilities [based on Murphy and Sauter (2004)]. Technology, organizational structures, and organizational procedures relate to the technology environment conditions. Individual skills and abilities relate to the individual perceiving and assessing demands from the environment. Each of the three levels refers to the entity affected by the measure, which is always implemented by the organization to prevent technostress (in line with principle 3 of (techno)stress prevention; see Sect. 3).
(2) Target group describes whether the technostress prevention measure targets I) all employees, II) management only, or III) other specific groups. This characteristic again relates to Principle 4 of (techno)stress prevention management as the first step to account for the heterogeneity of individuals. Specifically considering management relates to principle 2 which places a specific responsibility on the leadership team that serves as a role model and determines essential aspects of the employees' work environment.
(3) Organization size determines if the technostress prevention measure is suitable for organizations with at least I) one or more, II) ten or more, III) 50 or more, IV) 250 or more, V) 500 or more, or VI) 2,500 or more employees. Principle 4 of (techno)stress prevention management suggests that not all organizations are alike. Including the organization size is one broad category in the direction of accounting for the heterogeneity of organizations. (4) Duration of implementation and time to operational use distinguishes between technostress prevention measures that take (I) less than 1 year, I(I) 1-3 years, or (III) more than three years to become operational. While the principles of (techno)stress management suggest that stress is not inevitable (principle (3) and that leaders should select and implement technostress prevention measures, this process cannot happen instantaneously. Characterizing the duration between the selection of a prevention measure to be implemented until this measure is operational captures one element of time lack in realizing the effects of technostress prevention management. (5) Time until the effect of the technostress prevention measure is realized in the operational business can take (I) less than half a year, (II) 0.5-1 year, or (III) more than one year. Individual and organizational health are interdependent (principle 1), and both individuals and organizations react (principle 4). However, this reaction is not instantaneous upon the use of prevention measures but may take time. Characterizing the time from the operational use of a prevention measure until the effects of the measure are realized on the induvial and organizational levels is the second element of time lack in realizing the effects of technostress prevention management. (6) Duration of the technostress prevention measure's effect may be (I) less than 1 year, (II) 1-3 years, or (III) more than three years. A technostress prevention measure implemented may have long-term effects but the effect may also fade out. One reason is that organizations are ever-changing, dynamic entities (principle 5). Hence, characterizing the duration of the effect aims at informing technostress prevention management regarding the likely need to repeatedly reassess the state of technostress and its prevention in an organization.
In both rounds of Phase 2, we evaluated the results based on the characteristics' mode (for nominal characteristics 1 and 2) and the characteristics' median (for ordinal characteristics 3, 4, 5, and 6). For characteristics with a nominal scale, the consensus is reached if 75% of the experts agree on the same characteristic (i.e., the mode). For characteristics with an ordinal scale, we defined consensus as an average absolute deviation from the median below 0.5. To further evaluate the consensus between the experts, we calculated Fleiss' Kappa for all six characteristics. The received values range from 0.01 to 0.43 in round 1 of Phase 2. Here, experts had not yet consensus. In round 2 of Phase 2, Fleiss' Kappa increased to 0.79-1.00. Except for organization size (0.79, substantial agreement), all Fleiss' Kappa values indicate an almost perfect agreement (Landis and Koch 1977). Again, consensus on the characterization of the technostress prevention measures was reached after two rounds.
In Phase 3, the experts were given a definition and an illustrative example of each technostressor as presented by Gimpel et al. (2020). The experts then selected the prevention measures expected to be most relevant for mitigating each technostressor. In the following, we discuss the relevance of prevention measures for specific technostressors. As the results are based on expert assessments in a small sample, we do not indicate absolute and quantifiable relationships but use relevance as a measure for giving a trend statement on the expected relationship. The trend statements serve as the foundation for potential statistical analyses in the future.
For primary technostress prevention measures, this means that the measure is expected to reduce the corresponding technostressor directly. For secondary prevention measures, we do not present this relationship because one technostressor can trigger several technostress responses and one technostress response can be triggered several technostressors. Therefore, the association of the relevance relationship between secondary prevention measures and technostressor is not meaningful. Based on the results, we assess both the set of primary technostress prevention measures relevant for each technostressor as well as the set of technostressors to which a single primary technostress prevention measure is expected to be relevant. We asked the experts to assess the relevance of the selected technostress prevention measures for the specific technostressor on a scale ranging from 0 ("not relevant") to 6 ("highly relevant"). To evaluate the connections between technostressors and technostress prevention measures, we counted how often each technostress prevention measure was selected for each of the nine considered technostressors. We declared consensus when five or more technostress prevention measures were selected by more than 75% of the experts (that is, at least ten experts). In the second round, all technostressors met this criterion. Accordingly, the Fleiss' Kappa values indicating the degree of agreement between the experts for each of the nine considered technostressors increased from 0.09 to 0.33 in round 1 to 0.72 (substantial agreement) to 0.94 (almost perfect agreement) in round 2 of Phase 3.
Lastly, based on our overall results, we present propositions that contain insights gained through applying the Theory of Preventive Stress Management to Table 3 Technostress prevention measures that are sorted by their basic approach to technostress prevention-the column 'Inhibition?' indicates if the measure has been mentioned in any technostress inhibition article, '#' indicates how many times the measure (e.g., input that helped to formulate the measure) was mentioned in the litera-  Table 3 presents the complete list of the 24 technostress prevention measures. The column "Inhibition?" indicates whether we extracted this measure from technostress inhibitor literature. The seven technostress inhibitor articles (see Table 5 in the appendix for details) each suggested two to seven technostress prevention measures with some overlap among the articles. Overall, twelve prevention measures mainly result from technostress inhibition literature. An additional eight measures result from the remaining articles we reviewed during Phase 0. Finally, four measures are self-developed during our research process bases on experts' feedback. Examples for measures that result, among others, from technostress inhibition article by Tarafdar et al. (2015) is measure 19 provide ICT support or measure 23 provide ICT training. An example for measures that resulted from remaining literature besides technostress inhibition literature, is measure 2, adapt a stress-sensitive digital workplace design. Measure 2 results from prevention measure candidates from Arnetz (1996) and Lee et al. (2014), who do not refer to technostress inhibition but still provide valuable insights for the development of technostress prevention measures. The column "#" indicates how many times the technostress prevention measure was mentioned in the literature; "-" indicates that the technostress prevention measure was self-developed. The four self-developed measures are measure 1, focus the ICT landscape, measure 5 use gamification, measure 10 agree on binding ICT usage guidelines, and measure 21 train technostress coping competencies. While measures 1, 5, and 10 were developed during the workshops in Phase 0, measure 21 was added based on experts' suggestions in Phase 1 of the Delphi study.
Measure 1 refers to the situation of having several redundant systems for the same task (e.g., video conferencing) and information distributed in the organization. The measure aims at reducing the number of systems to a reasonable level and especially involves employees in selecting appropriate systems and ICTs for the given task (e.g., to also offer room for practices like "bring your own device" if this simplifies the process and collaboration). Measure 5 use gamification targets the ICT design by including game elements, for example, levels, points, rewards, or badges. ICTs might also react with humor in certain situations. In this way, users are introduced to ICTs through playful behavior. It is important to note that through the Delphi study, we found that participation in such games should be voluntary to avoid (perceived) performance monitoring. Measure 10 agree on binding ICT usage guidelines goes in a similar direction as measure 1 but emphasizes the importance of transparency on which ICT is used for which purpose and under what conditions. Measure 21 train technostress coping competencies is also highly relevant in the Theory of Preventive Stress Management and separates the constructs of coping from preventive technostress measures (Quick et al. 1997). Sufficient coping skills are highly important to effectively deal with technostress (Salo et al. 2020;Schmidt et al. 2021;Weinert et al. 2020). The preventive measure trains employees' individual technostress coping behavior being part of their technostress response. Coping includes, for example, positive reappraisal, changing the perception of an IT event, or seeking help from colleagues Pinsonneault 2005, 2010).
In comparison to the technostress prevention measures synthesized from literature during phase 0, we merged the two technostress prevention measure candidates foster sensitization regarding technostress and foster self-reflection regarding technostress into foster sensitization and self-reflection regarding technostress based on experts' input during the Delphi study (measure 24, Table 3), and also added a new measure 21, as mentioned above. Besides these apparent changes to the list of technostress prevention measures, the experts' qualitative feedback in the Delphi study's first phase suggested changes concerning the configurations of single technostress prevention measures. One expert, for example, pointed to the importance that, "especially for inexperienced employees with an IT problem, the IT helpdesk [measure 19, provide ICT support; note from the authors] staff should convey a feeling of trust and respect". Many of these suggestions are reflected in the descriptions of the measures. In addition, multiple experts stressed that organizational offers suggested by measures 17 (train effective self-management and time management) and 24 (foster sensitization and self-reflection regarding technostress) should not be mandatory, as that might even increase employees' stress.
In addition, various experts emphasized the relevance of a diverse portfolio of measures. Although measures targeting individual-level change are important, these measures bear the risk of outsourcing the responsibility for technostress prevention from the organization to the individual. Therefore, the complementary implementation of organizational and technical changes is essential. The 24 technostress prevention measures give a comprehensive overview of available measures in organizational settings and indicate the diversity of technostress prevention opportunities. The list of technostress prevention measures is structured along with the differentiation into primary and secondary prevention and the distinction between measures enabling changes to technologies, organizations, or individuals. The overview of references referring to the measures is presented in the appendix in Tables 5 and 6.

Characterization of technostress prevention measures
To better structure and compare the different types of technostress prevention measures, experts characterized them on a set of characteristics during the Delphi study. Each technostress prevention measure's complete characterization is presented in Table 9 in the appendix. The first characteristic (entity of change) relates to the entity affected by the technostress prevention measure and comprises the technological, organizational, and individual levels (Murphy and Sauter 2004). This differentiation stems from Murphy and Sauter's (2004) effort to structure interventions that avoid negative influences on worker health and safety. Measures at the technological level concern the implementation and use of well-designed ICTs that serve their purpose. At the organizational level, measures focus on changing organizational structures, processes, and guidelines (e.g., code of conduct and operating instructions). Lastly, changes on the individual level comprise technostress prevention measures that create a change in the individual, for example, their behavior. The manifestations of this characteristic are distributed relatively evenly across the 24 measures. The subheadings in Table 3 structure the classification (e.g., measures underneath the subheading "Enabling Changes to Individuals" refer to measures that affect the individual level). Six of the 24 measures address the technological level (e.g., measure 3, apply humancentered release management), eight address the organizational level (e.g., measure 12, establish reachability management), and ten address the individual level of prevention (e.g., measure 17, train effective self-management and time management). Technostress prevention measures on all three levels (technological, organizational, individual) can act as primary or secondary prevention (Murphy and Sauter 2004; Pirkkalainen et al. 2019;Salo et al. 2017;Weinert et al. 2020). However, the characterization shows very different frequencies of the three levels when comparing the two types of prevention. For primary prevention, the technological level, the organizational level, and the individual level are addressed by an about even number of primary prevention measures (5, 7, and 5 respectively). In contrast, secondary technostress prevention measures primarily address the individual (1, 1, and 5 measures for the technological, organizational, and individual levels, respectively).
Primary technostress prevention measures on the technological level alter the technological environment in such a way that employees experience fewer technology environmental conditions perceived as demanding. Complementary, secondary technostress prevention measures on the technological level refer to measures that aim at providing employees with technological resources to improve their response to technostress (e.g., helping employees by reminding them to take breaks). Similar accounts for primary and secondary prevention measures that target the organizational environment. While individual-level secondary technostress prevention measures mainly address the techno-stressed individual's internal factors (e.g., knowledge, skills, experience), individual-level primary technostress prevention measures primarily target a change of other individuals' behaviors, thus, shaping the social environment.
Target group and organization size concern the applicability of the measure for specific people and organizations. On the characteristic of the target group, we observe no fundamental differences between primary and secondary technostress prevention. In terms of the appropriate organization size, nine measures are assessed to be suitable for organizations with as few as ten employees, 13 measures for at least 50 employees, and two measures require at least 250 employees to be conducted effectively. While most measures in our study (21) are seen as suitable to all employees, three are relevant to management only. Interestingly, most primary technostress prevention measures (13/17, ~ 76%) are deemed applicable only for organizations with at least 50 employees. With secondary prevention, five of seven measures (~ 71%) are already applicable with as few as ten employees. This indicates that the build-up of employee resources to better react to technostressors (i.e., secondary prevention) is often already feasible with only a few individuals. In contrast, measures initiating large-scale technological and organizational changes require more advanced organizational structures and larger organizations.
Lastly, the characteristics of implementation duration, time from implementation until effect realization, and effect duration take a time perspective. The results suggest that half of the measures can be implemented and brought to operational use in less than one year (e.g., measure 16, train mentors for digital topics). In contrast, eleven measures require 1-3 years, and one measure even three years (measure 11, Develop team norms for the use of ICTs). These numbers indicate that the initial effort for many measures is seen as relatively low, which is important for reducing the barrier to successful prevention. Similarly, the time required until the effect of the respective measure becomes apparent in operational business is estimated as less than half a year for ten measures (e.g., measure 21, train technostress coping competencies), up to one year for 13 measures, and more than one year for only one measure (measure 6, foster a cooperative culture). The effect of six technostress prevention measures is characterized to last less than one year, while 18 measures show a positive effect between one to three years (e.g., measure 1, focus the ICT landscape). Regarding primary vs. secondary technostress prevention measures, a clear tendency can be observed. The average implementation duration and time until effect realization are longer for primary than secondary technostress prevention. However, the effect duration for primary prevention was assessed to last 1-3 years for 14 of the 17 measures (~ 82%), while of the secondary technostress prevention measures, only four of seven measures (~ 57%) last 1-3 years and the rest less than one year.
Next, we asked the experts for a relevance assessment on which primary technostress prevention measures are expected to target which technostressor. For each technostressor, we identify two to six measures that the experts assessed as the most relevant. The resulting scores serve as a basis for confirmative quantitative statistical analyses on the relevance of different technostress prevention measures on the set of technostressors. Table 4 presents these relations and graphically displays the average relevance (scale from 0 to 6) of the measures for the respective technostressor for all primary technostress prevention measures. The measures are sorted according to their accumulated relevance score. Table 10 in the appendix presents the complete list of numerical relevance ratings and the number of experts who assessed the measure as highly relevant.
When assessing the relations, no two technostressors share the same set of relevant technostress prevention measures. However, some patterns exist. For example, two technostress prevention measures can potentially help prevent complexity, unreliability, and role ambiguity. Both prevention measures (numbers 4 and 1) target the technological environment. When they reduce the complexity and unreliability of the technology, the need for employees to acquire new skills for working with complex and unreliable technology (and, hence, role ambiguity) are reduced. Similarly, interruptions and invasions share multiple prevention measures. Accordingly, these pairs of technostressors can be addressed through similar prevention measures.
Each technostress prevention measure is selected as a technostressor's most relevant measure between zero to four times. Interestingly, two primary technostress prevention measures (measure 14, train managers for leading distributed team members, and measure 7, develop a mission statement for digital collaboration) are not selected to be among the most relevant for any technostressor. This result does not mean that the measures are generally not relevant for preventing technostress. However, it does indicate that industry experts do not expect them to be among the most relevant measures for addressing one of the nine considered technostressors. In contrast, two technostress prevention measures (measure 4, apply human-centered ICT design, and measure 13, train managers to successfully lead in the digital working world) are assessed as very or highly relevant to four different technostressors. This finding indicates that the two are rather general technostress prevention measures that are suitable for multiple technostressors. In that, measure 4 relates to changing the technological environmental conditions and measure 13 relates to the prominent role of leadership in preventive stress management (i.e., principle 2 of (techno)stress management). Other measures (e.g., measure 9, agree on binding ICT usage guidelines, and measure 10, consciously manage ICT-related change) are only relevant to one technostressor. These measures are expected to be specialized in addressing a particular technostressor.

Propositions on preventive technostress management
Based on our results, we provide five propositions on preventive technostress management. The propositions are an abstraction of the results presented above. We see propositions as proposed relationships between constructs on the theoretical plane as compared to hypotheses on relationships between variables on the empirical plane (Bhattacherjee 2012, Chapter 2). The propositions are theoretical in the sense of relating to the understanding of a theory as relationships among constructs along with arguments substantiating the relationships and a definition of the scope of validity of the relationships (Whetten 1989). The purpose of the propositions here is to provide statements on potential relationships between technostress prevention constructs that stem from our empirical results and are compatible with existing theory on technostress and preventive stress management. The propositions are tentative and conjectural relationships between constructs; they need further theorizing and, after being transferred to research hypotheses, empirical testing for being fully integrated in theories of technostress. The propositions relate to abstract constructs on the theoretical plane. Hence, they cannot be tested empirically directly. Nevertheless, the propositions can serve as a foundation for developing testable hypotheses in the form of relationships between variables on the empirical plane (Bhattacherjee 2012, Chapter 2). The empirical test of hypotheses derived from the propositions is an indirect test of the propositions that will allow to reflect on their truth and usefulness. Once validated through the corresponding empirical test of related hypotheses, the propositions may serve as basis for designing preventive technostress management tools, methods, and programs.
For all following propositions, the scope is technostress in a work context with the organization as the entity that can prevent technostress among its employees. The arguments substantiating the proposed relationships derive from the

Proposition 1
The five principles of preventive stress management also apply to technostress. Quick and Quick (1979) intensively discuss work stress and its prevention. Given their publication before the widespread dissemination of digital technologies to work places, technology is not explicitly in focus. Quick et al. (1997) however do shortly discuss the role of early production technologies and a mismatch between people, task, technology, and structure as factors to consider in work stress research. Therein, the principles of preventive stress management are expected to apply to technostress, as technostress is a specific form of general stress. As such, the proposition is straightforward and in line with existing theory. Nevertheless, we suggest that stating it explicitly is important to direct the attention of technostress researchers towards (principle 1) the interdependence of individual and organizational health, (principle 2) the responsibility of leaders for individual and organizational health, (principle 3) the ability to prevent (or inhibit) individual and organizational stress, and (principle 5) the need for constant adaptation of preventive technostress management to keep up with ever-changing and dynamic organizations (Quick et al. 1997). Additional empirical evidence for the relevance of principles 2 and 3 can be found in the presence of prevention measures solely relevant to leadership (measures 13-15, principle 2), as well as the overall positive Delphi expert assessment of many prevention measures regarding their effectiveness in preventing technostress (principle 3). Principle 4 refers to the individuality of organizations and persons in their reaction to technostress (Quick et al. 1997). We do not see a need to direct attention to this principle as it is already front and center in technostress research (Beaudry and Pinsonneault 2005;Chen et al. 2019;Pirkkalainen et al. 2019;e.g., Salo et al. 2020).

Proposition 2 Technostress prevention measures address the technology, the organization, and/or the individual. Secondary technostress prevention measures focus primarily on the individuals.
The differentiation of addressing technology, organizations, and/or individuals in prevention derives from literature (Murphy and Sauter 2004;Quick et al. 1997). Historically, the focus was mainly on distinguishing individuals and organizations (Quick et al. 1997). However, with the digitalization of work, the specific attention to technologies as a central component of work stress has emerged (e.g., Ayyagari et al. 2011;Ragu-Nathan et al. 2008;Tarafdar et al. 2019). This general differentiation is empirically supported by the developed set of technostress prevention measures in this study, given that it spans all three categories. The focus of secondary prevention on individuals is primarily induced from the characterization of the technostress prevention measures in this study. Five of the seven (71%) secondary prevention measures concern the individual level. Further, this insight integrates well with existing stress prevention theory. Secondary technostress prevention targets individuals' technostress responses, i.e., physiological, psychological, and behavioral responses of individuals experiencing technostress (Quick et al. 1997). The behavioral response might be influenced by the technology or organization. However, physiological and psychological response are happening within the individual -hence the focus of respective prevention measures on individuals (LaMontagne et al. 2007;Salo et al. 2017). Omissions in proposition 2 are also noteworthy: The proposition does not refer to addressing the environment, customers, the employees' family or the like. Thus, the proposition suggests a clear focus for identifying and designing further technostress prevention measures.

Proposition 3 Primary prevention measures differ with respect to the number of technostressors they are relevant for.
Focusing on the relationship between technostressors and primary technostress prevention measures, we found patterns regarding the relevance of primary technostress prevention measures to reduce the frequency, duration, or/and intensity of technostressors. Our results indicate that not every measure addresses all technostressors equally. Some measures are relevant for one or a few specific technostressors, while other measures target a broader set of technostressors. For example, measure 8 (Introduce an employee data security concept) is explicitly concerned with increasing data security and reducing related risks. The measure is highly focused on one specific downside of digitalized workplaces and, as a result, targets only one technostressor: Invasion of privacy. In contrast, measure 13 (Train managers to successfully lead in the digital working world) addresses the very general topic of digital leadership. Strong digital leaders can help improve technostress stemming from all types of technostressors by identifying sources of technostress in time. The measure thus affects several technostressors (i.e., insecurity, invasion, invasion of privacy, and uncertainty). Table 4 gives the first indication of this matching based on the Delphi study's expert assessment, which requires future empirical research to verify. While existing research has not yet systematically matched the full list of technostressors with the list of technostress prevention measures, few studies have investigated the effectiveness of some subset of technostress prevention measures for certain technostressors (e.g. Galluch et al. 2015;Schmidt et al. 2021;Valta et al. 2021). Their work supports the proposition that technostress prevention measures support different numbers of technostressors. For example, Valta et al. (2021) demonstrate social support systems and contact persons as measures addressing many technostressors. In contrast, the reduction of e-mail traffic is shown to be very useful only for two of their technostressors. In conclusion, we see a need for future research to go beyond our Delphi study's relevance assessment and empirically establish the effectiveness of the different measures. Supporting technostress prevention in practice will require developing a body of knowledge on how to best select and implement a portfolio of technostress prevention measures which addresses the most important technostressors in a given organization.

Proposition 4 Technostressors differ with respect to whether they are addressable through only a few or many primary technostress prevention measures.
Considering technostressors as the baseline (columns in Table 4), a few observations emerge from the empirical results in our study. Not all technostressors share the same number and relevance levels of technostress prevention measures through which they can be targeted. Consequently, some technostressors might be more difficult to address, requiring highly specialized measures for prevention. Others, however, exhibit a larger set of measures as potential prevention tools. For example, invasion is indicated to be addressable through several different measures. The technostressor addresses a vague feeling of having to be connected and there are many ways to change this perception. Unreliability, on the other hand, is only matched with three measures in our study. As unreliability describes technostress from technical failure, the options for prevention are more limited: improve technical reliability or improve coping with failures. Complementing our observations, Valta et al. (2021) has highlighted results that support this hypotheses. In their study, for example, overload is demonstrated to be addressable by multiple technostress prevention measures in their sample, while cyberbullying is only addressed by one. As with proposition 3, we deem it important to point to this multifaceted interplay of primary technostress prevention measures and technostressors to spur future research in the direction of providing knowledge on managing a portfolio of technostress prevention measures.
Proposition 5 Compared to secondary prevention, primary technostress prevention measures represent a longer-term approach with a longer effect duration, but also require higher initial efforts.
Deep-diving into the characteristics of primary and secondary prevention measures in Table 8 of the appendix, we identified that primary measures tend to require higher average initial efforts for implementation and effect realization than secondary prevention measures. While for primary prevention only six out of seventeen (35%) technostress prevention measures require less than one year for implementation, six out of seven (86%) secondary technostress prevention measures were characterized with the shortest implementation time category. Simultaneously, however, the experts in our study indicated that primary technostress prevention measures yield longer-term positive effects. This leads to the proposition that primary technostress prevention is more focused on a long-term approach to preventing technostress, while secondary technostress prevention measures on average present a shortand mid-term solution to preventing technostress. Existing literature has not yet analyzed the differences between effect and implementation duration when comparing primary and secondary prevention. However, the nature of both prevention types offers some arguments supporting this proposition. Secondary prevention measures mostly address individuals and their response to technostressors, including their skills, competencies and experiences (Salo et al. 2017). Trainings and workshops that improve an individual's coping or IT skills can be conducted within a matter of hours or days thus leading to a low implementation and effect duration effort. On the downside, high employee turnover rates, particularly in IT professionals, lead to a frequent loss of knowledge and skills gained from secondary technostress prevention measures and reduce the effect duration period in this prevention type (Wang et al. 2022). Primary prevention measures on the other side often target the technological and organizational environment (Quick et al. 1997;Salo et al. 2017). Therein, changes stemming from prevention measures often affect overarching organizational and technological structures comprising many people and entities. While this makes their implementation more complex and time-consuming, it also reduces the dependency on individuals, thus increasing the effect duration period.

Discussion
Reducing the adverse outcomes of technostress is essential. In an ideal world, organizations invest in addressing technostress at an early level and throughout the process (Brivio et al. 2018). To sustainably reduce technostress, organizations need to implement ex-ante measures to prevent future stressful situations caused by technostressors, and they need to support their employees in responding to technostress. To drive knowledge on prevention in the technostress field and assist practitioners in implementing successful technostress prevention management, we bring together different research strands of technostress inhibition and stress management and enrich the combined perspective empirically. With our study, we follow Tarafdar et al.'s (2019) call for 1) a more thorough investigation of how altering technological aspects in an organizational environment can prevent technostress and 2) applying a methodological approach in technostress research that complements the current focus of technostress research on surveys.

Contributions and implications for research
The study contributes to research in four ways: 1. Embedding technostress inhibitors in the larger context of preventive technostress management 2. Providing a structured overview of 24 technostress prevention measures 3. Characterizing of the 24 technostress prevention measures in terms of their basic approach to preventive technostress management, their applicability in practice, and their relevance in targeting technostressors. 4. Formulating propositions on preventive technostress management First, we embed existing knowledge on technostress inhibition in the larger context of preventive technostress management by bringing together theory on technostress inhibitors (e.g., Ragu-Nathan et al. 2008;Sarabadani 2018) with the three types of stress prevention (primary, secondary, tertiary) from the Theory of Preventive Stress Management. The research strand of technostress inhibition theorizes inhibitors to either (1) act as antecedents to technostressors (Jena 2015;Tarafdar et al. 2011Tarafdar et al. , 2015, moderate the relationship between technostressor and outcomes (Ragu-Nathan et al. 2008), or decrease adverse outcomes directly (Jena 2015;Ragu-Nathan et al. 2008;Tarafdar et al. 2011). We synthesize this existing knowledge and combine it with insights of the Theory of Preventive Stress Management which are very similarly located in the general stress process as the three theorizations of technostress inhibition. For example, Tarafdar et al. (2015) integrates technostress inhibitors as antecedents to technostressors, including the three inhibitors: "Facilitate technical literacy", "provide technical support", and "facilitate technology involvement". While Ragu-Nathan et al. (2008) and Fuglseth and Sørebø (2014) mention the same three technostress inhibitors, the integration into the technostress process differs. Based on our definition of primary and secondary technostress dimension, we use these insights to summarize and structure applicable technostress prevention measures on the same level of detail in their role of the technostress process. Based on the insights of the different research strands of technostress inhibition and preventive stress management, we develop a conceptual model of preventive technostress management detailing the target points of the three types of technostress prevention. Hereby, we provide five guiding principles of technostress management. These are derived from the Theory of Preventive Stress Management and are new to the technostress literature.
Second, we provide a structured list of 24 actionable technostress prevention measures which consists of a synthesis of inhibitor literature (twelve technostress prevention measures), a literature-based expansion of eight further technostress prevention measures mentioned in related technostress literature, and an empirical expansion of four new technostress prevention measures. Therefore, while 20 of the 24 technostress prevention measures presented in this paper have in some form been mentioned in technostress literature (e.g., provide ICT training or provide supportive ICTs) (e.g. Adam et al. 2017;Pfaffinger et al. 2020), their domain varies between technostress inhibition (e.g., Fuglseth and Sørebø 2014; Tarafdar et al. 2015), mitigation (e.g., Day et al. 2012;Hung et al. 2015), coping (D'Arcy et al. 2014; e.g., Pirkkalainen et al. 2019), or other related constructs (e.g., Benlian 2020; Richter and Richter 2020). Regarding the research strand of technostress inhibition, the inhibitors named by, among others, Tarafdar et al. (2015), Ragu-Nathan et al. (2008), or Fuglseth andSørebø (2014) were used as inputs for, among others, the prevention measures 3 apply human-centered release management (primary prevention), measure 19 provide ICT support (secondary prevention), and measure 23 provide ICT training (secondary prevention). Beyond the named technostress inhibitors, the articles mention further prevention measure candidates, that went into the description of our prevention measures. For example, next to the three above mentioned technostress inhibitors, Fuglseth and Sørebø (2014) state: "However, the study also indicates that managers should actively encourage employees to try out new ICT, and reward employees for using new ICT" and "In addition, managers should emphasize teamwork and encourage ICT knowledge sharing". These insights went, among others, into the prevention measures 13 train managers to successfully lead in the digital working world, measure 15 provide role models with technological changes, and measure 22 offer platforms to exchange experience on ICT use. This example illustrates that our study goes beyond existing literature on technostress inhibitors and related studies. We expand on prior literature by intensively analyzing existing studies and collecting all possible content as prevention measure candidates to then synthesize them by clustering them on the same level of detail and discussing them with experts. The engagement with the experts results in adding further technostress prevention measures beyond literature. See Table 5 and Table 6 in the appendix for a complete overview of all references and their measures that served as an input for our technostress prevention measures.
The first two contributions lead to the following implications for research: Researchers working on analyzing or designing technostress inhibitors can use the prevention framing and approach literature on general stress prevention to obtain further theoretical grounding for their technostress research. Further, they can use our set and structure of technostress prevention measures as a broad overview of measures in the otherwise fragmented literature. Finally, researchers working on stress prevention in other domains besides technostress might consider our set of technostress prevention measures as inspiration for identifying similar prevention measures in other specific stress contexts or in abstracting them from technostress to general stress prevention research.
Moving to the third contribution, we structure and characterize all 24 measures in terms of their basic approach to technostress prevention (i.e., primary and secondary technostress prevention as well as the entity affected: technology, organization, or individual), their applicability, and their relevance in targeting technostressors (for primary measures). Thereby, we contribute a common starting ground for addressing technostress prevention from an organizational view. Due to their importance for future technostress research, we point out two specific facets of the set of characterized technostress prevention measures: The importance of leaders and the lack of a one-fits-all approach. The importance of leaders has been emphasized for preventive stress management (Macik-Frey et al. 2007;Quick 1992). Because of the critical role of leaders, it can be beneficial for organizations to also target primary stress prevention measures to leaders (e.g., executive coaching and peer support) (Hargrove et al. 2011). Arguably, leaders play a similarly important role in technostress prevention as in the prevention of stress in general. Looking at the most relevant primary technostress prevention measures, we find that measure 13, Train managers to successfully lead in the digital working world is the second most relevant (Table 4). Further concerning prior findings on primary stress prevention measures in general, the provision of instrumental (e.g., buddies or mentoring programs), informational (e.g., improvement of the flow of information), and emotional (e.g., increasing emotional understanding among employees) support proofed effective as primary preventive interventions (Quick and Quick 1984a). In the specific context of technostress, we also identified all types of the mentioned support measures with a specific focus on ICTs: Measure 15, provide role models with technological changes, and measure 16, train mentors for digital topics (8 th and 9 th most relevant primary technostress prevention measure), relate to the first category of "instrumental support". Informational support is addressed by the third relevant technostress prevention measure, measure 1, focus on the ICT landscape (Table 4) by reducing the complexity and avoiding redundancies of information. Lastly, emotional support is mainly addressed by measure 6, foster a cooperative culture (6 th most relevant primary technostress prevention measure). While the most relevant primary preventive stress measures can be found in the nine most relevant primary technostress prevention measures, we additionally found purely ICT-specific measures like the most relevant measure 4, Apply human-centered ICT design.
Our efforts in characterizing the different technostress prevention measures in terms of their applicability also help structure the field of preventive technostress management. We offer the possibility to describe measures or groups of measures on a shared set of characteristics. This possibility is important to better compare and classify similar measures in future research. Technostress research can make use of this grouping by not having to assess every measure individually but by being able to assess groups of measures that are similar in their characteristics. We found that primary technostress prevention measures require high initial efforts but yield long-term effects. Therein, primary technostress prevention measures deem a suitable long-term approach for designing workplaces that are technostress free. In contrast, most secondary prevention measures target individuals and require fewer implementation efforts. Thus, they present an opportunity for short-and mid-term prevention of technostress by enabling employees to better react to technostressors in the phase before they have been eliminated permanently. For future research, the implementation and effect duration classification also help design more appropriate studies on testing the actual effectiveness of the different measures. For example, a measure that takes one year to create a positive effect cannot reasonably be assessed in a six-month field study.
Additionally, we provide initial evidence that technostress prevention measures are expected to address different technostressors, and prevention is no one-fits-all mechanism. Hereby, we related the primary technostress prevention measures to nine established technostressors to increase practical applicability and to build the foundation for extensive empirical analyses in future research. Therein, we built on, validated, and extended the few studies that assessed selected prevention measures' potential for specific technostressors (e.g. Galluch et al. 2015;Schmidt et al. 2021;Valta et al. 2021). For example, our results confirm and expand the findings of Valta et al. (2021), who investigated seven measures (e.g., ICT training, contact person) to reduce single technostressors. The authors found that, for example, homogenizing the ICT landscape (here: measure 1, focusing the ICT landscape) reduces the technostressor complexity. This goes along with our results (e.g., measure 1 is highly relevant for complexity, see Table 4). We expanded the results by finding more technostress prevention measures that are expected to be relevant for preventing complexity (e.g., measure 3 apply humancentered release management, measure 4 apply human-centered ICT design, and measure 10 consciously manage ICT-related change). Stemming from the descriptive results of the expert panel, our results serve as the first foundation for relevance. Scholars in the field can build on these results and further theorize on the relationship between individual technostress prevention measures (or sets of such measures with similar characteristics) on specific technostressors. Finally, we contribute by abstracting our theoretical and empirical results to five propositions on preventive technostress management. These are contributions to technostress research, not to the general Theory of Preventive Stress Management. Combining our results with other such contextualizations of specific forms of stress, it might eventually become possible to consolidate the specific findings into a more general context-contingent theory of preventive stress management (Hong et al. 2014). Our five propositions are grounded in theorizing and empirical evidence. We call for future research engaging with these propositions to refute, refine, or validate them. The guiding principles of preventive technostress management and proposition 1 provide a new framing and a new lens to researching technostress at work. Future technostress research should further build on stress prevention theory to advance our knowledge on organizational health, the responsibility of leaders for health, and the constant management and adaptation of a portfolio of technostress prevention measures.

Implications for practice
Organizations have a moral, legal, and economic obligation to address work-related technostress among employees. Moral in the sense of having the responsibility to offer a workplace that is safe and healthy. Legal in the sense that countries like Germany have established laws on an organization's duty to protect employees' physical and psychological well-being. Economic in the sense that technostress can impair an individual's health and work performance and negatively impact an organization's performance (Tarafdar et al. 2015). As adverse outcomes from technostress are not acute events but develop over time, it is important to prevent technostress throughout the process, rather than only reacting after adverse outcomes arise. This indicates a need to support organizations in their prevention efforts by providing actionable fine-grained knowledge on preventive technostress management and measures to implement.
Our research demonstrates the diversity of available prevention measures and, thus, supports organizations in finding a set of measures applicable to their specific setting. Especially for organizations new to technostress prevention, our set of measures offers an information source to understand what different aspects prevention can comprise. Most measures are primary technostress prevention measures. This is noteworthy because secondary prevention is prevalent in job-related stress prevention practice, even though primary prevention is deemed more effective as it tackles stress at its source (LaMontagne et al. 2007). By offering a rich set of primary technostress prevention measures, we support organizations in establishing a comprehensive approach to preventive technostress management.
In terms of characterization, our study shows that most measures are available for organizations of all sizes. However, the measures differ strongly in the technostressors they address, their implementation costs, or their required expert knowledge. Further, most measures require relatively little time until implementation and effect realization. These are encouraging news for practice that might take away some of the burdens when starting with technostress prevention. The characterization of measures also serves as decision support for organizations. When choosing measures, one can easily filter measures by the relevant criteria (e.g., size or entity of change) to be presented with a set of appropriate measures.
The results on the difference in relevance of single technostress prevention measures and specific technostressors have substantial implications for practice. The relevance of technostress prevention measures is expected to be significantly impacted by the technostressor responsible for the technostress. Hence, as the first step to technostress prevention, organizations must identify the critical technostressors within their workforce. Only once an organization knows which technostressors cause technostress to what extent they can effectively select prevention measures. Different measures or a portfolio of prevention measures might be appropriate depending on whether one or multiple technostressors are prevalent. Especially if resources are scarce (typically), it is crucial to prioritize the most severe technostressors concerning prevention.

Limitations
Like any research paper, our study is subject to limitations. First, our results are based on the expertise of a limited number of 13 panelists and, before the Delphi study, a structured literature review and 17 experts in focus groups. We are convinced that our panel is diverse because panelists stem from different organizations of different industries. Still, we can make no formal claim about the panel's representativeness. Even though our Delphi study panel's structure and size fit our research purpose, the measures' relevance (Table 4) should be interpreted only as a rough first estimate. These relevance statements need to be evaluated in more extensive quantitative empirical research in the future to make definitive, generalizable claims. The effectiveness of a technostress prevention measure may highly depend on individual and organizational differences and characteristics (principle 5). For example, demographics, professional experience, or computer self-efficacy potentially affect how strongly people experience technostress and how effective a prevention measure is. To address this, future research should also assess the significance of such confounding variables, similar to existing studies such as Tams et al.'s (2018) work.
Second, we examined the assumed effects of primary technostress prevention measures on nine technostressors. While the selection comprises the most frequently studied technostressors, the literature holds further technostressors (Maier et al. 2012). Also, since the research field is currently rapidly advancing, new technostressors were proposed in the meantime. One example is Fischer et al. (2021), who developed ten technostressor categories, including new constructs like social environment or technical support. Consequently, some newer technostressors found no consideration in the Delphi study. Further technostressors could be included in future works.
Third, we did not narrow the organizational scope to industries or company sizes. We kept the scope broad because this overview of technostress prevention is aimed to be relevant for all types of organizations. However, some of our results might depend on the respective industry, especially concerning primary technostress prevention measures. This should be considered when applying the measures in organizations, and an additional individualization of measures to industries could be required.
Fourth, our study derives work-related technostress prevention measures that enable technological, organizational, and individual changes. On the technological level, the prevention measures are mostly generic in that they generally refer to ICTs. It would be valuable to assess specific ICTs (e.g., e-mail applications or data management software) as a next step. One could also develop additional prevention measures that target any technostress from the respective ICTs or group of ICTs.
Fifth, in this paper, we address individual technostress prevention measures. In real-life scenarios, organizations will apply portfolios of such measures. To increase applicability, it would thus be precious to create a) quantification of the effect of different prevention portfolios (including any positive and negative interdependencies) and b) a handbook/ guideline on how to develop and implement such a portfolio in each setting.
Finally, we derive propositions of preventive technostress management, which have an empirical basis in our study. However, they should be transformed to research hypotheses for empirical testing in future research.

Conclusion
Understanding technostress and its adverse outcomes have emerged as a popular and important research endeavor. Research has extensively investigated the potential effects of technostress on employees' health and organizational performance (Fuglseth and Sørebø 2014; Ragu-Nathan et al. 2008;Srivastava et al. 2015;Tarafdar et al. 2007Tarafdar et al. , 2015. Further valuable studies have also looked into ways of inhibiting the adverse outcomes of technostress (e.g. Ragu-Nathan et al. 2008;Weinert et al. 2020).
Our study complements existing research in that we combine the research strands of technostress inhibition and further related technostress-reducing literature with preventive stress management. Thereby, we focus on addressing the ex-ante prevention of technostress from an organizational perspective. We apply a Delphi study yielding a set of 24 relevant prevention measures that specifically address workrelated technostress. Our study characterizes these measures concerning their basic approach to preventive technostress management, applicability in practice (e.g., organizational size, target group, duration of implementation, realization duration, and effect duration) and their relevance in targeting technostressors. On a more abstract level, we provide guiding principles and proposition for preventive technostress management. This paper's contribution lies in embedding technostress inhibitors in the larger context of preventive technostress management by creating a theoretical basis for technostress prevention, a unification of existing prevention and technostress mitigation or inhibition studies, and a step towards structuring the dynamics underlying prevention measures. For organizations, we offer valuable support to fulfill their moral, legal, and economic responsibility to reduce technostress among their employees.

Appendix
See Figs. 5 and Tables 5, 6, 7, 8, 9 and 10.        A person in charge of psychological risk assessments working for an occupational health management service provider responsible for over one million employees Both Part7 Consultant with a focus in working-time and work organization focusing on SME Both Part8 Psychologist and trainer working for an occupational health management service provider responsible for over one million employees Both Part9 An emergency psychologist working for an occupational health management service provider responsible for over one million employees Second Part10 A researcher with a focus in working-time and work organization at a federal institute focusing on occupational safety and health Second Experienced employer representative working for the employers' association for a major industry in the country

Exp3
Occupational physician in the field of healthiness and well-being for a large workers' compensation company

Exp4
Former vice-chairman of a works council and lecturer at a training institute for works councils focusing among others on remote work and stress management

Exp5
Employee representative working for a national trade union center for a large trade union in the country

Exp6
Scientific director of a federal institute focusing on occupational safety and health management

Exp7
Senior researcher at an institute focusing on remote work and qualification

Exp8
Head of health management at the human resources department for a company in the construction industry

Exp9
Senior safety specialist at the human resources department for a company in the construction industry responsible for corporate health measures, among others stress management and remote work Exp10 Occupational safety specialist at the human resources department for a company in the baking industry, responsible for stress management in the company Exp11 Systemic consultant in and with organizations for a company in the baking industry regarding health and stress management Exp12 Head of occupational safety at the human resources department for a company in the construction industry responsible for corporate health measures, among others during remote work Exp13 26 years experience with psychosocial counseling and psychological coaching for employees and managers in large companies as a social consultant with a focus on, e.g., work-life-balance, stress, remote working   Author contributions All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by MB, HG, and CR. The first draft of the manuscript was written by RS and MB and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding Open Access funding enabled and organized by Projekt DEAL. This research project is funded by the German Federal Ministry of Education and Research (BMBF) as part of the PräDiTec project in the "Healthy -for a lifetime" Program (funding number 02L16D030) implemented by the Project Management Agency Karlsruhe (PTKA). The authors are responsible for the content of this publication.

Availability of data and material
The data that support the findings of this study are available from the corresponding authors upon reasonable request.
Code availability Code for data cleaning and analysis are available from the corresponding authors upon reasonable request.

Conflicts of interest/Competing interests Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.