Digital and Digitized Interventions for Teachers’ Professional Well-Being: A Systematic Review of Work Engagement and Burnout Using the Job Demands–Resources Theory

Abstract

Teachers’ work engagement and burnout are crucial for both teachers and students. Traditional interventions have reduced burnout and increased engagement. However, with the rise in digital interventions and their advantages in scalability, cost-effectiveness, higher reach, and fidelity, we aimed to explore these aspects further. Thus, our research questions were: What are the core components of teachers’ digital and digitized interventions for work engagement and burnout? How are these interventions implemented, focusing on the mode of delivery, support systems, fidelity, and dosage? We examined the core components and implementation factors, including the mode of delivery, support systems, fidelity, and dosage of digital interventions for teachers’ work engagement and burnout. A systematic review of the literature was conducted using Cochrane guidelines and PRISMA reporting. Of 1761 studies, 15 were eligible, and six were included, but none examined work engagement as an outcome variable. Moreover, core components included mindfulness, stress management, emotional intelligence, social–emotional competencies, organizational skills, and technological competence, all targeting personal resource development. These findings align with Domitrovich et al.’s framework and JD-R theory, emphasizing support systems and personal resources. Using the Job-demands resources theory and Domitrovich et al.’s framework, we found that all six studies focused on burnout using the Maslach Burnout Inventory. Four reported significantly decreased emotional exhaustion, two increased personal accomplishment, and one decreased depersonalization. Two digital interventions showed no significant changes in the burnout dimension. The small sample size limits conclusions, but the results indicate a fundamental difference between digital and digitized interventions regarding design, and the necessary support systems. Our findings indicate that core components must target teachers’ personal and job resources and job demands. Moreover, fidelity, dosage, and duration measures are crucial, along with facilitated group discussions and technical assistance for successful outcomes on burnout. These findings are relevant for practitioners, researchers, and policymakers, extending beyond education. Future research directions and implications are discussed.

1. Introduction

Teachers’ professional well-being is essential for the teachers, students, and children’s learning and thriving (Jennings & Greenberg, 2009; Maricuțoiu et al., 2023). Although many teachers report that they find their profession highly meaningful (e.g., Martela & Pessi, 2018), their professional well-being steadily declines (McCallum, 2020). Teacher stress is an international challenge (Agyapong et al., 2022). The decline in teachers’ professional well-being has generated growing interest from researchers and policymakers regarding how to best improve and support it (Schleiker, 2018).

Well-being includes subjective well-being (e.g., Diener, 1984), psychological well-being (Ryff, 1989), and how we “feel good and function effectively” across various life domains (Huppert & So, 2013). Professional well-being is a multifaceted construct that reflects the quality of an individual’s experience in their professional role, encompassing the emotional, psychological, social, and sometimes physical dimensions of functioning at work (e.g., Bautista et al., 2023). Professional well-being involves experiencing positive perceptions and constructive conditions at work and beyond, enabling workers to thrive and reach their full potential (e.g., Chari et al., 2018). Teachers’ professional well-being is a multidimensional construct that can be conceptualized and operationalized differently (Collie, 2025). For instance, a study by Yıldırım (2014) identified teachers’ professional well-being as consisting of factors such as cooperation among staff, fair and helpful assessment and feedback, a positive school climate, student-oriented teaching practices, classroom climate, and professional development. Hascher and Waber (2021) state that a single definition is lacking within the field. Thus, we adopt the definition of teachers’ professional well-being as teachers’ positive evaluation of and healthy functioning in their work environment (Collie et al., 2015). Specifically, we focus on the dimensions of burnout and work engagement as aspects of teachers’ professional well-being.

Teachers’ work engagement and burnout are essential antecedents for students’ learning and thriving (Maricuțoiu et al., 2023). Therefore, burnout and work engagement are critical factors in professional well-being. Many teachers experience an imbalance between the demands of their job and the personal resources available to them (McCarthy, 2019). Due to this imbalance, teaching is an emotionally demanding profession with high levels of burnout (e.g., Madigan & Kim, 2021). Burnout can be defined as “a prolonged response to chronic emotional and interpersonal stressors on the job” (Maslach et al., 2001, p. 397) and consists of three main factors: emotional exhaustion, depersonalization/cynicism, and personal accomplishment/inefficacy (Maslach et al., 2001; Maslach & Leiter, 2016). Burnout is often contrasted with work engagement (Schaufeli & Bakker, 2004; Gonzalez-Roma et al., 2005). For instance, teachers with high work engagement participate in more social interactions with students than less engaged teachers (Klassen et al., 2013). Moreover, in a study by Hamre and Pianta (2005), teachers with high work engagement showed greater emotional and instructional support, and pupils with high-risk problems performed just as well as those classified as low-risk peers. In contrast, the same study found that teachers with low work engagement exhibited lower emotional support and pupils with difficulties performed significantly worse.

Unfortunately, teachers experiencing burnout “may develop a callous, cynical attitude towards students, parents, and colleagues (depersonalization) and eventually grow to feel they are ineffective teachers (lack of personal accomplishment)” (Jennings & Greenberg, 2009, p. 498). The potential consequences of teacher burnout, as described by Jennings and Greenberg (2009), can extend beyond individual teachers to affect students, schools, and society. Moreover, in a recent study of 300 Norwegian teachers and 6014 students, teacher burnout was negatively associated with student-teacher-reported classroom climate (Jensen & Solheim, 2020). Additionally, socially and emotionally competent teachers appear to foster students’ social–emotional competence, well-being, and academic achievement (Carvalho et al., 2021). Owing to the potential consequences of teacher burnout and work engagement, developing effective interventions to prevent burnout and increase teacher engagement is valuable. Over the last two decades, several interventions have been developed, revealing that core components such as meditation, mindfulness practices, and stress reduction techniques contribute to positive changes in teacher well-being (e.g., see the meta-analysis by Iancu et al., 2018; review study by Hagermoser Sanetti et al., 2021). Core components are the “active ingredient” in interventions, driving change in participants (Blasé & Fixsen, 2013; Fixsen & Blasé, 2020).

Digital and digitized interventions are gaining popularity due to their potential for scalability, replicability, user friendliness, higher fidelity, and lower cost. We define digitized interventions as interventions delivered using technology, while digital interventions are interventions specifically designed to present the content in a manner only possible using technology. With the increased use of technology, there is a growing trend to deliver health-focused interventions digitally (Philippe et al., 2022). Thus, this systematic review explores the core elements of digitized and digital interventions, their mode of delivery, implementation characteristics, and their impact on reducing burnout and increasing teacher work engagement.

Digital and Digitized Interventions Aimed at Teachers’ Burnout and Work Engagement

With the advent of technology and the need to explore new ways of collaborating during the COVID-19 pandemic, the demand for information and communication technology (ICT) has increased. Consequently, an increasing number of digital interventions have been introduced. Digitally delivered interventions are defined as those that use technology to facilitate or support behavioral change (West & Michie, 2016). Digital and digitized interventions offer several advantages over traditional ones. For instance, they are potentially more easily scalable, as they do not depend on a trainer or psychologist to deliver the intervention; they could also be more cost-effective and have a higher reach (Moshe et al., 2021; Newby et al., 2021). Additionally, these interventions can be delivered asynchronously, allowing them to meet participants’ personal time preferences. Participation is preferable during working hours to avoid increasing teachers’ heavy workloads (Hagermoser Sanetti et al., 2021). However, digital and digitized interventions also have several disadvantages. A lack of skills and abilities, along with prolonged exposure to information technology, can result in technological or technostress (Salanova et al., 2013). Furthermore, technostress at work can be defined as a “negative psychological state associated with the future use or threat of ICT. This experience is related to feelings of anxiety, mental fatigue, skepticism, and inefficacy” (Salanova et al., 2013, p. 423). Several types of technostress and dimensions, such as techno-complexity and techno-overload, are likely to contribute to burnout (Kaltenegger et al., 2024) Consequently, digital training programs can be perceived as a demand rather than a means of preventing burnout. Moreover, some findings indicate that older age and/or having a neurotic personality may increase the likelihood of technostress (Pagán-Garbín et al., 2024; Srivastava et al., 2015).

It may seem like an oxymoron to deliver an intervention to reduce burnout and increase work engagement, which could inherently generate more stress due to the mode of delivery via technology. Given this risk, optimizing digitized interventions to counteract potential downfalls is essential, particularly for teachers. Time and energy are scarce resources, and as noted previously, burnout can have potentially dire consequences for the teachers and their students. However, no previous reviews have explicitly focused on digital and digitized interventions, burnout, or work engagement that we could locate. Considering teachers’ limited time and energy, creating a return on investment for their time is paramount. Our research questions are: What are the core components of teachers’ digital and digitized interventions for work engagement and burnout? How are these interventions implemented, focusing on the mode of delivery, support systems, fidelity, and dosage? Hence, we consider this systematic review a study of digital and digitized interventions, their core components, corresponding support systems, and their implementation to be a valuable contribution to the research field, practitioners, and policymakers.

2. Theoretical Background

2.1. Job Demands–Resources (JD-R) Theory, Burnout, and Work Engagement

We selected the work-oriented Job Demands–Resources (JD-R) theory (Demerouti et al., 2001) as the theoretical framework for this study because it effectively integrates a positive focus on work engagement with a negative focus on burnout. This framework has a broad scope and is flexible (Schaufeli, 2017). In line with JD-R theory, every occupation is characterized by specific job demands and resources, “Job demands refer to the physical, social, or organizational aspects of a job that require sustained physical or mental effort and are therefore associated with physiological and psychological costs” (Demerouti et al., 2001, p. 501). In contrast, job resources refer to a job’s physical, psychological, social, or organizational aspects that may be functional in achieving work goals, regulating the impact of job demands, and stimulating personal growth and development (Bakker & Demerouti, 2017). These job demands and resources impact employees’ health and well-being, playing a role in two energetic processes: the health impairment process and the motivational process. The health impairment process focuses on how job demands exhaust employees and cause chronic stress, which can lead to burnout and psychological or physical illnesses. In contrast, the opposite includes energy, involvement, and efficacy. Burnout may have adverse physical and psychological effects (Salvagioni et al., 2017). This implies that individuals experiencing burnout may feel emotionally exhausted, develop cynical thoughts about one’s job, and take longer than expected to complete tasks (e.g., Jennings & Greenberg, 2009). As a result, burnout may result in anxiety, depression, lifestyle illnesses such as type 2 diabetes, and cardiovascular disease (see meta-study Salvagioni et al., 2017; Shirom et al., 2005).

Conversely, the motivational process focuses on how job resources motivate and inspire employees to learn, grow, and develop to meet work goals, resulting in work engagement and high performance (Bakker & Demerouti, 2007). According to Bakker et al. (2008), work engagement consists of three factors—vigor, dedication, and absorption—as measured using the Utrecht Work Engagement Scale (UWES) (Schaufeli & Bakker, 2003; Schaufeli et al., 2006). This approach suggests that work engagement is an independent and distinct concept rather than the opposite of burnout, although it is negatively related to burnout. The UWES was developed to assess the positive dimension of work engagement (Schaufeli & Bakker, 2003). Xanthopoulou et al. (2007) added personal resources (e.g., resilience and optimism) to the theory. Other established personal resources within JD-R theory include mindfulness (Grover et al., 2016), self-regulation (Bakker & de Vries, 2020), and emotional intelligence (Mérida-López et al., 2022).

The JD-R theory has been successfully applied in prior research focusing on teachers’ well-being, making it particularly relevant to our study. The JD-R model is regarded as a robust process model due to its flexibility and practical utility, as it integrates contextual and individual factors (Granziera et al., 2020). Confirming the JD-R model, Hakanen et al. (2006) conducted a study with 2038 Finnish teachers, revealing that high job demands lead to burnout, and that abundant job resources enhance work engagement. Additionally, Van Wingerden et al. (2017) explored how a job crafting intervention improved personal resources such as self-efficacy, optimism, and resilience, which led to increased work engagement and performance. These personal resources play a crucial role in managing job demands and sustaining motivation. Finally, Han et al. (2020) highlighted that emotion regulation strategies can be specifically targeted in JD-R-based interventions. Despite its widespread application, the JD-R theory has limitations, including questions about the generalizability of resources and demands across diverse cultures (Rattrie & Kittler, 2014). The lack of clarity surrounding the definitions of demands and resources and the role of personal resources may lead to unclear classifications regarding whether these are considered resources or outcomes (Schaufeli & Taris, 2014). Additionally, different personality profiles may perceive challenge and hindrance demands differently, particularly concerning technology (Kunzelmann et al., 2025). Regardless of these limitations, we find that the strengths of the JD-R theory outweigh its limitations. Hence, we applied the JD-R theory in the current study to analyze digital and digitized studies, focusing on burnout and work engagement regarding the interventions’ aim of increasing job and personal resources as opposed to decreasing job and personal demands.

2.2. Implementation Factors in Efficient Interventions

According to Fullan (2013, p. 111), implementation is a process that consists of “putting into practice an idea, program or set of activities new to the people attempting or expected to change (…)”. To analyze an intervention’s implementation, we have chosen the conceptual framework of Domitrovich et al. (2008), who advocate that focusing on implementation quality can increase the outcome for participants and provide valuable knowledge for the researchers by distilling the crucial components both in an intervention and during an intervention’s implementation process. Implementation quality refers to how well different intervention components are delivered (Humphrey et al., 2016, p. 6). The multilevel conceptual framework depicts several factors that may affect interventions and the quality of their implementation. These factors are divided into three levels: macro level (e.g., policies and funding, university/community), school level (e.g., school climate and organizational health, resources, school characteristics), and individual level (e.g., professional characteristics and psychological factors). The model focuses on an adequate support system for intervention, standardization of delivery, core elements of content, and intervention delivery (Domitrovich et al., 2008). An adequate support system is crucial for communication, inspiration, and motivation (Domitrovich et al., 2008).

According to Humphrey et al. (2016), measuring the quality of the intervention implementation, the support system, its dimensions, and the factors affecting implementation can be seen as a multidimensional construct consisting of eight different dimensions (fidelity, dosage, adaptation, quality, reach, responsiveness of the participants, program differentiation, and monitoring of control/comparison groups). These constructs can be defined as follows: Fidelity measures the degree to which an intervention and its support system are conducted as planned, while dosage refers to the duration and frequency of an intervention’s implementation. Adaptation is the nature and extent of changes made to an intervention (Humphrey et al., 2016). Rarely can an intervention be implemented exactly as planned, and sometimes adaptations may even be advantageous; however, they can also lead to quality deterioration. To assess the level of adaptation, the intervention must be standardized and communicated clearly and in detail (Humphrey et al., 2016). It is important to differentiate between implementation adaptation and content adaptation, as implementation adaptation may positively activate core content (Kirk, 2020). Participants’ responsiveness refers to the degree to which they engage in an intervention (Humphrey et al., 2016). Reach is the rate and scope of participation in an intervention. Program differentiation is the extent to which interventional activities can be distinguished from other practices. Finally, control/comparison groups are monitored to measure differences in post-intervention outcome variables. Implementation quality refers to how well different components of an intervention are delivered (Humphrey et al., 2016, p. 6). For a successful intervention, it is essential to ensure social validity. Social validity originates from Wolf (1978), who proposed a framework consisting of three dimensions for assessing interventions: the social significance of goals, procedures, and effects (Leko, 2014). Social validity anchors the vision, purpose, and acceptance of an intervention, thereby improving the quality of the implementation process (Foster & Mash, 1999). The sum of both core components and implementation outcomes in an intervention is the element that drives change (Carroll, 2020). Thus, interpreting an intervention’s outcomes depends on understanding which aspects are delivered and how well they are conducted (Durlak & DuPre, 2008, p. 328). However, only a few studies have used adequate resources to monitor all the above-mentioned dimensions (Domitrovich et al., 2008).

2.3. Core Components Focusing on Job Demands, Personal and Job Resources

As mentioned earlier, implementing the core components and their content is crucial for successful interventions. The content of the core components can be analyzed using the framework of JD-R theory (Bakker et al., 2023) to determine whether they aim to increase job and personal resources and/or decrease job or personal demands. According to Fixsen and Blasé (2020), the core components create a desirable change, provided that the intervention maintains sufficient implementation fidelity and is implemented with high quality.

Most evaluated interventions for the well-being of teachers are targeted at the individual and not at the organizational-wide level (Naghieh et al., 2015). A review by Avola et al. (2025) found that the content of the interventions to teacher well-being and burnout since 1996 primarily focused on individual well-being and that only some interventions incorporated “communal activities.” Similar findings are supported in a review study by Hagermoser Sanetti et al. (2021), who found that stress reduction interventions focused on meditation and mindfulness and not increasing job resources or decreasing job demands. However, one systematic review by Naghieh et al. (2015) found organizational-wide interventions for teachers, focusing on increasing job resources and decreasing job demands (changing tasks alongside stress management, organizational characteristics, and performance bonuses and mentoring options). They found low-quality evidence of improvements in teachers’ well-being. In contrast, Van Wingerden et al. (2017) conducted an intervention using job crafting and individual relaxation techniques. Job crafting is when employees change their own jobs by increasing social/structural resources, increasing challenging job demands, or decreasing hindrance demands (Tims et al., 2012). The organizational-wide intervention resulted in increased work engagement and performance among teachers. According to Schaufeli et al. (2023), individual-focused interventions generally affect burnout complaints better than organizational-wide interventions. However, due to a paucity of organizational-wide interventions, care must be taken when making conclusions. Moreover, they strongly propose more interventions focused on teams- and organizations. Not only for mild burnout complaints but also for burnout disorder (Schaufeli et al., 2023).

As mentioned, many interventions are focused on the individual teachers. For instance, Zarate et al. (2019) conducted a meta-analysis on mindfulness training on teacher well-being. There were 18 mindfulness-based interventions (MBIs) included, published between 1999–2017. They found that interventions varied in dosage, frequency, and delivery. However, all studies included mindfulness practices such as meditation, breathwork, and visualization. The effect sizes were small for depression and burnout but moderate for stress. Similar findings were reported by Hagermoser Sanetti et al. (2021), who examined 18 articles published between 1987 and 2016. Their findings indicated that the most evaluated stress-reduction intervention for teachers incorporated meditation. These interventions positively impacted mindfulness, mental health symptoms, personal accomplishments, and emotional exhaustion. Agyapong et al. (2023) analyzed 40 articles published between 1974 and 2022. They concluded that mindfulness-based interventions, alone or with yoga or cognitive-behavioral therapy, were the most effective for reducing teacher burnout. These interventions were most effective when tailored to individual needs and combined with organizational support. In conclusion, increasing job resources with personal resources appears to be the most efficient way to achieve this goal. These findings are consistent with those of Iancu et al. (2018), who conducted a meta-analysis of 23 controlled studies to examine the effectiveness of interventions aimed at teacher burnout. They found that the overall effects were small but statistically significant. Mindfulness interventions had a significant impact on both exhaustion and personal accomplishment. The cognitive-behavioral approach and mindfulness/meditation techniques significantly alleviated emotional exhaustion. Von der Embse et al. (2019) analyzed 24 articles from 1998 to 2017. They found that the interventions that led to desirable changes in teachers’ well-being were within the mindfulness and cognitive-behavioral domains. Based on these studies, strong evidence suggests that mindfulness-based and cognitive behavioral interventions effectively reduce teacher burnout and stress. Despite the plethora of reviews of interventions for teachers focusing on burnout and work engagement, as well as the increasing use of technology, reviews focusing on the digital mode of delivery using the JD-R theoretical framework appear scarce.

2.4. What Is the Optimal Duration and Dosage?

In addition to the core components, the intervention’s implementation, duration, and dosage are paramount. According to a systematic review by Iancu et al. (2018), interventions lasting less than one month had minimal effects on burnout measures. In contrast, Klingbeil and Renshaw (2018, p. 507) found that the dosage of mindfulness-based interventions had diminishing returns on burnout beyond 24 h. Another study reported a medium effect in reducing emotional exhaustion with 20 h across 10 weeks (Cooley & Yovanoff, 1996). Finally, a scoping review by Agyapong et al. (2023) focused on interventions to reduce stress and burnout among teachers in articles published between 2018 and 2022. Their main finding was that four days of intensive mindfulness training, instead of the regular eight-week program, could be effective in enhancing teacher engagement. Lang et al. (2020) found that some teachers reported higher post-intervention stress levels due to their increased level of awareness. Their sample consisted of preschool teachers who received a three-hour digital intervention on stress-reduction techniques. However, the results generally demonstrated that the intervention was beneficial, even at a low dosage. Conversely, Hidajat et al. (2023) did not find a clear conclusion regarding the relationship between dosage and outcomes in their review. Based on the above research, the optimal duration and dosage of interventions to reduce burnout and increase teacher work engagement remain inconclusive.

2.5. The Present Study

This systematic review examines digital and digitized interventions to reduce teacher burnout and increase their work engagement. As the application of technology in interventions is an emerging trend, this study seeks to provide insights into the future development of such interventions. We analyzed the selected studies’ core components and implementation factors, including the mode of delivery, support systems, fidelity, and dosage. Accordingly, our research questions were as follows:

RQ1. 

What are the core components of teachers’ digital and digitized interventions for work engagement and burnout?

RQ2. 

How are these interventions implemented, focusing on the mode of delivery, support systems, fidelity, and dosage?

To analyze and conceptualize the key elements of a successful digital and digitized intervention for teachers addressing burnout and work engagement, it is essential to identify the core components embedded in these interventions. We aimed to analyze whether the core components included in this scoping review were designed to enhance personal or job resources and/or a decrease in personal or job demands. According to the JD-R theory, resources are associated with a motivational process resulting in work engagement, whereas demands are linked to a health impairment process that may lead to burnout over time (Bakker et al., 2023). It is essential to scrutinize how these interventions are implemented. We consider several factors that may impact an intervention’s implementation quality, such as delivery, core elements, and support systems (Domitrovich et al., 2008), as well as delivery factors, such as fidelity, dosage, and duration (Humphrey et al., 2016).

3. Methods

3.1. Study Design, Data Sources and Search Strategy

This systematic review followed the PRISMA guidelines (Moher et al., 2009). We identified and examined pertinent studies using the JD-R theory and implementation quality. The results were synthesized within this framework, and we evaluated the quality of the studies, adhering to the approach set forth by Gough et al. (2017). All the included data were publicly available; thus, no ethical approval was obtained. A systematic review methodology was selected to gain an overview of the available digital interventions focusing on teachers’ burnout and work engagement. We applied a multistep approach to the search process. PICO elements (Population, Interventions, Comparators, Outcomes) were used to develop a search strategy outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins & Green, 2008). First, preliminary searches were conducted in Oria, the University of Stavanger’s database, and Google Scholar to identify relevant studies and pilot various search strings. Second, the search string shown in Appendix A was applied to the following databases: ERIC, Psych Info, Scopus, Web of Science, and Academic Search Premier. We also reviewed meta-analysis studies. The screening of articles based on titles and abstracts, extraction, and data synthesis in this systematic review was conducted using EPPI Reviewer (a software application for reviews) and Microsoft Excel. The studies were assessed based on pre-established inclusion and exclusion criteria. The search string was built on keywords used in previous reviews and relevant articles, with additional keywords added to focus on this systematic review. A search specialist from the library and a specialist from the Knowledge Centre for Education at the University of Stavanger were consulted before the search string was completed. Appendix A provides the search string for each database.

The search string was centered around keywords combined with Boolean Operator AND:

• All Fields (teacher* or instructor* or lecturer*)
• All fields (intervention* or rct or “control trial*” or course or program)
• Title (Work NEAR/5 Engagement) OR (Job NEAR/5 Engagement) OR (Burnout)

“Digital” was not included in the initial search but was used in subsequent screening of the title and abstract to ensure an adequately encompassing search. The database search was limited to peer-reviewed original articles published in English with no date restrictions.

3.2. Study Selection and Quality Appraisal

Following the search conducted from 20 to 22 February 2023, all identified articles were collated and uploaded to EPPI, and duplicates were removed. All uploaded articles were screened on titles and abstracts, until only 101 articles remained. Then, 98 full-text articles were screened and reported according to PRISMA guidelines (Page et al., 2021) to determine eligibility based on predetermined inclusion and exclusion criteria (three articles were unavailable in full text; see

Table 1. Inclusion and exclusion criteria.

Inclusion Criteria	Exclusion Criteria
The type of participants is in-service teachers teaching in grades 1–13 (including Kindergarten in the US and in Australia)	Type of Participants: All occupational groups other than teachers working in-service.
Context: Digital and digitalized interventions implemented for in-service teachers.	Types of Interventions: Professional well-being interventions that do not include digital technology in the delivery method.
Types of Interventions: Both digitalized and digital interventions aim to increase teachers’ professional well-being. These include technology, online apps, PDFs, CDs, and game-based interventions.	Types of Studies: Qualitative studies and conceptual/theoretical papers will be excluded from the review.
Types of Studies: Quasi-experimental and experimental studies will be included in the review.	Comparator/control: Studies with no control or pre-and post-test.
Comparator/control: The studies included will include a randomized control group, an assigned control group, or only pre-and post-test measures.	Type of Publication: Grey literature, dissertations, not peer-reviewed articles, reports, or articles in languages other than English.
All years of publications.	Outcome Measures: Outcome measures outside teachers’ job engagement and/or burnout.
Type of Publication: Full-length empirical peer-reviewed articles published in peer-reviewed journals, with no date restrictions.
Language: English
Outcome Measures: Interventions to promote teachers’ professional well-being with job engagement and/or burnout outcome variables.

). Any disagreements between the authors were discussed until resolved, resulting in an interrater agreement of 100%. The initial interrater reliability was 80%. We convened to reach a consensus on the articles we initially disagreed on, explaining our reasons for including or excluding them based on the predetermined inclusion and exclusion criteria. Involving a third reviewer was considered if we did not reach consensus. The less experienced reviewer expressed concerns about prematurely excluding articles, which led to the main inter-rater discrepancy.

Given that the study encompasses both Randomized Controlled Trials (RCTs) and Quasi-Experimental (QE) research designs, the Mixed Methods Appraisal Tool (MMAT) (Hong et al., 2018) was considered the most suitable appraisal instrument. The purpose of the quality assessment was to evaluate and provide insights into the overall methodological quality of the included studies. The MMAT assessed the methodological quality of six studies, evaluating each on five items reported as “yes,” “no,” or “can’t tell.” These items focus on sampling, the risk of non-response bias, data collection methods, measurement techniques, and alignment with research questions. Studies were appraised based on their adherence to MMAT criteria. The authors applied the MMAT independently before meeting to discuss and reach a consensus. An overall score for each study is not reported according to the recommendation by Hong et al. (2018, p. 1).

3.3. Data Extraction

Data were extracted from the six included studies (Ansley et al., 2021; Oliveira et al., 2022; Pozo-Rico et al., 2020; Round et al., 2022; Matiz et al., 2020; Zadok-Gurman et al., 2021) in the systematic review by two independent reviewers using Excel. Coding was first applied to three articles to pilot-test the instrument, and several codes were added. This was an abductive iterative process anchored to the research questions and the principles of critical realism (Fletcher, 2016). An exploratory approach, within the theoretical framework of the JD-R theory (Bakker et al., 2023), was used to analyze the core components with a focus on job and personal resources as well as job and personal demands. Furthermore, we applied Domitrovich et al.’s (2008) implementation quality conceptual framework to synthesize the intervention modes of delivery, support systems, fidelity, dosage, and duration. Narrative data synthesis was performed to analyze the following: country, study design, study size, sex, underlying theory, intervention length, self-reporting scales, digital modalities, and primary outcomes (

Table 2. Overview of included studies.

Author, Country, Publication Year	Study Details	Intervention Details	Underlying Theory and Definition of Burnout	Main Effects and Burnout Scale Used Relevant to This Systematic Review (Other Effects were Found but Not Reported Here):
(Ansley et al., 2021, USA)	Type: randomized controlled trial (RCT) Participants: 51 Women (W): 10 Men (M): 41 Measurement: Pre- and Posttest Digital modalities: Video files/audio files, written instructions, online open learning platform Other Materials: Course guide, workbook, course schedule, and suggested pacing.	Duration in weeks: 4 Dosage of intervention: 30 min per module Intervention Group (IG): 2 modules weekly (total hours required: 4; participants reported 8 h a week to practice coping strategies, although they were not mandatory) Control Group (CG): No treatment was received. Coach/Facilitator: No Synchronous hrs.: No Asynchronous hrs.: Yes Goal/focus: Teaching teachers coping resources to manage stress (Lazarus, 1984) and ameliorate social–emotional competencies that promote positive learning experiences, increase teacher efficacy, and lower burnout rates through mindfulness, relaxation, cognitive restructuring, social support, and physical exercise. Intervention feasibility: Treatment acceptability (7 items scale): Time engaged in independent practicing coping strategies weekly (in between work) Support: After 48 h, an email is sent out, followed by a reminder again after 48 h. If no response is received, the participant is withdrawn. Enrolled received a welcome email and, course pacing guide, and optional workbook. Weekly emails with updates about progress, and reminder emails if behind. Five weeks to complete the program.	Underlying theory: Based on previous stress reduction interventions, including mindfulness, relaxation, and cognitive restructuring for teachers. Burnout definition: Maslach’s three dimensions: Emotional exhaustion (EE), personal accomplishment (PA), depersonalization (DP) (Maslach & Jackson, 1981)	Effect: No significant effects. Burnout scale: Masclach burnout inventory-educator survey (MBI-ES; Maslach et al., 1986)
(Matiz et al., 2020, Italy)	Type: Quasi-experimental Participants: W 58 Measurement: Pre- and Posttest Digital modalities: Audio recording, email, calls, videos sent via the internet Other Materials: Audio files, books, and articles per request from the participants.	Duration in weeks: 8 Dosage of intervention: IG and CG: 8 group meetings (2 h per meeting), two face-to-face meetings, and 6 video lessons delivered via the Internet. Daily meditations are 30 min, and activity is reported every two weeks. Based on baseline scores on resilience, participants were divided into low (LR) and high-resilience (HR) groups. Coach/Facilitator: Facilitator Synchronous hrs.: Yes Asynchronous hrs.: Yes Goal/focus: Test the MOM on anxiety, depression, affective empathy, emotional exhaustion, psychological well-being, interoceptive awareness, character traits, and mindfulness. Support: Phone calls, discussions, books, and articles to read per request.	Underlying theory: Mindfulness-Oriented Meditation (MOM) Burnout definition: Maslach’s three dimensions. (Maslach & Jackson, 1981)	Effect: The LR group significantly lowered EE. The HR group experienced higher PA. DP no change. ηρ2 EE = 0.240 significant change over time, and change overall decrease Burnout scale: Italian translation of the Maslach Burnout Inventory Educators Survey (MBI-ES) (Maslach et al., 1986).
(Oliveira et al., 2022, Portugal)	Type: Quasi-experimental Participants: 81 W: 78 M: 3 Measurement: Pre- and post-test. 3 Months- and 6 months after posttest. Digital modalities: Zoom/Teams, Moodle platform. Other Materials: None reported.	Duration in weeks: 10 Dosage of intervention: Weekly 2.5-h group sessions and 2.5 h asynchronous hours for ten weeks. (Combined 50 h) IG: Three clusters of teachers according to perceived organizational climate CG: No treatment Coach/Facilitator: Trained and certified instructor Synchronous hrs.: Yes Asynchronous hrs.: Yes Goal/focus: Increasing social and emotional competence and occupational health., e.g., increased self-regulation, positive relationship, conflict management skills and increased emotional well-being, decreased occupational stress, and emotional exhaustion symptoms. Support: No specific support was reported in the article.	Underlying theory: SEL for teachers based on three main theoretical frameworks: Emotional intelligence theory (Salovey & Mayer, 1990), the Transactional model of stress and coping (Lazarus, 1984), and Self-determination theory (Deci & Ryan, 1985). Burnout definition: Emotional exhaustion is emphasized (Maslach et al., 1996).	Effect: Significant effect on EE. DP and PA are not reported. Beta = −0.84, SE = 0.40, 95% Burnout scale: Maslach burnout inventory-Educators survey (Maslach et al., 1996; Portuguese version).
(Pozo-Rico et al., 2020, Spain)	Type: RCT Participants: 141 W: 80 M:61 Measurement: Pre- and Posttest Digital modalities: Moodle platform for pre and posttest. E-learning Moodle platform for discussions, online teaching. Other Materials: Not reported.	Duration in weeks: 14 Dosage of intervention: IG: Not reported CG: No treatment. Coach/Facilitator: Trainer Synchronous hrs.: Yes Asynchronous hrs.: No Goal/focus: Coping with stress, preventing burnout, improving their information and communications technology, and introducing the principles of EI in the classroom. Support: Not specified in the article.	Underlying theory: Introducing emotional intelligence to reduce stress and prevent burnout in teachers for their own well-being and to introduce it into the classrooms. Burnout definition: Maslach three dimensions (Maslach & Jackson, 1981)	Effect: Significant decrease in EE and DP, and increase in PA. ηρ2 EE = 0.63, DP = 0.76, PA = 0.46 Burnout scale: Maslach Burnout Inventory (MBI; 22-item), (Maslach et al., 1996).
(Round et al., 2022, UK)	Type: RCT Participants: 66 W: 54 (35 teachers) M: 12 The remaining participants were fulltime workers of other occupations. Measurement: Pre- and Posttest, and anxiety test before and after each of the three days of writing. Digital modalities: Not specified, only via the internet or on an online platform Other Materials: None specified.	Duration in weeks: 0.4 Dosage of intervention: IG: 20 min positive expressive writing in three days (combined 60 min) CG: 20 min neutral writing in three days Coach/Facilitator: No Synchronous hrs.: No Asynchronous hrs.: Yes Goal/focus: To test positive expressive writing on burnout, job satisfaction, anxiety, perceived stress, and self-reported physical symptoms. To measure baseline differences in burnout and perceived stress between teachers and non-teachers. Support: Emails were sent out to remind the participants of participation in the study.	Underlying theory: Written emotional disclosure (Pennebaker, 1997). Positive writing for reducing stress and anxiety and increasing well-being. (Allen et al., 2020). Burnout definition: Maslach three dimensions (Maslach & Jackson, 1981)	Effect: No significant or interactive effects on burnout for the teachers, nor any other difference for the participants of other occupations. Burnout scale: Maslach burnout inventory MBI, three dimensions (Maslach & Jackson, 1981)
(Zadok-Gurman et al., 2021, Israel)	Type: Quasi-experimental Participants: 60 W: 58 M: 9 Measurement: Pre- and Posttest Digital modalities: Not specified, only via the internet or an online platform. Other Materials: No other materials reported. Control group received an IBSR book upon posttest completion.	Duration in weeks: 20 Dosage of intervention: IG:10 biweekly meetings 2.5 h/meeting and biweekly individual sessions with a facilitator 1h/session for 20 weeks (combined 35 hrs). CG: Participated in other courses unrelated to the intervention. Coach/Facilitator: Facilitator Synchronous hrs.: Yes Asynchronous hrs.: Yes Goal/focus: Increase teachers’ well-being and assess the intervention’s effect on resilience, burnout, mindfulness, and stress among teachers during the COVID-19 pandemic. Support: No specific support reported.	Underlying theory: Blended inquiry-based stress reduction (IBSR), mindfulness and cognitive reframing intervention on teachers’ well-being. Burnout definition: Maslach three dimensions (Maslach & Jackson, 1981)	Effect: Significant decrease on EE. PA had no change, and DP was not measured. EE: Cohens d = 0.752 Burnout scale: Maslach Burnout Inventory (MBI), Maslach et al. (1996).

). The two authors convened a post-completion data synthesis to examine and finalize the Excel documents. Any disagreements were deliberated upon until they were resolved. Two articles were added through mutual reference mining (Matiz et al., 2020; Pozo-Rico et al., 2020), while were removed due to the lack of burnout or work engagement scales for outcome measures and insufficient digital delivery of the intervention.

4. Results

As presented in the Methods section, we identified 1761 studies upon completing our database search. Figure 1 illustrates the PRISMA flowchart of the article selection process. A total of 358 duplicate studies were excluded. Furthermore, 1403 articles were screened based on their title and abstract, resulting in the exclusion of 1302 articles. Hence, 101 articles were retrieved, though three could not be obtained. Thus, 98 articles were screened, and 83 were excluded for failing to meet the inclusion criteria. The process yielded 15 eligible articles, of which 11 were excluded due to missing relevant delivery modes or outcome variables, and two articles were added through reference mining. Finally, after applying the inclusion and exclusion criteria, six studies were included in the current review. However, none of these studies examined work engagement as an outcome variable. We did not limit our search to any publication year. However, all included studies were published between 2020 and 2022 (the search was completed in February 2023). First, we present an overview of the study characteristics. Next, we present how the interventions were implemented, followed by our analysis of the core components and their impact on burnout dimensions.

4.1. Study Characteristics and Quality Assessment

Design, Countries, Sample Sizes, and Quality Assessment (MMAT)

The study designs included RCTs (n = 3) (Ansley et al., 2021; Round et al., 2022; Pozo-Rico et al., 2020); and quasi-experimental trials (n = 3) (Zadok-Gurman et al., 2021; Oliveira et al., 2022; Matiz et al., 2020). The included studies were conducted in six countries: the United States (Ansley et al., 2021), Portugal (Oliveira et al., 2022), United Kingdom (Round et al., 2022), Israel (Zadok-Gurman et al., 2021), Italy (Matiz et al., 2020); and Spain (Pozo-Rico et al., 2020). Sample sizes ranged from 51 to 141 participants per study, with 454 participants across all studies. The studies included only nationally representative samples (See Table 2 for an overview).

MMAT has different criteria regarding research design types. According to these guidelines, three studies were classified as RCT (Ansley et al., 2021; Round et al., 2022; Pozo-Rico et al., 2020) and three were classified as QE (Matiz et al., 2020; Oliveira et al., 2022; Zadok-Gurman et al., 2021). Following the MMAT, all studies had clear research questions, and the data allowed for answering these questions. Moreover, two of the RCT studies incorporated a fidelity measure (Ansley et al., 2021); in the study by Round et al. (2022), they imposed a time limit on the computer for the writing assignment, which we interpreted as a fidelity measure. All QE studies employed fidelity measures (Oliveira et al., 2022; Matiz et al., 2020; Zadok-Gurman et al., 2021). Notably, only one RCT study (Ansley et al., 2021) detailed the randomization process according to the standards set by Hong et al. (2018).

4.2. Sample Recruitment and Inclusion Criteria for Participation in the Studies

Five studies reported the modes of recruitment (Ansley et al., 2021; Zadok-Gurman et al., 2021; Pozo-Rico et al., 2020; Matiz et al., 2020; Round et al., 2022), but one study did not (Oliveira et al., 2022). The studies reported that the participants were recruited through email or other forms of electronic marketing via teaching centers, schools, or social media, targeting teachers who were experiencing stress and were interested in participating in a professional well-being intervention. Moreover, the inclusion criteria for participation in the studies differed. The first study used the criterion of having no prior experience with the technique introduced in the intervention (Zadok-Gurman et al., 2021), while the second study required that the participants be employed as elementary school teachers (grades 1–4) during the intervention and met three exclusion criteria: (1) teachers who did not have an assigned class, (2) teachers in coordinating and/or supporting roles within the school cluster; and (3) teachers responsible for teaching extracurricular activities (Oliveira et al., 2022, p. 8). The third study required fluency in English, an absence of any previous clinically diagnosed mental health condition, and full-time employment over 18 years of age (Round et al., 2022). The fourth study required that the participants had attended a particular technological course the previous year (Pozo-Rico et al., 2020). Finally, two studies had no requirements beyond being a teacher to participate in the interventions (Ansley et al., 2021; Matiz et al., 2020). Two studies offered compensation for participation, such as gift cards (Ansley et al., 2021) and continuing educational training credits (Oliveira et al., 2022), whereas the remaining studies did not. Hence, participation requirements and rewards were heterogeneous.

4.3. Covariates Age and Sex on Burnout

Although the covariates of age and sex were not part of the research questions, we chose to include them in our analysis, as previous studies have found their relationship with burnout and technostress to be inconclusive. We found that the following four studies did not analyze age and sex in conjunction with burnout measures: Ansley et al. (2021), whose participants were primarily under 35 years old (10 women and 41 men); Matiz et al. (2020), whose sample consisted of 58 female teachers (mean age = 50.8, SD = 8.0); Pozo-Rico et al. (2020), with 80 women and 61 men (mean age = 38.4 years, SD = 6.84); and Round et al. (2022), with 54 women and 12 men (mean age = 38.1, SD = 12.3). Additionally, no differences in age and sex in conjunction with the impact on burnout measures were found in the study by Oliveira et al. (2022), with 78 women and three men (mean age of 46.21, SD = 4.83) and in the study by Zadok-Gurman et al. (2021), with 58 women and nine men, aged between 34 and 67 years (mean age = 45). Only two of the six studies analyzed whether age and sex affected burnout measures. These two studies found no conclusive correlations between age, sex, and burnout. However, the sample size was too small to draw definitive conclusions.

4.4. How Were the Interventions Delivered?

We present our findings regarding the delivery of content in the interventions, the duration and dosage of the interventions, and their impact on burnout measures. In addition to our discoveries regarding the core components of the included studies, we examined their application of JD-R theory, their impact on burnout measures, and their fidelity measures. First, we examined how the intervention content was delivered, focusing on the digital modalities used and the respective support systems. We were also interested in whether content was delivered asynchronously or synchronously. One advantage of asynchronous hours is that participants can choose their preferred participation time.

Table 3. Digital modalities, format for training, and support system.

		Studies
		1. (Round et al., 2022)	2. (Ansley et al., 2021)	3. * (Matiz et al., 2020)	4. (Oliveira et al., 2022)	5. * (Zadok-Gurman et al., 2021)	6. (Pozo-Rico et al., 2020)
Digital modalities	Videos/audio files		x	x
	Zoom/Teams e-learning Moodle platform				x		x
	Not specified, only via the internet or on an online platform	x	x			x
Training hours	Training during asynchronous hours	x	x	x	x	x
	Training during synchronous hours			x	x	x	x
Group training format	Group discussions w/facilitator or coach online meeting				x	x	x
	Group discussions w/facilitator or coach physical meeting			x		x
Individual training/ homework	Individual training/ homework	x	x	x	x	x
Materials and communication	E.g., emails, materials, articles, books.	x	x	x

* Changed to a digitized mode of delivery owing to COVID-19.

presents an overview of the digital modalities used in this study.

Our findings revealed that the degree of digital modalities in delivering interventions varied greatly. The coding for the type of digital modality was divided into the following categories: videos/audio files, Zoom/Teams, and the Moodle platform, or cases where no specific platform was reported other than that being delivered via the Internet or online. Phones, E-mails, and SMS were also used to communicate between facilitators/coaches and participants. Two studies (Zadok-Gurman et al., 2021; Matiz et al., 2020) changed their mode of delivery from in-person to digitized after the intervention had started owing to the COVID-19 lockdown, while the remaining interventions were planned using technological modalities from the design stage. These two interventions were delivered entirely digitally, allowing participants to engage at their convenience during asynchronous hours (Ansley et al., 2021; Round et al., 2022). According to our definition, these two studies are digital interventions. We consider the remaining four studies digitized interventions (Oliveira et al., 2022; Matiz et al., 2020; Pozo-Rico et al., 2020; Zadok-Gurman et al., 2021). These four interventions were delivered in a blended format, implying that some of the content was delivered through digital modalities; however, the intervention content was also discussed in groups during synchronous hours led by trained coaches/facilitators. These discussions took place in different modalities: online (Oliveira et al., 2022; Pozo-Rico et al., 2020), in-person (Matiz et al., 2020), or a combination of both (Zadok-Gurman et al., 2021). Zadok-Gurman et al. (2021) offered individual and group meetings with each participant, whereas the other studies did not include one-to-one training with a trainer/coach. However, four studies reported assigning individual work between sessions to further bolster and activate the content of the interventions (Ansley et al., 2021; Oliveira et al., 2022; Matiz et al., 2020; Zadok-Gurman et al., 2021).

We investigated how communication was administered in each study. For instance, in the study by Ansley et al. (2021), participants received a welcome email within 24 h of enrollment, a course guide, and an optional workbook. If they missed a session, they were individually messaged about the suggested pace of the course, in addition to regular weekly updates and plans. Although the intervention was delivered entirely digitally, communication was extensive, with reminders and extra materials. In a study by Matiz et al. (2020), participants were asked to request phone calls, discussions, books, and articles to read outside of scheduled meetings. In the study by Round et al. (2022), emails were sent to remind the participants of their involvement, and a computer program timed them so that they did not need to keep track of the time themselves. In three studies, no specific support system was reported (Oliveira et al., 2022; Pozo-Rico et al., 2020; Zadok-Gurman et al., 2021). Thus, the potential for disparate degrees of activation and integration of content depended on the heterogeneity in duration and dosage and the required homework between sessions. Finally, the greatest distinction was the presence or absence of group discussion sessions led by a trained facilitator who could answer questions, listen, encourage, and motivate the participants throughout the interventions. In conclusion, the reported support systems varied across studies, regardless of the duration and digital delivery mode.

4.5. Duration and Dosage of the Interventions

The current study found that the duration of interventions spanned from three days (Round et al., 2022) to 140 days (Zadok-Gurman et al., 2021), with a mean of 65.8 days (9.4 weeks). The shortest dosage consisted of three sessions of 20 min each (one hour combined) (Round et al., 2022), whereas the longest dosage was up to 50 h (Oliveira et al., 2022). However, dosages varied independently of the duration of the interventions (See

Table 4. Core components using the JD-R theory.

Study	Personal Resources			Job Resources		Job Demand
	Stress Mastery	SEL/ EI	PE	ICT	Org. Skills	Techno Stress
(Ansley et al., 2021)	x	x	x
(Matiz et al., 2020)	x
(Oliveira et al., 2022)		x			x
(Pozo-Rico et al., 2020)	x	x		x	x	x
(Round et al., 2022)	x
(Zadok-Gurman et al., 2021)	x

Stress mastery = Mindfulness, meditation, visualization, yoga, stress reduction, cognitive restructuring. SEL = Social–emotional learning. EI = Emotional intelligence. ICT: Information Communication Technology. Org. skills = Organizational skills. JD-R: PR = Personal resources. JR = Job resources. JD = Job demands. PD = Personal demands.

for an overview). For instance, one study lasted 70 days with a dosage of 50 h (Oliveira et al., 2022), whereas another study lasted 140 days with a dosage of 45 (Zadok-Gurman et al., 2021). Hence, the dosage was similar; however, the duration of the latter study was twice as long. As mentioned, the shortest intervention, conducted by Round et al. (2022), had a three-day duration and a total dosage of one hour. The second shortest study required four mandatory hours, with an intervention duration of 28 days (four weeks) (Ansley et al., 2021). The medium-long interventions lasted for 56 days (eight weeks) for 16 h (Matiz et al., 2020) and 70 days for 50 h (Oliveira et al., 2022). Finally, the two longest interventions lasted 98 days with 14 modules of unspecified duration (Pozo-Rico et al., 2020) and 140 days with 45 h of dosage (Zadok-Gurman et al., 2021). The two shortest interventions did not result in significant changes in burnout measures; these studies lasted one hour (Round et al., 2022) and four hours (Ansley et al., 2021), respectively. However, the remaining four interventions significantly decreased emotional exhaustion (Matiz et al., 2020; Oliveira et al., 2022; Pozo-Rico et al., 2020; Zadok-Gurman et al., 2021). Moreover, two studies reported a significant increase in personal accomplishments (Matiz et al., 2020; Pozo-Rico et al., 2020), whereas only one study found a significant decrease in depersonalization (Pozo-Rico et al., 2020). This study was the only one that incorporated core components to decrease job demands. At this point, the intervention duration and dosage appear to reach a point of diminishing return, which is difficult to discern based on the available data. Nevertheless, the results indicated that interventions lasting at least eight weeks, with a minimum dosage of 16 h, were effective in producing measurable changes.

4.6. Analyzing Core Components Using the JD-R Theory

We evaluated the core components using JD-R theory as our framework to determine whether each core component aimed to increase personal or job resources or decrease job or personal demands. The results showed that the core components of all the interventions included in the current study primarily focused on increasing personal resources. Personal resources included stress mastery techniques, such as cognitive restructuring, stress management, mindfulness, and meditation. Additionally, social–emotional learning (SEL) and emotional intelligence (EI) were reported in three studies (Ansley et al., 2021; Oliveira et al., 2022; Pozo-Rico et al., 2020), whereas physical exercise was reported in only one study (Ansley et al., 2021). Two studies emphasized increasing job resources: resolving organizational conflict (Oliveira et al., 2022) and increasing technological competence pertinent to daily work tasks (Pozo-Rico et al., 2020). Only one study targeted decreasing job demands related to technostress by providing ICT training (Pozo-Rico et al., 2020). However, no other studies have focused on reducing job demands.

A brief overview of each intervention is provided below: We begin with studies that focused exclusively on increasing personal resources (Ansley et al., 2021; Matiz et al., 2020; Round et al., 2022; Zadok-Gurman et al., 2021). First, Ansley et al. (2021) focused on teachers’ self-reflections, the development and selection of coping mechanisms to manage stress, and the enhancement of social–emotional competencies that promote positive learning experiences. These competencies include increased teacher efficacy and lower burnout rates through mindfulness, relaxation, cognitive restructuring, social support, and physical exercise. Although it was a digital intervention, the participants received structured support through written materials, such as a workbook, course guide, instructions, and instructional videos. They defined their goals and developed plans accordingly. The intervention consisted of eight modules, information topics, activities, and independent practices. The control group did not receive any treatment. Second, Matiz et al. (2020) conducted an eight-week Mindfulness-Oriented Meditation (MOM) training program. The program was not explicitly tailored for teachers. All participants received the same training but were divided into two groups for comparative analysis, depending on whether they scored low or high on resilience at baseline. Resilience can be defined as the ability to “bounce back” from adversity (Pooley & Cohen, 2010). Third, the stress reduction intervention by Round et al. (2022) required participants in the positive writing group to “write about their thoughts and feelings surrounding intensely positive previous experiences” (p. 5). They could write about the same or different experiences. Participants in the neutral writing group were assigned daily writing topics. On day 1, they were instructed to write about plans for the rest of the day; on Day 2, they described the shoes they wore; and on Day 3, they wrote about their bedrooms. Writing sessions were limited to 20 min per condition, enforced by a timer on the computer screen, which prevented participants from exceeding the allotted time. Finally, Zadok-Gurman et al.’s (2021) intervention was an inquiry-based stress reduction (IBSR), which is an adaptation of Byron Katie’s The Work, Four Powerful Questions (Katie, 2002). This intervention focused on increasing personal resources for better stress management. It emphasized identifying stressful thoughts, writing them down, and critically examining them by answering the following four questions: Is it true? Can I know if this is true? How do I react when I believe that it is thought? Who would I be without thinking? Finally, participants identified possible evidence of the opposite.

As mentioned previously, two studies focused on increasing job resources: Oliveira et al. (2022) and Pozo-Rico et al. (2020); the latter also focused on decreasing job demands. Oliveira et al. (2022) conducted a contextually focused intervention, considering teachers’ perceptions of the work environment. The teachers belonged to the three school clusters. The intervention targeted personal and job resources within social–emotional competencies, including personal organization and time management, emotional awareness and regulation, conscious communication, conflict management, and personal leadership. Pozo-Rico et al. (2020) conducted the intervention with the broadest focus, aiming to increase personal and job resources and decrease ICT job demands. They emphasized increasing digital skills and emotional intelligence among teachers in the intervention group, while the control group did not receive any treatment. The content concentrated on the principles and potential applications of ICT competency, lesson planning, practical teacher tools for dealing with adverse emotions, understanding emotions, self-realization, emotional awareness, and expressing emotions positively.

4.7. Effects of Interventions on Burnout Measures

All included studies used a version of the Maslach Burnout Inventory (MBI) (e.g., Maslach et al., 1996; Maslach & Jackson, 1981).

Table 5. Duration, dosage, content focus, effects, and fidelity.

Study	Duration of Days	Dosage Required Hours	Content Focused 1, 2	JD-R	Effects and p-Value on Burnout	Fidelity
* (Round et al., 2022)	3	1	1	PR	No significant effect.	The computer program only allowed 20 min of writing activity at a time.
* (Ansley et al., 2021)	28	4	1	PR	No significant effect. Emotional exhaustion: ηρ2 = 0.09 (p = 0.051) Depersonalization: ηρ2 = 0.07 (p = 0.071) Personal Accomplishment: ηρ2 = 0.05 (p = 1.30)	Two independent reviewers analyzed the program’s content using a fidelity checklist developed by the first author.
(Matiz et al., 2020)	56	16	1	PR	Emotional exhaustion: ηρ2 = 0.240 (p = 0.001) Group Time: ηρ2 = 0.108 (p = 0.01) Personal accomplishment: 0.157 (p = 0.002)	Self-recorded meditation practices in diaries
(Oliveira et al., 2022)	70	50	1 and 2	PR, JR	Emotional exhaustion β = −0.84 significant, SE = 0.40, 95%.	A trained observer at all 30 training sessions completed at SSOG. The second trainer filled in 1/3 of the time.
(Pozo-Rico et al., 2020)	98	*14 modules, but no specified length for each	1 and 2	PR, JR, and JD	Emotional exhaustion: ηρ2 = 0.73 Depersonalization: ηρ2 = 0.79 Personal accomplishment: ηρ2 = 0.97 (p = 0.001)	None reported.
(Zadok-Gurman et al., 2021)	140	45	1	PR	Emotional exhaustion: d = 0.752 (p = 0.01) Personal accomplishment: Borderline significant	Reports that all lessons were standardized and assessed.

Legend: Content-focused: 1 = Individual. 2 = Context. JD-R: PR = personal resources. JR = Job resources. JD = Job demands. PD = Personal demands. * = digital intervention.

presents an overview of the included studies and their respective results on the MBI. These studies reported their results using different statistical measures: partial eta squared (Ansley et al., 2021; Matiz et al., 2020; Pozo-Rico et al., 2020), beta coefficient (Oliveira et al., 2022) and Cohen’s d (Zadok-Gurman et al., 2021). The conventional cut-off values were set as follows: Cohen’s d = 0.2, small, 0.5 = medium, and 0.8 = large. For partial eta-squared values, the cutoff values were: 0.01 = small, 0.06 = medium, and 0.14 = large (Ellis, 2010, p. 41). As shown in Table 5, only two studies reported all three dimensions of the MBI (Ansley et al., 2021; Pozo-Rico et al., 2020).

The studies are included in Table 5 and organized according to the duration of the interventions. The first two studies, which had the shortest durations and focused on increasing personal resources, did not show any significant changes (Round et al., 2022; Ansley et al., 2021). These are both digital interventions. The study by Matiz et al. (2020), which also focused on increasing personal resources, resulted in a large effect size (ηρ² = 0.240) for the decrease in emotional exhaustion; the decrease was higher in the low-resilience group than in the high-resilience group. Both groups also experienced a substantial increase in personal accomplishment, with the high-resilience group experiencing a greater increase than the low-resilience group (ηρ² = 0.157, p = 0.002). There was no change in the depersonalization measures. Oliveira et al. (2022) focused on increasing personal and job resources and reported a significant effect (β = −0.84, SE = 0.40), with a 95% confidence interval indicating a decrease in emotional exhaustion. Teachers with the most supportive colleagues also had the lowest burnout baseline emotional exhaustion scores, while teachers with the worst work environments benefited the most from the intervention and experienced decreased emotional exhaustion. The findings demonstrated that teachers’ perceived context affected the intervention results. Depersonalization and personal accomplishments have not yet been reported in all the studies. Pozo-Rico et al. (2020) reported significantly favorable changes in all three dimensions. The intervention focused on decreasing job demands and increasing job and personal resources. With a large effect size across all scales, the study reported a reduction in emotional exhaustion (ηρ² = 0.73), a decrease in depersonalization (ηρ² = 0.79), personal accomplishment (ηρ² = 0.97, p = 0.001). Finally, the study by Zadok-Gurman et al. (2021), which focused on increasing personal resources, reported a significant medium effect size for a decrease in emotional exhaustion (d = 0.752, p = 0.01). Personal accomplishments did not change. However, depersonalization was not reported.

4.8. Fidelity Measures and Use of Control Groups

Measuring fidelity and using control groups were addressed differently across the interventions. A fidelity checklist was used in one study (Ansley et al., 2021). Two trained observers completed an observation grid to evaluate fidelity in one study (Oliveira et al., 2022). Self-reports of adherence to interventions were also documented (Matiz et al., 2020). A timer was set for 20 min of writing time on the computer, which helped ensure adherence to protocol. We interpreted that as a fidelity measure (Round et al., 2022). In contrast, the study by Zadok-Gurman et al. (2021) reported that all sessions were standardized according to the training manual and assessed to maintain consistency. One study did not report any specific measures (Pozo-Rico et al., 2020). Although it can be argued that five out of the six studies did report fidelity measures, how they were assessed varies. Not only did the fidelity measures vary, but we also found that the control groups were treated differently across the studies. In the studies conducted by Ansley et al. (2021), Oliveira et al. (2022), and Pozo-Rico et al. (2020), the control group did not undergo training. In contrast, Matiz et al. (2020) divided participants into two groups depending on their baseline scores (high or low resilience); both groups received the same training. In the study by Round et al. (2022), the control group participated in the intervention but completed a different writing task while adhering to the same parameters: 20 min writing sessions over three days. Furthermore, in a study by Zadok-Gurman et al. (2021), participants in the control group attended a course for an equivalent duration; however, the same information was not covered. Thus, establishing a clear pattern is difficult owing to the substantial variation in the fidelity reporting and the differing treatments of the control groups.

4.9. Summary of Main Findings

This study explored digital and digitized interventions aimed at reducing teacher burnout and increasing their work engagement. We examined the core components and their implementation. All studies focused on burnout as a dimension, using the Maslach Burnout Inventory (Maslach & Jackson, 1981). Four of the six studies reported decreased emotional exhaustion, these were all digitized interventions with support systems (Matiz et al., 2020; Oliveira et al., 2022; Pozo-Rico et al., 2020; Zadok-Gurman et al., 2021), two reported increased personal accomplishment (Matiz et al., 2020; Pozo-Rico et al., 2020), and one reported decreased depersonalization (Pozo-Rico et al., 2020). Applying the JD-R theory to analyze our results, we found that core components focusing on increasing personal resources—specifically stress management and emotion regulation (cognitive restructuring, stress management, mindfulness, and meditation)—were the most prevalent, as they were included in all six studies. Only two studies have focused on increasing teachers’ job resources through courses in ICT and organizational skills (Oliveira et al., 2022; Pozo-Rico et al., 2020). Moreover, only two studies—Pozo-Rico et al. (2020) and Matiz et al. (2020)—reported increased personal accomplishment. Finally, only one study measured baseline technostress and attempted to reduce job demands related to technostress (Pozo-Rico et al., 2020). To our knowledge, this is the only study to report a reduction in depersonalization. A common denominator among the four studies that significantly decreased emotional exhaustion (Matiz et al., 2020; Oliveira et al., 2022; Pozo-Rico et al., 2020; Zadok-Gurman et al., 2021) was the inclusion of discussion groups facilitated by a coach/trainer either via online platforms or in person. These results align with the conceptual framework proposed by Domitrovich et al. (2008), which emphasizes that an adequate support system is crucial for communication, inspiration, and motivation among participants. The two digital studies did not result in significant changes (Round et al., 2022; Ansley et al., 2021).

5. Discussion

Dosage, Core Components, and Effects on Burnout

This systematic review aimed to investigate the core components and implementation of digital and digitalized interventions to enhance teachers’ work engagement and reduce burnout. Once the core components are established, it is essential to determine the optimal duration, dosage, mode of delivery, support systems, and fidelity of interventions to achieve maximum effectiveness. As reported in the results, four out of the six studies included in this review showed a significant reduction in teachers’ emotional exhaustion (Oliveira et al., 2022; Matiz et al., 2020; Pozo-Rico et al., 2020; Zadok-Gurman et al., 2021). These interventions included various core components, such as stress reduction and mindfulness techniques. Using the JD-R theory, studies have focused on increasing personal resources. For instance, increasing teachers’ mindfulness and/or developing emotional intelligence has been the primary focus of several studies (Ansley et al., 2021; Matiz et al., 2020; Pozo-Rico et al., 2020; Zadok-Gurman et al., 2021), while two studies specifically emphasized emotional intelligence development (Oliveira et al., 2022; Pozo-Rico et al., 2020). These findings indicate that scaffolding teachers’ personal resources appears to be valuable in decreasing emotional exhaustion and aligns with the JD-R theory, as personal resources can buffer against job demands and reduce emotional exhaustion (Bakker et al., 2008; Xanthopoulou et al., 2007). Even when the intervention was not tailored to the occupational challenges specific to teachers, as seen in Matiz et al. (2020), it still led to a decrease in emotional exhaustion. These results are consistent with previous research, demonstrating that mindfulness is a powerful technique for stress reduction (e.g., Carvalho et al., 2021; Grossman et al., 2004).

The included study by Ansley et al. (2021) employed equivalent core components to the ones that achieved significant effects. However, the main difference was that the intervention was delivered entirely digitally and did not include support systems like a discussion group. Thus, we speculate that the content of the intervention could not be effectively integrated and internalized through discussions with other participants, nor could a facilitator/coach adapt the approach to suit specific teachers in a group. Similarly, Round et al. (2022) found no significant result; they implemented a digital stress reduction intervention without facilitating group discussions. The two digital intervention studies had the shortest durations and dosages among the six included studies. The study by Ansley et al. (2021) had a very low dosage of only four hours of mandatory duration. The study by Round et al. (2022) lasted only 20 min per day over three days, with a total duration of 60 min. Additionally, we theorize that the content of the intervention was not particularly relevant to reducing burnout in these teachers, as the exercise required participants to “write about the thoughts and feelings surrounding intensely positive previous experiences” (p. 5). Teachers were allowed to write about the same event three times. However, writing about the same event three times in three consecutive days may have provided insufficient exposure to positive emotions. This finding is not unexpected, as standardized mindfulness training typically consists of eight weeks of eight group meetings, each led by an instructor, along with daily individual practice (Kabat-Zinn, 2003). Another study reported a decrease in emotional exhaustion with a medium effect at a dose of 20 h across 10 weeks (Cooley & Yovanoff, 1996), which appears to be a suitable medium dosage for achieving the desired effect. According to Klingbeil and Renshaw (2018, p. 507), the dosage of mindfulness-based interventions exhibited diminishing returns beyond 24 h. However, care must be taken when generalizing these findings, as they were not specifically conducted on digital interventions. We could not discern whether the two longest interventions included in this systematic review (Zadok-Gurman et al., 2021; Oliveira et al., 2022) reached a point of diminishing returns before the post-test, as this was not measured in the respective studies.

However, Pozo-Rico et al. (2020) yielded the most robust results, as they achieved significant results in all three dimensions of the Maslach Burnout Inventory. Unfortunately, the exact dosage was unclear; they only reported that the intervention consisted of 14 modules but did not specify each module’s duration. They did not report any fidelity measures. We speculate that the results of Pozo-Rico et al.’s (2020) study could be attributed to the content of the core components, independent of the intervention’s duration and dosage. Nevertheless, our results are only partially congruent with the review by Hidajat et al. (2023), which found no difference in the results of mindfulness-based interventions based on duration or dosage. Based on the findings of our systematic review, the optimal dose and duration of mindfulness training for teachers should be at least eight weeks and more than 16 h in dosage to create positive change. This is partially aligned with the findings of the scoping review by Iancu et al. (2018), which found that interventions lasting less than a month were the least effective.

As the introduction mentions, few traditional, in-person interventions have focused on increasing job resources. The same seems to be the case for digital and digitized interventions. Only two studies in this review have incorporated core components to increase job resources: organizational and ICT skills (Oliveira et al., 2022; Pozo-Rico et al., 2020. However, it remains unclear why the teachers in Oliveira et al.’s study did not experience an improvement in their personal accomplishments, only a decrease in emotional exhaustion. This can be attributed to the nature of the core components. In contrast, teachers in the experimental group in Pozo-Rico et al. (2020) reported increased personal accomplishment. This could be explained by the mechanisms of the motivational process in JD-R theory, which states that increased job resources lead to increased personal accomplishment. The study by Matiz et al. (2020) also reported increased personal accomplishment for highly resilient teachers, although the core components only consisted of increasing their personal resources. Thus, the JD-R theory was only partially supported in this sample.

Interestingly, only one study examined the perception of organizational climate (Oliveira et al., 2022). Organizational climate refers to the characteristics perceived by personnel, making each work context unique and influencing worker behavior (Hoy & Tarter, 1992). The results demonstrated that the perception of organizational climate impacts both baseline measurements of emotional exhaustion and intervention effects (Oliveira et al., 2022). This aligns with previous research indicating that a positive organizational climate predicts fewer teacher burnout symptoms (Skaalvik & Skaalvik, 2018). As the remaining studies consisted of teachers from different schools who were randomly or simply assigned to experimental or control groups, we could not further compare the results regarding this variable.

As mentioned, a key study was the intervention by Pozo-Rico et al. (2020), which had the broadest focus, targeting the increase in personal and job resources and the decrease in the job demands associated with technostress, a contextual job demand. This is the only study that addressed the reduction in depersonalization. This aligns with previous research, as this burnout dimension depends on the work context and is the most challenging to address through interventions (Maslach & Leiter, 2016). Thus, we expected the depersonalization dimension to remain unchanged in all interventions except those focusing on decreasing job demands. Three studies (Ansley et al., 2021; Matiz et al., 2020; and Round et al., 2022) reported no changes in depersonalization. Two studies did not report the depersonalization scale (Zadok-Gurman et al., 2021; Oliveira et al., 2022). More research focusing on digital and digitized interventions impacting depersonalization and personal accomplishment may prove helpful in designing better interventions for teachers.

6. Implementation of the Interventions

6.1. Mode of Digital Delivery and Group Discussions

The digitized interventions utilized technology to varying degrees, with some incorporating discussion groups and other supporting materials to bolster and activate the content of the interventions (see Table 3). Two digital interventions (Round et al., 2022; Ansley et al., 2021) and one digitized intervention (Zadok-Gurman et al., 2021) provide vague details regarding the digital platforms used, as they only use descriptions such as “via online” or “digital platforms.” One plausible explanation could be that one of these interventions (Zadok-Gurman et al., 2021) was not initially planned for technological delivery and deviated from its original research design during the intervention. Participants were required to make delivery adaptations. The two digitally delivered interventions included in this study (Ansley et al., 2021; Round et al., 2022) did not result in a significant change in the burnout dimension. We argue that the difference between a digitized and digital intervention seems to impact the results, although it is too small a sample to draw firm conclusions. Digital interventions also appear to warrant a support system, but perhaps of a different character than digitized interventions. Another plausible explanation could be that a digital modality offering asynchronous hours might be of less value to teachers than to other occupational groups. Moreover, findings have revealed that an intervention delivered through a digital modality can introduce an additional demand, particularly affecting older teachers with less experience, confidence, and technological competence (Pagán-Garbín et al., 2024). However, we were unable to draw any conclusions regarding age or sex in conjunction with technostress. Thus, our results align with previous findings that indicate that it remains unclear whether viable differences in sex, age, personality profiles, and proneness to technostress exist (Pagán-Garbín et al., 2024). This trend warrants further investigation. While digital training provides many advantages, offering technical assistance can greatly enhance the learning experience for many individuals who lack technological experience. Falloon (2020) highlights that feelings of inadequacy, fear of criticism, and reluctance to seek help can obstruct digital learning. In the study by Pozo-Rico et al. (2020), one inclusion criterion was that teachers had completed an ICT course the previous year, and this intervention also incorporated an ICT course. These teachers received scaffolding in ICT competence, which may have positively influenced their burnout scores.

Support from trainers/coaches varied across the interventions, depending on their design. The following studies conducted group sessions: Oliveira et al. (2022), Matiz et al. (2020), Pozo-Rico et al. (2020), and Zadok-Gurman et al. (2021), with the latter also offering live discussions and individual training sessions (Zadok-Gurman et al., 2021). All of these studies resulted in significant changes in emotional exhaustion. Therefore, we argue that studies incorporating facilitated group discussions could have enhanced learning and provided a venue for belonging and peer support, which have been shown to foster teacher engagement and act as a job resource (Skaalvik & Skaalvik, 2018). Thus, conducting interventions for teachers within the same school may promote feelings of belonging while fostering individual growth, which might be optimal. This finding is aligned with Kolfschoten et al. (2012), who highlight the importance of trained facilitators for managing group dynamics, ensuring psychological safety, and promoting meaningful engagement.

6.2. Fidelity Measures and Use of Control Groups in the Studies

The fidelity of the study, the use of control groups, and the outcome measures are essential for interpreting the results. Measuring fidelity is important to avoid type III errors, which means erroneously concluding that there was no change due to the intervention content when the failure was due to implementation failure. The quality assessment indicated that the six studies analyzed exhibited good methodological quality and low potential bias risks. As previously reported, five of the six studies reported a type of fidelity measure. In the study by Zadok-Gurman et al. (2021), which included individual training sessions, the possibility for individual adaptations could be highly beneficial but also more challenging to maintain equality among participants. However, how these standardized sessions were assessed is unclear. In addition, the lengths of the modules in the study by Pozo-Rico et al. (2020) and the homework required in this study were unknown. Thus, the possible content adaptations are also unknown. Finally, the study conducted by Pozo-Rico et al. (2020) had an inclusion criterion requiring that teachers completed ICT training the previous year, which may have led to a sample biased towards more motivated and tech-savvy teachers compared to the general teacher population. Thus, testing this intervention with a broader range of teachers could provide valuable insights. However, caution must be exercised when generalizing the results, as the inclusion and exclusion criteria varied among the different studies. Hence, we cannot fully evaluate these results because of missing information. In addition to fidelity measures, using control groups to measure whether core components lead to changes may be essential.

Random assignments to the intervention and control groups help overcome selection bias, which would otherwise arise from placement or self-selection (White, 2013). All interventions employed either RCT or quasi-experimental designs. As two studies did not report how the randomization was conducted, this is unknown (Round et al., 2022; Pozo-Rico et al., 2020). In the studies by Ansley et al. (2021), Oliveira et al. (2022), and Pozo-Rico et al. (2020), the control group did not undergo training. Matiz et al. (2020) divided participants into two groups depending on their resilience baseline scores (high or low); both groups received the same training. In the study by Round et al. (2022), the control group participated in the intervention but completed a different writing task, writing task while adhering to the same parameters—20 min writing sessions over three days. Furthermore, in a study by Zadok-Gurman et al. (2021), participants in the control group attended a course for the same duration, but the same information was not covered.

6.3. Limitations

This study has several limitations. First, there could be a geographical bias, as all samples were from Europe, the US, and Israel, with none from Africa, Asia, or Australia. Second, we could not conclude the importance of asynchronous hours due to the small sample size. Third, only articles published in English were included, which may have resulted in the omission of the relevant research. Publication bias could be a limiting factor, as we excluded unpublished work (Cumming, 2014; Rosenthal, 1979). Fourth, all studies used self-reporting scales, which present the risk of skewed responses due to social desirability bias. Fifth, the sample included in the study had fewer males than females. However, there are more female teachers than male teachers. However, sex was not found to be a significant variable in these studies, nor was the focus on exploring sex-specific preferences for interventions. As a result, we were unable to explore any sex differences. Women are more prone to experiencing and developing symptoms of stress, anxiety, depression, and burnout than men (e.g., Matiz et al., 2020). Sixth, the generalizability of the findings in this study is limited due to the small number of studies and the low sample sizes within them. Finally, technological limitations and individual skills, such as digital literacy, may have influenced participants’ engagement with the interventions and outcomes (Hoffmann et al., 2014; Kintu et al., 2017). Given these limitations, future studies should incorporate larger sample sizes. In addition to conducting a meta-analysis of effect sizes, further comparisons of fidelity measures and longitudinal data could provide valuable insights. Moreover, mixed-methods studies that include qualitative interviews with teachers could uncover in-depth information about personal perceptions of technology used in interventions.

6.4. Future Research

This study has several implications for future research. First, a methodological concern arises from the variations in when the follow-up measurements were taken. Oliveira et al. (2022) showed little to no results were detected immediately after the completion of the intervention, with effects only becoming measurable after one month or longer. When the effect is only measurable after some time after the completion of an intervention, it is referred to as the sleeper effect (Van Wingerden et al., 2016). The recommended time for measurement is one to three months after the completion of the intervention, as the SEC found that most effects emerge later (Oliveira et al., 2022). Follow-up after the post-test to capture any potential sleeper effects is important (Iancu et al., 2018; Schaufeli & Taris, 2014) in ensuring confidence and reliability of the intervention processes (e.g., Durlak et al., 2015). Only one study measured follow-up after the post-test (Oliveira et al., 2022). According to Hall and Hord (2020), the timing of assessments may be of concern when fine-tuning optimal interventions. We were unable to draw any conclusions, as only two studies were conducted after the post-test. Second, the academic year should also be considered, as empirical findings suggest that teachers experience more exhaustion and have lower work engagement at the end of the academic year, regardless of personal and job resources, as well as job demands. Controlling for baseline measurements at the beginning of the academic school year can help determine whether these patterns persist. Hence, if the end of an intervention coincides with the end of the academic year, one might find that the potential effects of the intervention are counterbalanced by the time of year, due to natural fluctuation in teacher exhaustion.

Third, the use of digital capabilities to customize teachers’ professional development can be significantly improved with the assistance of AI. Just-in-time adaptive interventions (JITAIs) have yet to be applied to deliver the most suitable level of content in the intervention for an individual at a particular moment, while minimizing unnecessary burdens (Nahum-Shani et al., 2018). Adaptations can be made for each teacher during group and individual sessions to inspire and enhance their development. While this has not been reported in the articles, it could be a valuable addition for teachers during the interventions. This suggests that the field is still in its infancy when capitalizing on the possibilities of customization and just-in-time technological capabilities. Fourth, technological competence varies greatly; thus, for one teacher, it can offer a valuable modality, whereas for another, it may impose an additional burden in an already overburdened work situation (Bondanini et al., 2020). To further explore this variable, the technostress scale could serve as a pertinent measure.

Fifth, digital and digitized interventions could decrease job demands and increase job and personal resources for each teacher. Interventions that decrease job demands through job crafting or other approaches can benefit teachers’ professional well-being (Bakker et al., 2023). However, only one intervention focused on decreasing job demands by providing training in ICT (Pozo-Rico et al., 2020). Changing habits in terms of thoughts, behaviors, and psychological mechanisms requires time and maturation. Therefore, providing professional and personalized development to teachers in cohorts within the same schools, along with long-term support, might be ideal for teachers to internalize changes and self-regulate new ways of thinking (e.g., Hall & Hord, 2020) with support from colleagues and leadership, school climate, and organizational health (Domitrovich et al., 2008; Tikkanen et al., 2022). Consequently, SEL interventions must adopt a whole-school approach, involving all school personnel (Collie, 2020). The most effective interventions target bottom-up and top-down management initiatives (Bakker & de Vries, 2020). Schaufeli et al. (2023) examined whether one subscribes to the definition of burnout on a continuum or whether there are two types of burnout (burnout complaints and burnout disorder). In the latter case, they recommend person-centered interventions for burnout complaints, such as relaxation and mindfulness (e.g., Iancu et al., 2018), and more context-focused interventions in addition to person-centered burnout disorders (Ahola et al., 2017). This systematic review uncovered limited data on teachers’ work engagement and digital and digitized interventions. Further research that includes contextual factors is needed, so interventions should complement individual and contextual/organizational-level dimensions. Recent studies have suggested a positive association between a closed or unhealthy organizational climate and lower personal resources (Collie, 2020), and interventions should address teachers’ contextual factors to achieve higher intervention acceptability/social validity (Evans et al., 2022). We encourage fellow researchers to explore the optimal support system for digital and digitized interventions, acknowledging they may differ from what in-person interventions warrant (Halvorsen et al., 2025).

7. Conclusions

The current study focused on digital and digitized interventions for teachers to decrease burnout and increase work engagement using the JD-R theory (Demerouti et al., 2001; Bakker et al., 2023). Unfortunately, we could not locate any digitally delivered interventional studies with work engagement as an outcome measure, so the study focuses on burnout. The findings indicate that digitized interventions decrease teachers’ emotional exhaustion, with effects ranging from medium to large, independent of duration, dosage, and mode of delivery, and contingent upon facilitated group discussions for the participants to integrate the content of the intervention. Although digitalized interventions offer clear advantages regarding asynchronous hours, high fidelity, and potential scalability, it is important to consider the potential pitfalls of technostress. To mitigate this risk, an adequate support system should be provided by offering discussion groups, training scheduling, explanation of content, practicing new stress-reduction techniques, technical support, and so on, adapted to different levels of needs among the participants.

In this study, the digital interventions did not result in significant changes, which we speculate was due to several reasons: (1) they had a shorter duration and dosage; (2) they did not offer facilitated group discussions, which could have impacted the integration of the content; (3) without a sense of community, teachers do not experience a sense of belonging, inspiration, or motivation; (4) social validity cannot be quickly reinforced without a facilitator; and (5) differences in core components/or dosage of these. Further research is needed to establish the salient factors. Moreover, our findings are partially congruent with the JD-R theory, which posits that increased personal resources may lead to decreased emotional exhaustion. However, the same results were obtained when the intervention was not explicitly tailored to the job demands of teachers. Additionally, we found that interventions focusing on increasing job resources and decreasing job demands may impact depersonalization and personal accomplishment measures, but there were deviations. Although the sample was limited, the results mainly support the JD-R theory.

This systematic review aimed to enhance the understanding of key components and best practices for developing the most effective digital or digitized interventions for researchers, practitioners, and policymakers. It allows teachers to progress at their own pace and integrate their growth into personal schedules, thereby reducing time and costs. Although technology offers a transformative shift in training and professional development, careful implementation is essential. Key steps include defining core components, establishing a robust support system that features facilitated group discussions and technical assistance, and clarifying the nature of the intervention—whether digital or digitized. To ensure successful intervention and implementation outcomes that enhance teacher work engagement and mitigate burnout.