This study was a multicenter, cross-sectional study, integrating a quantitative survey (5-point Likert scale) with a qualitative analysis of open-ended responses to capture the complexities of real-time information processing.

Critically, the entirety of data collection was conducted contemporaneously in March 2026, strictly during the active, live phase of the escalating armed conflict involving the United States, Israel, and Iran, frequently referred to in US media as the “US-Israel-Iran War,” which featured an intensive missile offensive that commenced on February 28 and continued until a ceasefire was established on April 7, 2026. Capturing responses entirely within this operational window ensured that participant evaluations reflected authentic, real-time decision-making dynamics under immediate threat rather than post hoc or retrospective recall. Throughout this period, Israel was under a sustained missile offensive, resulting in numerous real-time MCIs that required immediate and complex operational responses from health care leadership. To target the specific MCI leadership echelons required for this study, the questionnaire was disseminated electronically through formal, national professional WhatsApp communities and established leadership forums, from 24 medical centers designated to receive patients during MCIs at the national level, including the formal networks of the Israeli Association for Emergency Medicine for emergency department directors and head nurses, the National Trauma Coordinators Network, the national emergency preparedness directors forum, the Israeli hospital executive leadership community, and the trauma unit directors network of the Israeli Association for Trauma and Emergency Surgery. These included specialized cohorts such as the “Forum of Acting Emergency Department Directors” and the “National Emergency Preparedness Officers Forum.” To ensure optimal participant engagement while minimizing digital survey fatigue, the questionnaire link was distributed a total of 2 times. The survey used a 5-point Likert scale (1=not at all, 5=to a great extent) to measure the frequency of social media use, levels of trust, and the resultant operational impact on leadership decision-making.

This study was conducted and reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) reporting guideline for cross-sectional studies.

Statistical analyses were conducted on complete available cases. Ordinal Likert-type items were summarized using medians and IQRs; continuous or derived scores were summarized using means and SDs where appropriate. Within-person platform-use comparisons were evaluated using paired nonparametric tests with Holm correction, and bivariate associations between composite predictors and outcomes were screened using Spearman rank correlations with Benjamini-Hochberg false discovery rate (FDR) correction.

For the additional analyses requested for the current revision, years of experience were examined against platform-use and trust items using Spearman rank correlations. Differences by professional qualification and current role were examined using Kruskal-Wallis tests; because several qualification categories were sparse, qualification categories were collapsed into registered nurse, physician, and other for this secondary analysis. The association between the availability of sufficient verification tools and exposure to false information was evaluated using the Fisher exact test, and associations between verification tools and ordinal decision influence items were evaluated using Mann-Whitney U tests. Associations between active search during missile launch or missile fall alerts and situational awareness were evaluated using Spearman rank correlations. These secondary analyses were considered exploratory, 2-sided, and interpreted with FDR-adjusted P values within each analysis family.

Adjusted cumulative logit models were fitted for manuscript-reportable ordinal outcomes using use intensity and trust and reliance as primary predictors, with experience and professional background variables included where estimable. Open-ended responses were summarized descriptively and used only as illustrative context rather than as a standalone qualitative strand.

The study was approved by the Tel Aviv Sourasky University Medical Center Institutional Review Board (approval ASAF-0172‐26) and conducted in accordance with the ethical principles of the Declaration of Helsinki and its later amendments. The requirement for written informed consent was waived by the institutional review board, and completion of the questionnaire was considered evidence of informed consent. The research instrument was developed and managed using the REDCap platform, a secure, web-based software application designed to support data capture and metadata-driven workflows for clinical research. All data were collected and stored anonymously to ensure participant privacy and confidentiality. No financial or other compensation was provided to participants.

The analytic sample comprised 76 respondents (Table 1). The largest professional qualification groups were physician (41/76, 53.9%) and registered nurse (25/76, 32.9%). The largest current role groups were hospital executive leadership (19/76, 25.0%) and ED director (12/76, 15.8%). The median experience in the current role was 7.0 (IQR 3.0‐13.0) years. Overall, 81.3% (61/75) reported confirming social media information against official sources before taking major action, and 65.3% (49/75) reported encountering false information during emergencies.

WhatsApp was the most frequently used platform (median 3.0, IQR 2.00‐4.00), followed by news channels (median 2.0, IQR 1.00‐3.00), whereas Telegram use was lower (median 1.0, IQR 1.00‐3.00; Table 2). All pairwise platform comparisons were statistically significant after correction except Telegram versus news channels. Among Telegram users, news channels and real-time field security channels were the most endorsed channel types.

Respondents rated social media as moderately helpful for situational awareness (median 2.0, IQR 2.00‐3.00). Perceived influence was modest for clinical decisions (median 2.0, IQR 1.00‐2.00) and operational decisions (median 2.0, IQR 1.00‐3.00), lowest for MCI announcement decisions (median 2.0, IQR 1.00‐2.00), and somewhat higher for team management during an MCI (median 2.0, IQR 1.00‐3.00).

The use intensity (α=0.741), trust and reliance (α=0.653), and decision influence (α=0.821) composites showed acceptable-to-good internal consistency. In FDR-adjusted screening, higher use intensity was associated with greater situational-awareness ratings (ρ=0.674; FDR P<.001) and greater operational-decision influence (ρ=0.672; FDR P<.001) (Figure 1).

Higher levels of trust and reliance were associated with greater clinical decision influence (ρ=0.637, FDR P<.001) and with the composite decision influence index (ρ=0.621, FDR P<.001; Table 3). To summarize the relative strength of the 2 composites per outcome, we plotted the difference between Spearman correlations (use intensity index minus trust and reliance index). The overall visual pattern was consistent across outcomes: use intensity aligned most strongly with situational awareness and operational decisions, whereas trust and reliance aligned more strongly with clinical decisions and the overall decision influence composite.

Additional secondary analyses addressed experience, professional role, verification capacity, and active searching (Table 1 and Multimedia Appendix 1). Years of experience were not significantly associated with individual platform use or trust items after FDR adjustment; the largest uncorrected signal was a weak inverse association with WhatsApp use frequency (ρ=-0.266; P=.03; FDR-adjusted P=.22). Professional qualification and current role also showed no significant differences across use and trust items after FDR adjustment, although current role differed in Telegram trust before multiplicity correction (Kruskal-Wallis H=14.350; P=.03; FDR-adjusted P=.21).

Reporting sufficient tools to confirm information was not associated with having encountered false information (Fisher exact test: odds ratio [OR]=1.091; P>.99), and it was not significantly associated with situational-awareness or decision influence items after FDR adjustment. In contrast, active search during alerts showed consistent positive associations with situational awareness, both during missile launch alerts (ρ=0.560; FDR-adjusted P<.001) and missile fall alerts (ρ=0.439; FDR-adjusted P<.001).

Figure 2 shows, for each outcome, the adjusted OR and 95% CI for a 1-unit increase in the use intensity composite (left panel) and the trust and reliance composite (right panel), with estimates aligned on a logarithmic axis and a reference line at no effect (OR=1).

The detailed numerical measures for ordinal models are summarized in Table 4. Use intensity remained the clearest independent correlate of situational awareness (OR 7.452, 95% CI 2.053‐27.056; P=.002), whereas trust and reliance did not add independent signal (OR 2.337, 95% CI 0.606‐9.017; P=.22). By contrast, trust and reliance was the dominant correlate of clinical decision influence (OR 6.811, 95% CI 1.481‐31.326; P=.01), while use intensity did not reach the conventional significance threshold (OR 3.267, 95% CI 0.873‐12.221; P=.08). Use intensity also showed the clearest adjusted association with operational-decision influence (OR 5.253, 95% CI 1.307‐21.109; P=.02), whereas trust and reliance was not independently associated (OR 1.840, 95% CI 0.48‐7.055; P=.37). Trust and reliance again provided the clearest adjusted signal for team management influence (OR 4.182, 95% CI 1.201‐14.563; P=.03), whereas use intensity was not independently associated (OR 1.049, 95% CI 0.338‐3.255; P=.93). For MCI announcement decisions, neither use intensity (OR 0.950, 95% CI 0.29‐3.116; P=.93) nor trust and reliance (OR 2.519, 95% CI 0.712‐8.905; P=.15) showed a clearly independent association.

A 2-profile exploratory clustering solution separated lower-engagement and lower-trust respondents from higher-engagement and higher-trust respondents; situational-awareness ratings differed across profiles (Kruskal-Wallis statistic 30.265; P<.001), and the decision influence composite differed across profiles (Kruskal-Wallis statistic 20.538; P<.001).

This study evaluated how hospital MCI leaders used informal digital platforms such as WhatsApp during wartime to guide real-time operational and clinical decisions. Our analysis revealed that platform use intensity directly enhanced situational awareness and operational decision-making. Conversely, clinical decision-making and real-time team management were strictly contingent upon participants’ level of trust and reliance on the information source.

The preliminary findings highlight a phenomenon that can be termed “informal tactical triggering.” This concept describes a scenario wherein rapid, informal information, such as a Telegram post or a WhatsApp video, reaches managers or decision-makers and acts as an operational trigger that jumpstarts the system into action well before official reports are received. A critical insight from the data suggests that during wartime, the initial data received via these platforms serve as a vital sensor for the system. Another clear signal from our analysis was behavioral: respondents who actively searched for information during missile launch and missile fall alerts reported higher situational awareness ratings.

Moreover, our secondary analysis revealed that professional experience, background, and qualifications were not significantly associated with informal platform use or trust patterns, presenting a divergence from established literature. Historically, digital health research suggests a pronounced generational and seniority gap in health care settings. For instance, Farsi (2021) [8] documented that senior or more experienced clinicians consistently exhibit greater skepticism, lower adoption rates, and heightened risk aversion regarding social media integration compared with their junior counterparts. In contrast, our findings demonstrate a uniform reliance on informal communication streams across all professional echelons and age groups during wartime MCIs. This phenomenon may indicate that under conditions of acute operational stress and extreme uncertainty, such as the real-time threats of an active missile offensive, traditional digital divides dissolve. During emergencies, the critical need for rapid situational awareness and informal tactical triggering supersedes individual personal and professional habits, thereby transforming informal platforms from a generational preference into a universal and operational necessity.

Furthermore, our multivariable ordinal regression models delineate a fundamental operational dichotomy: platform use intensity independently governs situational awareness and macrolevel operational decisions, whereas clinical decisions and direct team management are strictly mediated by subjective trust. This dissociation reflects the practical mechanics of how emergency leaders operate under extreme uncertainty. Macrolevel logistics and system priming, such as clearing hallway bottlenecks, reserving operating suites, or holding off-duty personnel on standby, carry an asymmetric risk profile in which early preparation is cheap, but unpreparedness is catastrophic. Consequently, commanders leverage raw, low-trust visual streams (eg, unverified scene footage) as an early “sensor” to prime physical capacity. In stark contrast, microlevel clinical execution, such as deploying invasive trauma teams, entails immediate patient safety risks. Hospital commanders therefore apply a rigorous cognitive threshold: informal digital intelligence triggers operational posture, but definitive clinical intervention requires high-trust validation and institutional corroboration.

This study has several limitations that should be considered when interpreting the findings.

First, the overall sample size is relatively small (N=76), and several role categories contained few participants. This may limit the statistical power of certain subgroup analyses and requires that these subgroup findings remain exploratory, which may also reduce broader generalizability. Relatedly, because the survey was distributed electronically across established, closed national leadership forums and professional group channels with overlapping memberships, the exact aggregate denominator of unique individuals who viewed the electronic invitation could not be determined with complete precision, precluding the calculation of a conventional response rate. However, both methodological constraints must be contextualized within the unique structural scale of the Israeli health care system. Nationwide, there are only 24 general hospitals with EDs designated as receiving sites during MCIs. Because the national pool of specialized trauma directors, ED managers, emergency preparedness officers, and executive leaders is inherently constrained (estimated at approximately 120‐150 eligible frontline commanders nationwide), a sample of 76 respondents does not merely represent a minor cohort, but rather constitutes a substantial and highly representative proportion of the country’s frontline wartime medical leadership, achieving an estimated substantive national coverage of more than 50%. Furthermore, although the overall cohort (N=76) represents a substantial proportion of frontline MCI leadership across all 24 authorized receiving hospitals, some specialized subgroups were represented by very small numbers of participants, including physician fellows (n=1), paramedics (n=1), and trauma coordinators (n=3). Therefore, subgroup comparisons, including nonsignificant Kruskal-Wallis tests, should be interpreted with caution, as the small subgroup sizes limit statistical power and may increase the risk of type II error. These analyses should therefore be considered exploratory, and nonsignificant findings should not be interpreted as evidence that meaningful differences between professional roles do not exist.

Second, because the questionnaire was distributed primarily through professional WhatsApp networks, this recruitment method could introduce selection bias toward respondents who are inherently more active on WhatsApp or favorable toward digital instant messaging. However, this potential bias is mitigated by the fact that these WhatsApp networks serve as the official, nationwide operational channels for Israeli health care leadership during emergencies, making them the most direct and realistic medium to reach this specific target population.

Finally, our analysis showed that having personal verification tools did not protect respondents from being exposed to false information. While this might seem like a limitation in our findings, it actually reveals an important insight: verifying real-time digital information cannot rely solely on the skills or tools of an individual clinician. Instead, it demonstrates that data verification must be built directly into hospital operations as an official organizational protocol.

The integration of informal digital intelligence into hospital MCI protocols represents an operational shift in emergency management. While formal prehospital channels remain the gold standard for verified data, informal digital streams, dominated by WhatsApp, have become indispensable environmental sensors for rapid situational awareness. Crucially, this study demonstrates that reliance on informal platforms is a universal operational reality that spans all age groups, professional roles, and seniority levels during acute wartime stress.

However, this reliance uncovers a critical vulnerability termed the verification paradox. While 81.3% (61/75) of decision-makers actively attempted to confirm digital intelligence before taking major actions, 68.4% (52/76) reported lacking adequate tools to verify information in real time. To bridge this systemic gap, health care systems should consider shifting from relying on individual ad hoc verification to establishing formal institutional protocols. Furthermore, disaster management frameworks could benefit from institutionalizing verification loops, which are structured operational protocols that rapidly cross-reference informal visual data from field scenes with incoming EMS radio traffic before committing clinical resources or over-triaging.

Finally, future disaster preparedness should address the inherent technical fragilities of relying on instant messaging apps. During kinetic warfare or large-scale mass casualty events, physical damage to cellular infrastructure, network congestion, or deliberate cyberattacks and communications blackouts pose risks to operational continuity. Emergency planners are therefore advised to ensure technical redundancy by integrating localized backup communication networks alongside digital intelligence protocols. A thorough understanding of informal digital channels, along with their strategic integration as operational triggers, can significantly enhance hospital readiness, provided it is supported by institutional verification frameworks and resilient communication infrastructure.