Introduction
Masking by healthcare personnel (HCP) is a core infection prevention and control (IPC) measure to reduce transmission of respiratory pathogens in healthcare settings. Respiratory viruses such as influenza, respiratory syncytial virus (RSV), and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are transmitted primarily by respiratory droplets and aerosols generated during routine patient care. Medical masks and respirators reduce HCP infection risk, supporting masking as an effective protective intervention. Reference Offeddu, Yung, Low and Tam1
During the COVID-19 pandemic, masking gained renewed prominence as both a source-control strategy and means of protecting HCP and patients. Early studies suggested even modest reductions in viral exposure achieved through widespread masking could substantially reduce transmission, particularly when implemented broadly across healthcare systems. Reference Goyal, Reeves and Thakkar2 In response, universal masking policies were rapidly adopted resulting in reductions in healthcare-associated respiratory viral infections (HARVIs). Reference Munigala, Ching and Wood3,Reference Wee, Conceicao and Foo4
With increasing vaccination uptake, rising population immunity, and declining SARS-CoV-2 incidence as the pandemic progressed, public health guidance and institutional policies shifted, leading many healthcare organizations to relax or discontinue universal masking requirements. Emerging data suggest discontinuation of universal masking may be temporally associated with increases in HARVIs, highlighting the continued relevance of masking beyond pandemic response. Reference Most, Phillips and Sebert5,Reference Pak, Chen, Kanjilal, McKenna, Rhee and Klompas6 Simultaneously, masking has been associated with operational and communication challenges, underscoring the need for carefully designed, risk-based approaches balancing IPC with feasibility and workforce considerations. Reference Mheidly, Fares, Zalzale and Fares7,Reference Lee, Cormier and Sharma8
Despite evidence supporting masking effectiveness, there is limited consensus on how masking policies for HCP should be implemented. Reference Chow, Lynch and Zerr9 Current approaches to operationalizing masking vary widely across institutions, with organizations relying on diverse indicators to guide decisions. Reference Chow, Lee and Lenahan10 Further, guidance from organizations such as the Centers for Disease Control and Prevention (CDC) and World Health Organization (WHO) lack standardized operational thresholds, leaving decisions largely to institutional discretion. 11
Data describing how healthcare institutions currently operationalize masking policies and the factors driving decision-making remain limited. This study aimed to characterize institutional masking policies for HCP across diverse healthcare settings and identify epidemiologic and organizational factors informing masking decisions to inform data-driven approaches to mitigating HARVI transmission.
Methods
Study design
We conducted a cross-sectional survey to characterize institutional HCP masking policies and identify epidemiologic and organizational factors informing masking decisions across primarily US healthcare organizations, with limited international participation (five responses from Saudi Arabia and one response from each of Lebanon and Qatar). Two independently administered but substantively identical surveys were distributed through the Association for Professionals in Infection Control and Epidemiology (APIC) and the Society for Healthcare Epidemiology of America (SHEA) Research Networks (ARN/SRN).
Setting and participants
Eligible participants included members of the SRN and/or ARN responsible for IPC programs and/or institutional policy development (eg, hospital epidemiologists, infection preventionists, healthcare administrators). Participating institutions included adult and pediatric acute-care hospitals, academic and non-academic facilities, behavioral health hospitals, rehabilitation hospitals, and long-term care facilities. The healthcare institution was the unit of analysis. Although individuals completed the survey, responses were intended to reflect institutional-level masking policies and decision-making frameworks. Respondents were instructed to submit a single response per institution.
Survey instrument
A structured, web-based survey was developed using REDCap (Supplemental Figure 1). Branching logic was used to tailor questions based on prior responses. Survey domains included institutional characteristics; HCP masking policies/strategies; and epidemiologic metrics and organizational/operational factors informing masking decisions. Responding organizations selected all states, territories, or regions in which they operated hospitals/facilities; these were aggregated into geographic regions (Table 1). Organizations could contribute to multiple regions. The survey was pilot tested by IPC experts within the ARN/SRN to assess clarity and content validity.
Demographic characteristics of responding healthcare organizations and survey responses1,2
1Respondents were instructed to complete the survey once per institution to reflect organizational-level policies.
2Due to branching logic restricted some questions to subsets of respondents, denominators vary by item and may be less than the total survey sample (n = 172); these are reported where applicable.
3Percentages may not sum to 100% due to rounding.
4Professional organization membership reflects respondent affiliation and is not mutually exclusive.
5For “Types of hospital(s) in organization,” “Type of mask required,” “Factors influencing the decision to implement masking,” and “Respiratory pathogens/conditions considered,” respondents could select more than one option; percentages therefore exceed 100%.
6Geographic region reflects all states or regions in which responding organizations reported operating hospitals or facilities; organizations could contribute to more than one region. See Supplemental Figure 2 for the state/territory break down of hospitals/facilities reported by respondents.
7Masking strategies reflect institutional policies outside of transmission-based precautions (eg, outbreak-specific isolation practices).
8For subsequent questions/items related to seasonal/situational masking implementation, responses were restricted to organizations reporting use of seasonal, situational, or combined masking approaches.
Abbreviations: APIC, Association for Professionals in Infection Control and Epidemiology; CDC, Centers for Disease Control and Prevention; ED, emergency department; HCP, healthcare personnel; ILI, influenza-like illness; RSV, respiratory syncytial virus; SHEA, Society for Healthcare Epidemiology of America; WHO, World Health Organization.
Survey distribution, recruitment, and data collection
Survey recruitment was conducted via e-mail outreach coordinated by the SRN/ARN. Invitation e-mails included a brief study description, survey link or QR code, estimated completion time, and a statement indicating voluntary participation. Data was collected over six weeks (SRN survey: August 21–October 2, 2025; ARN survey: September 26–November 7, 2025). Responses were anonymous. SRN facility identifiers were collected solely to ensure one response per institution.
Statistical analysis
Descriptive statistics were used to summarize organizational characteristics and masking practices. Categorical variables were reported as counts and percentages. For checkbox (“select all that apply”) items, each option was analyzed as a separate binary variable (selected vs not selected), with percentages reflecting the proportion of institutions endorsing each option. Reference Jann12
Exploratory inferential analyses were conducted to assess whether broad patterns in masking strategies and decision-making factors differed by organization type, hospital size, and geographic region. Given the descriptive aims of the study and anticipated heterogeneity in institutional practices, these analyses were not intended to estimate effect sizes, establish directionality, or support causal inference. Masking strategy (a mutually exclusive categorical variable) and individual decision-making factors (treated as binary indicators) were compared across organization types and bed-count categories (<200, 200–400, 400–800, and >800 beds) using χ 2 tests of independence (Supplemental Tables 2–4). Because sparse cell counts were anticipated, p-values were estimated using Monte Carlo simulation with 5,000 replicates to improve validity when asymptotic χ 2 assumptions may be violated. Reference Kim and Agresti13,Reference Agresti14
χ 2 tests provided global assessments of association across groups and do not convey the magnitude or direction of differences, adjust for confounding, or model multivariate relationships. Monte Carlo–simulated p-values improve inference under sparse data conditions but do not address these broader limitations. No adjustments were made for multiple comparisons, as all inferential analyses were exploratory. Reference Rothman15–Reference Althouse17
All data cleaning, recoding, statistical analyses, and table generation were conducted in R (version 4.3.2) using the dplyr, tidyr, purrr, gt, flextable, and officer packages. 18
Results
Respondents from 172 unique institutions completed the survey, representing a response rate of 41% (172/425), including 51 responses from the SRN (54% of 95 organizations) and 121 from the ARN (37% of 330 organizations) (Table 1). Most respondents were infection preventionists (65%, n = 111) or hospital epidemiologists (25%, n = 43). Academic/teaching institutions comprised the largest organizational type (42%, n = 73), followed by non-profit private (33%, n = 57) and public/government institutions (12%, n = 21). Most organizations operated four or more hospitals (54%, n = 93), 46% (n = 79) reported more than 800 beds, and most set masking policies at the system level (63%, n = 108).
Masking approaches and influencing factors
The most common masking strategy was a seasonal/situational, risk-based approach (57%, n = 98). Other strategies included masking only for certain patient groups (12%, n = 21) or combined approaches (11%, n = 19), while 7% (n = 12) of institutions reported no formal masking policy (Table 1). When masking was required, surgical masks were the most commonly required type of mask (98%, n = 146/149).
Among organizations using seasonal/situational masking strategies, the most frequently cited factors informing decisions were the presence of outbreaks/clusters of respiratory viral infections among patients or HCP (37%, n = 63), public health guidance (33%, n = 56), and HCP illness/absenteeism (27%, n = 47). Other major factors were ED visit and hospital admission trends for specific respiratory pathogens and influenza-like illnesses (ILI), each cited by 24% of responding organizations (n = 41). When ED and hospital admission data were used, the primary pathogens monitored were SARS-CoV-2 (45%, n = 78), influenza (44%, n = 76), and RSV (30%, n = 51). Among 117 organizations reporting seasonal/situational or combined masking approaches, most (81%, n = 95) considered multiple factors when making masking decisions, with a median of 5 factors (interquartile range [IQR]: 2–6; range: 1–15); 19% (n = 22) relied on a single factor.
Free-text responses: epidemiologic thresholds to inform masking decisions
Threshold data were derived from free-text responses entered in response to structured prompts within the survey (Supplemental Figure 1). Substantial heterogeneity was observed in the metrics used and thresholds applied among organizations using ED visit and/or hospital admission data to inform masking decisions. Most respondents did not report fixed quantitative thresholds, instead describing reliance on contextual, seasonal, or year-to-year trends in respiratory illness activity.
When quantitative thresholds were provided, they varied widely and most commonly involved the proportion of ED visits or hospital admissions attributable to ILI or specific respiratory pathogens. Among institutions reporting ED-based criteria, the median masking trigger was 15% of ED visits due to either ILI or specific respiratory pathogens (interquartile range [IQR]: 10–25%; range: 7–50%). Similar heterogeneity was observed for hospital admission–based thresholds (median: 15%; IQR: 10–25%; range: 2–50%).
Free-text responses also revealed substantial variation in how thresholds were conceptualized. Some organizations described tiered or escalation-based frameworks, in which masking requirements intensified as respiratory illness burden increased, while others relied on relative changes from institutional or local/regional baselines. Additional approaches included the use of absolute counts or rates and pathogen-specific thresholds for COVID-19, influenza, or RSV. Several institutions reported multi-metric, trend-based decision-making that integrated ED and admission data with test positivity, wastewater surveillance, HCP absenteeism, staffing capacity, or external public health benchmarks.
Among organizations considering HCP metrics to inform masking decisions, variability was observed in the measures used and the thresholds applied. Common factors included HCP absenteeism, sick calls, respiratory illness–related call-outs, and staff clusters/outbreaks. Although some institutions reported quantitative absenteeism thresholds, the majority had no set thresholds and relied on monitoring trends, occupational health review, or leadership discretion rather than fixed criteria. Vaccination status, staff symptom status, and confirmed respiratory infections were also frequently considered.
Wastewater surveillance was used less commonly and primarily as a supplementary, trend-based signal. When employed, monitoring focused on SARS-CoV-2, with some organizations incorporating influenza and RSV.
Masking strategies and factors by organization size
Masking strategies varied by hospital size, though differences were not statistically significant (Supplemental Table 1). A seasonal/situational, risk-based approach was the most common strategy across all hospital size categories, ranging from 50% (n = 19) in 400–800 bed hospitals to 62% (n = 24) in hospitals with <200 beds. Masking limited to specific groups (eg, symptomatic or unvaccinated staff) was most frequently reported by 200–399 bed hospitals (19%, n = 3).
There was wide variability in factors influencing masking decisions by organization size; however, none reached statistical significance (Supplemental Table 2). Consideration of HCP illness/absenteeism was most common among 400–800 bed hospitals (40%, n = 15) and least common among 200–399 bed hospitals (13%, n = 2). Reliance on local/regional public health guidance was more frequent in hospitals with >800 beds (38%, n = 30) and least common in 200–399 bed facilities (13%, n = 5). Hospital administration recommendations were cited most often by 400–800 bed hospitals (21%, n = 8) and least by 200–399 bed hospitals (6%, n = 5).
Masking strategies and factors by organization type
Descriptive analysis demonstrated variation in masking strategies by organization type, although differences were not statistically significant (Supplemental Table 3). Seasonal/situational, risk-based masking was the most common approach across all types, ranging from 44% (n = 4) in academic/teaching hospitals to 63% (n = 36) of non-profit private hospitals. Requiring masking only for staff in clinical or direct patient care areas regardless of seasonal/situational factors was uncommon, reported by 2% (n = 1) of non-profit private hospitals and none of the public/government, academic/teaching, or “other” organization types. Masking limited to specific areas/departments was most frequently reported by academic/teaching hospitals (33%, n = 3) and least frequently by non-profit private hospitals (2%, n = 1). Having no formal masking policy was most reported by “other” types of organizations (17%, n = 2) and for-profit private hospitals (7%, n = 5).
Several factors influencing masking decisions differed significantly by organizational type (Supplemental Table 4). Use of local/regional public health guidance was reported by most public/government hospitals (57%, n = 12) and “other” organizational types (50%, n = 6) but was uncommon among for-profit private hospitals (11%, n = 1) (P = .03). Public/government hospitals were also more likely to cite using WHO guidance (19%, n = 4) (P = .05) and hospital administration recommendations (27%, n = 6) (P = .04). Consideration of HCP illness/absenteeism was commonly cited by “other” organizational types (58%, n = 7) and public/government hospitals (38%, n = 8) (P = .01).
Masking strategies and decision-making by geographic region
Regional patterns in HCP masking strategies varied descriptively across geographic areas (Supplemental Tables 5 and 6). Seasonal/situational masking was most frequently reported by organizations operating in the Pacific (76%, n = 19) and Mid-Atlantic regions (71%, n = 10), and least frequently among organizations in the Southwest (22%, n = 2). The absence of a formal masking policy was reported most often by organizations in the Midwest (12%, n = 7), Southeast (10%, n = 3), and Northeast (6%, n = 2) and was not reported by organizations with hospitals or facilities in the Mid-Atlantic, Mountain West, Pacific, or international/other regions.
Similarly, masking-related decision-making factors differed descriptively by region (Supplemental Table 5). Reliance on local, state, provincial, territorial, or regional public health guidance was most commonly reported by organizations in the Pacific (52%, n = 13), Mid-Atlantic (43%, n = 6), and international/other regions (44%, n = 4), with lower frequencies in other regions. Hospital administration recommendations were cited most often by international/other organizations (56%, n = 5). Use of a fixed annual date to initiate masking was reported more frequently by organizations in the Pacific (20%, n = 5) and Mountain West (19%, n = 3) than in other regions. Because responding organizations could report operations in multiple regions, regional comparisons are presented descriptively and should be interpreted as patterns rather than independent, mutually exclusive group differences.
Discussion
We found substantial heterogeneity in masking policies for HCP and in epidemiologic and operational criteria used to inform masking decisions across healthcare institutions. Although most institutions reported using a seasonal or situational, risk-based approach to masking, there was wide variability in which metrics were considered and whether formal thresholds existed. Most responding organizations did not report using fixed or standardized thresholds to guide masking implementation or discontinuation, instead relying on contextual judgment, trend interpretation, or leadership discretion. These findings underscore the absence of a shared, operational framework to guide masking decisions in healthcare settings, resulting in widely divergent approaches to masking across and within healthcare institutions.
Masking remains a foundational IPC intervention for mitigating transmission of respiratory pathogens in healthcare settings. A robust body of evidence demonstrates masking reduces exposure to respiratory droplets and aerosols, decreases HCP infection risk, and functions effectively as source control to protect patients and HCP. Reference Offeddu, Yung, Low and Tam1,Reference MacIntyre, Seale and Dung19–Reference Chu, Akl and Duda21 Observational studies demonstrate that universal masking has been associated with reductions in HARVIs, including infections caused by influenza, SARS-CoV-2, and other pathogens. Reference Klompas, Morris, Sinclair, Pearson and Shenoy22–Reference Yan, McClure and Aslam24 Recent studies suggest discontinuation of universal masking may be associated with increased HARVIs, reinforcing the continued importance of masking in healthcare settings beyond emergency pandemic response. Reference Most, Phillips and Sebert5,Reference Pak, Chen, Kanjilal, McKenna, Rhee and Klompas6
Despite this evidence base, our findings demonstrate masking policies are currently operationalized through highly variable and often informal decision-making processes. Institutions reported using a diverse array of indicators, including outbreaks/clusters, HCP absenteeism, ED visits, hospital admissions, wastewater surveillance, and public health guidance, frequently without predefined thresholds. Even among institutions relying on similar data sources, thresholds were either absent or ranged widely. This wide variability may contribute to inconsistent implementation, challenges in communicating the rationale for masking decisions to HCP and patients, and potential inequities in protection for patients and staff across healthcare systems.
The lack of consensus guidelines defining operational criteria for masking implementation and discontinuation represents a critical gap in current IPC policy. Existing guidance from agencies such as the CDC, WHO, and state health departments appropriately emphasizes transmission-based precautions and PPE but does not provide standardized, scalable thresholds for routine masking outside of outbreak settings. 25–Reference Landelle, Birgand and Price27 As a result, healthcare organizations are left to independently interpret evolving epidemiologic data, often balancing competing considerations such as workforce well-being, patient experience, communication challenges, and staffing constraints. Prior studies demonstrating communication and interpersonal challenges associated with masking further emphasize the importance of transparent, predictable, and evidence-informed policies. Reference Mheidly, Fares, Zalzale and Fares7,Reference Lee, Cormier and Sharma8
Our findings suggest a clear opportunity for professional societies, public health agencies, and healthcare epidemiology experts to collaborate in developing consensus-based, data-driven frameworks to inform masking policies for HCP. Importantly, such frameworks should not be viewed as prescriptive algorithms or rigid triggers, but rather as decision-support tools that can inform multidisciplinary discussions among infection prevention teams, clinical leaders, and senior hospital leadership. A consensus approach could identify a core set of commonly used metrics, such as respiratory virus–associated ED visits, hospital admissions, test positivity, HCP absenteeism, and outbreak activity, paired with suggested threshold ranges or tiered escalation models, while preserving flexibility to account for institutional mission, patient populations, and local contexts. Reference Chow, Lee and Lenahan10 These frameworks should be flexible and adaptable enough to accommodate local contexts and patient populations while promoting greater consistency, transparency, and equity across healthcare settings. This adaptability is particularly relevant for centers caring for highly immunocompromised populations, where masking decisions may be appropriately implemented earlier or more broadly despite similar community epidemiology. From a policy perspective, shared, flexible frameworks may enhance transparency, promote consistency without uniformity, reduce confusion and policy fatigue, and facilitate clearer communication among leadership, HCP, patients, and the public regarding the rationale for masking decisions.
This study has several limitations. Participation was voluntary and limited to organizations engaged in SHEA and APIC research networks, which may overrepresent institutions with greater IPC resources and/or engagement. Although the institutional sample size was sufficient to demonstrate substantial heterogeneity in masking policies, subgroup and stratified analyses were constrained by modest sample sizes within individual categories, limiting statistical power. As a result, non-significant findings may reflect limited power rather than true similarity across organization types, hospital sizes, or geographic regions. Inferential analyses relied on global χ2 tests of independence, which do not provide information regarding effect size, directionality, or multivariate relationships. These analyses were exploratory and intended to support interpretation of descriptive findings rather than to generate definitive comparative conclusions. In addition, masking policies and decision-making frameworks were self-reported and may not fully capture informal practices, rapid policy changes, or nuanced operational decisions during periods of fluctuating respiratory virus activity. While respondents were instructed to submit one response per institution and facility identifiers were used within the SRN survey to identify duplicate responses, residual duplication cannot be fully excluded. Despite these limitations, the wide geographic representation, diversity of healthcare settings, and consistent patterns of heterogeneity observed across respondents support the relevance of these findings.
Conclusions
Healthcare institutions employ heterogeneous and frequently non-standardized approaches to HCP masking decisions, despite strong evidence supporting masking as an effective IPC strategy. The absence of consensus, operational guidance results in decentralized and variable policies which may limit the effectiveness and acceptability of masking interventions. Development of standardized, evidence-based, and adaptable consensus guidance is a critical next step to support coherent, transparent, and equitable masking policies protecting patients and HCP in the evolving postpandemic landscape.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/ice.2026.10469.
Acknowledgments
The authors thank the members of the Association for Professionals in Infection Control and Epidemiology (APIC) and the Society for Healthcare Epidemiology of America (SHEA) Research Networks for their time, expertise, and participation in completing this survey.
Financial support
None reported.
Competing interests
All authors report no conflicts of interest relevant to this article.
Previous presentation of the data/findings
None/Not Applicable.
Open access
