Health crisis communication represents a specialised domain within crisis communication, focused specifically on public health threats and the role of health authorities in managing information during emergencies (Hyer & Covello, 2005). While initially developed for disease outbreaks, health crisis communication frameworks are equally applicable to armed conflicts, which generate both direct health consequences and indirect effects (Jayasinghe, 2024; Mowafi, 2011; WHO, 2024).

While the topic of wars and conflicts as a public health crisis is not new, the indirect health consequences continue to be neglected; wars and armed conflicts constitute complex public health crises that extend beyond immediate casualties (Jayasinghe, 2024; Mowafi, 2011). Modern warfare's shift to the domestic front disproportionately affects civilians (Goto et al., 2022; Webster & Neal, 2022). These conditions demand sophisticated health crisis communication, as citizens must make informed health decisions while navigating misinformation that adversaries may weaponise (Patel et al., 2020; WHO, 2024). Contemporary armed conflicts increasingly affect civilian populations within their home territories, creating dual imperatives for health authorities: protecting public health while managing information flows to populations experiencing conflict directly (WHO, 2024).

The WHO recognised the critical role of crisis communication in public health settings long before the COVID-19 outbreak (Hyer & Covello, 2005; Rubinelli et al., 2023). Guidelines published by the WHO in 2005 emphasised the importance of engaging the public and helping them make informed decisions through effective media communication (Hyer & Covello, 2005). A meta-analysis of crisis communication research identified health communication and social media as two of the most extensively studied topics in this field (Upadhyay & Upadhyay, 2023).

Temporal frameworks, such as the Crisis and Emergency Risk Communication (CERC) model, organise crisis communication by phases (such as pre-crisis, initial, maintenance, resolution, and evaluation). They provide guidance for analysing functional message types or platform-specific strategic variations in digital environments (Reynolds & Seeger, 2005). However, these frameworks may be insufficient in war settings. Hence, the literature gap we aim to fill.

The widespread adoption of social media has significantly transformed public communication over the past two decades. Responding to public demand and transparency initiatives, public institutions now use social media to share information and promote engagement (Kass-Hout & Alhinnawi, 2013; Thackeray et al., 2012; Yavetz & Aharony, 2020).

Social media emerged as a valuable public health tool by the late 2000s. By March 2011, 30 American state health departments had already used social media platforms, though primarily for top-down, one-way messaging (Neiger et al., 2012; Thackeray et al., 2012).

Over the years, social media usage in public health has evolved, with applications in disaster management, disease tracking and surveillance, risk assessment, and health promotion and intervention (Blackstock et al., 2016; Neiger et al., 2012). While public health crises, such as Ebola, Yellow Fever, and Zika outbreaks, have demonstrated social media's effectiveness, they have also increased misinformation (Toppenberg-Pejcic et al., 2019; Wang et al., 2019).

The use of social media in public health and risk communication underwent significant changes with the outbreak of the Coronavirus disease 2019 (COVID-19) pandemic in early 2020, which impacted global societal, economic, and health status (Muhsen et al., 2021; WHO, 2020). Considering the pandemic and the resulting uncertainty, it became crucial for governments to take action and involve the public in health-related decisions that affected their daily lives while also keeping them informed (Liu & Tong, 2020; Ting et al., 2020). Government institutions worldwide have utilised various social media applications, including MIM apps such as Telegram (Nwammuo & Nwachukwu, 2021) and WhatsApp (Johns et al., 2023; Liu & Tong, 2020). However, high reliance on social media was associated with increased vulnerability to misinformation, particularly among older adults (Joshua & Vethanayagam, 2023). The high levels of uncertainty and public reliance on social media necessitate the development of effective health crisis communication practices (Rubinelli et al., 2023).

Social media are also associated with the rapid spread of false information, both misinformation and disinformation (Patel et al., 2020). The World Health Organization (WHO) has termed this an infodemic - an overabundance of information, some of which is accurate and some of which is not - during a global crisis and a period of low trust (WHO, 2020). The WHO has communicated with the global public during the pandemic through various channels, including Facebook, Instagram, Twitter, and Weibo (WHO, 2020).

During the COVID-19 pandemic, even before the subsequent 2022 Russian invasion, Ukraine faced Russian disinformation campaigns targeting health crisis communication and fuelling societal polarisation (Patel et al., 2020). The pandemic, the Russian invasion, and the disinformation campaigns compounded this complexity (Choudhary et al., 2022).

Ensuring improved access to health information was recognised as key to crisis communication (Maddah et al., 2022). Health authorities face an array of communication challenges, and the mitigation of misinformation stands as a crucial concern.

Mobile instant messaging (MIM) apps, such as WhatsApp, Telegram, and WeChat, have become integral to modern communication, offering diverse functionalities across personal, public, commercial, and organisational contexts with unique affordances.

WhatsApp, owned by META Corporation, is the most popular messaging service application in more than 100 countries, with over 2.5 billion active users (Iqbal, 2024), functioning as a global communication and business platform (Johns et al., 2023, p. 3). WhatsApp channels are utilised by journalists, news organisations, and governmental authorities alike (Johns et al., 2023, p. 115-118). During the COVID-19 pandemic, WhatsApp's broad reach and popularity prompted governments to create official channels to relay accurate information (Johns et al., 2023, p. 117).

Initially developed in 2013, Telegram is a cloud-based messaging application that offers several options for engagement, including private one-on-one conversations, group chats, and private and public channels controlled by admins (Walther & McCoy, 2021). Unlike other platforms, Telegram maintains minimal content takedown policies, removing only pornographic material and some violent rhetoric from public channels (Telegram).

Importantly, many Telegram online users seek information and passively engage with the content on the channels (Bloom et al., 2019), making it an effective platform for government entities to disseminate safety measures, health advisories, and emergency protocols in real time (Tindall & Diepeveen, 2024).

Activists have adopted Telegram in various political contexts (Bareikyte & Makhortykh, 2025; Santos et al., 2024). However, its minimal content takedown policies also enable authoritarian regimes and terrorist organisations to spread disinformation (Auslender, 2023; Wijermars & Lokot, 2022).

Telegram and WhatsApp differ in their affordances. Telegram is considered a hybrid platform for both private and public messaging services, in contrast to WhatsApp, which remains primarily private (Nobari et al., 2021). Notably, actual public reach in both platforms may extend beyond registered subscribers.

Concerning Israel, formal data demonstrates that 99% of the Israeli population uses WhatsApp, and 71% use Telegram. Hence, both platforms effectively reach national audiences (ISOC-IL, 2026).

Both MoH channels operate as one-way broadcast channels, where content is published exclusively by authorised administrators. While Telegram technically supports emoji and reply reactions, the MoH channel turned off this feature, limiting subscribers to passive reception only. On WhatsApp, subscribers could react to messages with emojis, the only feature available in this app. This asymmetric structure, common to official governmental MIM channels, reflects a deliberate choice to prioritise controlled information dissemination over interactive engagement (Ben-Gigi, 2024).

Israel has faced numerous crises since its inception, so it is better equipped to handle them (Yavetz & Bronstein, 2023). Israel presents a particularly data-rich digital environment for this case study. Prior to October 7, 2023, platform penetration among the Israeli population was already high: 97% used YouTube, 95% used WhatsApp, 90% used Facebook, and 54% used Telegram, making Israel one of the higher Telegram-adoption countries globally (Wiener & Stegman, 2023). The war dramatically accelerated this trend: Telegram penetration rose from 54% to 70% within a year, driven by a 200% spike in downloads in the war's opening days, with national communications provider Bezeq reporting a twofold increase in active Telegram devices and a 50% rise in average daily logins (Auslender, 2023; Wiener & Stegman, 2024).

The MoH maintains an active presence across multiple digital platforms, including Facebook, Instagram, YouTube, X (formerly Twitter), and TikTok, in addition to its MIM channels. As of 2026, the MoH's Facebook page leads its digital presence with approximately 560,000 followers, followed by Instagram (84,000), Telegram (78,500), and YouTube (25,700), making the MoH the leading Israeli government ministry in terms of digital following across all platforms. The MoH's Telegram channel, with approximately 82,000 subscribers at the time of the study, is thus best understood not in isolation but as part of a broader multi-platform communication strategy, in which MIM channels serve distinct functions. As Ben-Gigi (2024) explains, while content across platforms largely reflects official MoH positions, MIM channels involve deliberate sectoral and cultural adaptations, particularly on WhatsApp, launched earlier than planned, and chosen specifically for its capacity to reach population groups, including ultra-Orthodox and Arab communities, that largely abstain from mainstream social media platforms.

The Iron Swords War dramatically reshaped Israel's media consumption landscape. In the first two days of the conflict alone, seven major Hebrew-language Telegram news channels collectively gained over 440,000 new subscribers, a 52% increase, while 36.5% of active users reported having joined the platform only after October 7 (Orpaz, 2025).

As early as 2018, the MoH launched a dedicated Telegram channel, which it used extensively during COVID-19 to publish daily case updates and safety measures (Ministry of Health Telegram channel); data later migrated to a comprehensive public dashboard (Ministry of Health: World of Data).

The large-scale terror attack conducted by Hamas on October 7, 2023, and the ensuing war, the first of its kind in Israel since 2006, resulted in at least 1,878 casualties, 253 abductions, 10,500 injuries, and 500,000 Israeli citizens displaced from their homes, as of May 28, 2025 (AFP, 2023; Ghert-Zand, 2023; Haaretz, 2023a, 2023b).

Therefore, the MoH Telegram channel provided transparent and timely information to the public, helping to counter the increasing spread of misinformation and disinformation on the platform. By the end of October 2023, the MoH became the first Israeli ministry to operate a WhatsApp Channel, earlier than planned, to maximise public reach (Ben-Gigi, 2024) (Ministry of Health WhatsApp channel).

Israel is a unique case study: a highly developed country with universal healthcare, high health literacy (Levin-Zamir & Baron-Epel, 2020), and extensive digital infrastructure (David, 2022), home to many ethnicities with evident health disparities. Although terror attacks and armed conflicts have been relatively frequent over the years, the Iron Swords War was unprecedented, with the highest number of civilian casualties and the most prolonged duration since Israel's Independence War. The MoH's expanding use of social media and MIM apps to communicate crisis risks to the public highlights the need for a comprehensive evaluation of this practice during these unprecedented times.

The MoH is the supreme governmental authority responsible for health policy implementation and system oversight in Israel, though its budget is managed by the Ministry of Finance (Sabbah, 2019).

This study explores how the MoH communicated during the Iron Swords War on Telegram and WhatsApp during the first 60 days, analysing the ministry's use of MIM apps, its risk communication approach, and platform-specific differences.

The study period spans October 7 to December 7, 2023, coinciding with the Hamas attack on Israel and its immediate aftermath. The MoH's WhatsApp channel launched on October 26, 2023.

The research corpus comprises all 271 posts made on the formal channels of the Israeli MoH, Telegram (n=216) and WhatsApp (n=55), during the first two months of the Iron Swords War.

Through a mixed methods design, this research integrates descriptive statistics, content analysis, thematic analysis, and a background interview, effectively harnessing the benefits of each (Creswell, 2014).

A background elite interview with the Head of the New Media and Internet Department at the MoH, Mrs. Nilly Ben-Gigi, was conducted on January 16, 2024, via Zoom, in a single one-hour session. The interview covered four thematic domains: digital content policy and platform strategy; differentiation of content across the MoH's digital channels; public engagement and audience response; and the management of MIM communication during the Iron Swords War. The interview was recorded with the participant's consent. Its relevant insights were integrated into the study as contextual background, informing the interpretation of platform-level strategic choices rather than serving as primary data for coding. This method enables us to obtain insights otherwise not documented and to better understand the institutional context of the findings (Menzies, 2025; Scally et al., 2021).

Content analysis enables systematic quantification of the studied phenomenon (Schreier, 2012; Vanderstoep & Johnson, 2008), and descriptive statistics allow precise presentation of the sample's attributes (Shayib, 2013). Thematic analysis enables the identification, analysis, and description of main themes within a dataset (Boyatzis, 1998; Braun & Clarke, 2006). This inductive approach intentionally forgoes the application of a pre-existing theoretical framework in favour of allowing conceptual categories to emerge from the data, aiming to generate a context-sensitive theory in understudied settings (Braun & Clarke, 2006; Corbin & Strauss, 2014).

Following the work stages defined in the literature, inductive category development was integrated with systematic coding (Shkedi, 2003), determining the presence and meaning of concepts and terms within the texts (Cole, 1988).

All messages were manually extracted from the publicly accessible MoH Telegram and WhatsApp channels and systematically archived in a structured Excel dataset throughout the study period (October 7 - December 7, 2023). As all messages were originally published in Hebrew, both authors, native Hebrew speakers, conducted all reading, coding, and analysis in the original language. Quoted messages appearing in this article were translated into English by the authors. Multimedia messages (videos and photos) were coded based on their accompanying textual content and, where relevant, their visible contextual elements.

Then, all 271 messages were independently coded by both authors and divided into one of two categories: War-related or Non-war-related, following an example of binary classification (Greenwald et al., 2024). The war-related category included messages that explicitly referenced topics relating to the October 7 attacks, the Iron Swords War, or direct consequences thereof. The non-war-related category included messages which addressed routine public health matters. Additionally, the messages' formats, either plain text, videos, or photos, were extracted. Intercoder reliability for both binary classifications was perfect (Cohen's kappa = 1.00, 100% agreement).

A collaborative content analysis was then conducted to examine the topics raised by the MoH across its Telegram and WhatsApp channels, with each message classified according to its topic. Data was collected and organised using Excel and demonstrated numerically, reflecting the occurrence of the specified categories.

A second, inductive thematic analysis was conducted jointly to identify functional communication strategies emergent from the data, without imposing predetermined theoretical categories. Following Braun and Clarke's six-step process (Braun & Clarke, 2006), and while referring to the entire message corpus (n=271), we thematically analysed the messages by creating frames to identify the communication health strategies employed by the MoH in both MIM channels.

Following the identification of three communication themes, Acknowledging, Recommending, and Instructing, each author thoroughly read all the messages and categorised them into one of the established classifications. Reliability tests performed independently for the two coders yielded intercoder reliability results of 85.45% and 91.20%, and Cohen's kappa scores of 0.758 and 0.810 for WhatsApp and Telegram, respectively, indicating substantive agreement (Coleman et al., 2024).

Discrepancies between coders were resolved through discussion and mutual consensus, with no unresolved disagreements remaining in the final dataset. Topical classification into the ten identified categories was conducted collaboratively by both authors rather than independently, given the categories' clear definitional boundaries and mutually exclusive nature. As a result, formal intercoder reliability was not calculated for this stage; the binary and ARI classifications, which were coded independently, serve as the primary reliability evidence for the study.

All messages were collected from publicly accessible official MoH channels; no user interaction, private data, or personally identifiable information was accessed or retained. The background interview was conducted with Mrs. Nilly Ben-Gigi in her professional institutional capacity; as is standard practice for elite interviews with public officials, no ethics committee review was required (Liu, 2018; Ntienjom Mbohou & Tomkinson, 2022; Vaagland, 2024). The participant was informed of the study's aims, consented verbally to recording, and approved the use of attributed quotations. Figures reproducing official MoH posts are public institutional communications reproduced for academic analysis.

All MIM messages were classified according to the ten topics identified: Blood, milk and device donations, Call centre and hospital info, Casualties, Fake news, Hostage return, Ministry announcements, Public health advisories, Public health tips, Recruitment, and Stress management and mental health (Table 2; Supplementary Figures 1-10).

On Telegram (n=216), the top topics addressed by the MoH were public health advisories (48, 22.2%), followed by ministry announcements (20.8%), and data on casualties (18.5%). Additional topics that constituted a significant part of the Telegram channel were hostage return information (16.2%) and call centre and hospital information (9.3%).

On WhatsApp (n=55), the top topics addressed were ministry announcements (29.1%), followed by public health tips (23.6%), and stress management and mental health (12.7%).

Notably, the topic of public health tips was addressed significantly on the WhatsApp channel but not on the Telegram channel.

We have identified three communication strategies the MoH employs to inform the public. These have been integrated into a communication strategies model called the ARI model, which stands for Acknowledging, Recommending, and Instructing.

The acknowledging strategy aims to consistently deliver important, reliable, and formal information to the public (Figure 3). In routine, this category includes messages such as committee meetings, personnel changes, and updates from Israel's representatives in global health organisations. During the war, this strategy also included messages relaying war-related information to the public, such as call centre details, casualty data, and updates on the medical status of returning hostages.

A representative example is a message sent on October 14, 2023, directly addressing circulating misinformation: “Misinformation, rumours, and false reports are circulating on social media requesting the public to donate food for medical personnel at Barzilay Hospital. This is an absolute fake. Please pay attention to receiving information from authorised sources only” (Supplementary Figure 4).

The recommending strategy includes softer health information, aiming to recommend rather than instruct (Figure 4). In routine, this category included messages such as physical activity tips for diabetes patients, safety measures for holidays, and the importance of hygiene in preventing infectious diseases. During the war, this strategy also included messages conveying war-related tips, such as coping with stress, the importance of resilience for older adults, and mental health first aid videos.

For example, a WhatsApp message sent on December 7, 2023, the final day of the study period, offered Hanukkah safety tips including guidance on candle supervision, fire prevention, and safe frying practices, closing with “Happy and safe Hanukkah holiday!” (Supplementary Figure 8). This message illustrates the dual informational and affective function characteristic of the recommending category.

The instructing strategy constitutes another layer of information provided by the MoH, which is more practical and pragmatic, aiming to instruct proper health behaviours (Figure 5). In routine, this category includes messages such as instructions to boil water in certain areas, recall of food products, and warnings about the presence of animals with infectious diseases (e.g., a jackal with rabies found in the kibbutz of Tel Yosef). A clear illustration is a message sent on November 7, 2023, announcing a mandatory boiling obligation for drinking water in Beit She'an following microbial test results, specifying its application to “drinking, cooking, medication preparation and tooth brushing” and remaining “valid until further notice” (Supplementary Figure 7). During the war, this strategy also included messages relaying war-related instructions, such as asking the public to donate blood and breast milk, warning against specific misinformation, and recruiting volunteers for understaffed hospitals.

While the model is hierarchical, coding followed the highest-order strategy present in each message. Messages with instructing elements were classified solely as such, regardless of lower-level functions. This ensures mutual exclusivity while maintaining the model's logic. These categories form a hierarchical pyramid structure (Figure 6), with each superior information layer incorporating the preceding one. The acknowledging category supplies fundamental data, the recommending category elaborates on this by incorporating advisory recommendations, and the instructing category, positioned at the highest level, combines both informative and advisory elements with direct behavioural instructions.

Recommending messages frequently integrate informational and affective dimensions simultaneously, guiding audiences while offering reassurance during acute stress, hence reflecting the inseparability of informational and psychological needs in health crisis communication.

Of the 216 Telegram messages sent by the MoH, 153 (70.8%) were classified as acknowledging, 52 (24.1%) as instructing, and 11 (5.1%) as recommending. Notably, the communication strategies employed by the MoH differed between war-related messages (n=162) and non-war-related, routine messages (n=54) (Figure 7). Acknowledging dominated war-related messages (137, 84.6%), followed by instructing (8.6%) and recommending (6.8%). However, instructing dominated non-war related, routine messages (38, 70.4%), while the recommending did not appear in these messages (Table 4).

Of 55 WhatsApp messages, 29 (52.7%) were classified as acknowledging, 18 (32.7%) as recommending, and 8 (14.6%) as instructing (Table 3). As in the MoH's Telegram channel, the communication strategies employed differed between war-related messages (n=38) and routine messages (n=17). Acknowledging dominated war-related messages (25, 65.8%), followed by recommending (28.9%) and instructing (5.3%). Non-war-related, routine messages differed by communication strategies; recommending was used the most (7, 41.2%), followed by instructing (35.3%) and acknowledging (23.5%) (Table 4; Figure 8).