Firearm-related deaths remain a pressing public health concern in the United States. In 2021, (1) firearm homicides reached their highest level in approximately 3 decades. Although firearm homicides have since decreased nationwide, provisional estimates of 2023 deaths remain elevated relative to pre-pandemic levels. Concerningly, firearm suicides have continued to increase each year and currently are at their highest level in over 50 years (1–4). Further, firearm injuries are the leading cause of death of children and teens in the U.S. (5). While the reasons for these long-term trends are complex, the large number of firearm injuries has prompted growing attention to the problem.

The infrastructure for firearm injury data in the U.S. remains limited, especially for nonfatal injuries, constraining our understanding of temporal patterns (6), especially as most healthcare-based data sources on firearm injuries lack granular temporal information (7). For example, neither the Agency for Healthcare Research and Quality’s (AHRQ) Healthcare Cost and Utilization Project (HCUP) administrative claims-based datasets (8) nor the National Electronic Injury Surveillance System (NEISS) (9) release information on the precise day or time of injury. While this helps prevent the disclosure of protected information at the individual level, it precludes researchers from understanding when injuries are most likely to occur.

Prior research on the temporal dynamics of firearm injury and other violence outcomes has observed variation according to time of day (10–12), day of week (10,13–17), and holidays (10,13), with higher rates of firearm injury occurring generally at night, on weekends, and on holidays. Other research in criminology has highlighted similar trends across multiple types of violent incidents and crime (18–20). However, definitions of the temporal variables (e.g., “night”) have varied due to the lack of temporal granularity, limiting comparability across studies. These studies also have limitations in representativeness and timeliness. For example, several of these studies focused primarily on urban areas (10,11,15–17), perhaps because these areas may have greater public availability of firearm injury data provided by law enforcement or media sources. These results, however, may not be generalizable to rural areas (21). Other studies utilizing healthcare-based data sources are limited by substantial lags in data availability and the potential for bias inherent in sampling-based methodologies (6,13,14). Emergency medical services data excludes patients who arrive to a hospital by other means, which would result in undercounting and may introduce other systematic biases (22).

In this study, we analyzed granular temporal patterns of emergency department (ED) visits for firearm injuries using data from a novel and large-scale syndromic surveillance data source, covering both urban and rural areas, to improve resource allocation. Specifically, we utilized data on the date and time of firearm-related ED visits in 30-minute intervals to understand patterns of firearm injury and interactions between the precise time of day, day of week, month, and holiday. This information expands upon previous research (19) by including detailed data on arrival time as well as incorporating both urban and rural areas. The timeliness of syndromic surveillance data combined with the detail they contain on the timing of patient encounters may be valuable for informing resource allocation (e.g., healthcare facility staffing), prevention, and response efforts.

From January 1, 2018, to August 31, 2023, a total of 93,022 firearm injury ED visits were identified out of a total of 125,862,881 ED visits overall. The overall rate of firearm injury ED visits was 73.9 per 100,000 ED visits (95% CI, 73.9 to 73.9).

Our findings highlight significant temporal clustering of firearm injury ED visits, emphasizing the importance of evidence-based resource allocation and the need for targeted interventions during peak times. Specifically, firearm injury ED visit rates were highest during evenings, weekends, summer months, and holidays, with important interactions between temporal patterns (e.g., time of day and holiday status).

The current study is the largest analysis of temporal patterns in firearm injury using ED data to date, and the first to utilize a data source that is simultaneously timely, granular, and inclusive of data from both urban and rural areas. These findings support and expand on previous research demonstrating differences in firearm injury incidence according to time of day (10–12), day of week (10,14,15), holiday status (10,13,15), and time of year (10). Because the FASTER program captures data from throughout each included state, this study is unique in its inclusion of areas other than major metropolitan centers. Rural areas of the U.S. have been relatively understudied in recent work (10); although the current analysis did not stratify by urbanicity, the inclusion of data from rural areas improves generalizability and suggests opportunities for future research.

These insights can inform healthcare staffing and emergency preparedness, potentially reducing mortality rates associated with firearm injuries. Identifying high-burden times for when firearm injuries present in ED settings is important for informing prevention and response efforts, particularly in the form of resource allocation in treatment settings, as delays in care are associated with increased mortality rates following severe gunshot wounds (32). Since time delay between an injury and accessing advanced trauma care is associated with increased mortality (22,32), our findings provide detailed information that can help inform staffing for trauma centers and the surrounding infrastructure (e.g., prehospital emergency services). Furthermore, since a large proportion of firearm injuries are assaults (33), services such as Community Violence Intervention (CVI) or hospital-based violence intervention programs (HVIP), which provide rapid intervention by a violence prevention specialist, could help de-escalate conflicts and prevent retaliatory violence that may occur in a community in a subset of the injuries studied here (34). The implementation of these and other CVI programs could also benefit from the detailed understanding of temporal patterns provided by this study by allocating prevention and response resources when they are most likely to be needed (35).

Other research has also suggested the role of climate (36–38) and weather (15–17) in firearm injury specifically and interpersonal violence in general, as a modifier of underlying temporal patterns. Higher rates of firearm injury have been observed in association with higher daily temperatures and long-term warming trends (15,37), and lower rates with precipitation (16,17). Increases in interpersonal violence have also been observed in association with deviations from historical climate patterns (37). Together, the interplay of local factors such as weather, urbanicity, variations in holiday observances, and potentially other unmeasured confounders highlights the need for further research using data that offer both temporal and spatial granularity. Furthermore, the time period examined included the onset of the COVID-19 pandemic. Previous research using both ED and EMS data (39,40) has documented significant increases in firearm injuries during this period. Additional research is needed to fully understand how the temporal patterns described here may shift over time.

This study has several limitations. First, data are limited to FASTER-funded jurisdictions, including 9 states and the District of Columbia, and are not nationally representative. Second, the syndrome definition used in this analysis might underestimate or overestimate ED visits related to firearm injuries because of variation in coding, reporting, and availability of visit-level data between facilities or over time. The syndrome definition also does not distinguish firearm injury intent, making it impossible to know the proportions of visits related to assaults, unintentional injuries, and self-directed injuries. Further research should explore intent-specific patterns of firearm injuries to refine prevention strategies. Third, temporal analysis of trends in ED visits for firearm injury offer only an approximation of temporal trends in the injuries themselves, as the time between injury and arrival to an ED may vary and could be influenced by unmeasured confounders, including community conditions (41–43). Fourth, the period examined in the current study (approximately 5 years, 8 months) results in a relatively small number of data points for holidays (e.g., 6 instances of July 4^th) compared to non-holidays. This may distort patterns observed for individual holidays as well as comparisons among holidays. Fifth, some individual patient-level and facility-level information were incomplete or not available (e.g., discharge disposition) and thus could not be analyzed for this study. Additionally, FASTER data available to CDC does not presently contain identifiers denoting multi-victim incidents, a particular type of incident that merits additional research. Finally, temporal patterns in healthcare utilization for other conditions could potentially distort the measurement of firearm injury rates as a proportion of total ED visits. Sensitivity analysis using raw counts of ED visits for firearm injury yielded largely equivalent results, and examining data as rates still can help inform the relative need for certain types of clinical services among all ED care being sought.

Understanding the factors contributing to the temporal patterns of firearm injury presents a valuable opportunity for future prevention efforts, and implementation of policies, programs, and practices grounded in the best available evidence can bolster states’ and communities’ prevention efforts. For example, some prevention programs account for temporal variation in demand for services, and this data could potentially be used to support such programs by identifying those at highest risk during high-risk periods, monitoring progress or evaluating program effectiveness, or facilitating information-sharing among programs and partners. CDC’s Community Violence Prevention Resource for Action (44) highlights a number of strategies and approaches for preventing community violence, including firearm injury. Strategies like creating protective environments involve approaches such as safe and secure storage of firearms and remediating vacant lots to reduce opportunities for violence. Strengthening youth and young adults’ skills through school-based skill building and job training and employment programs can reinforce positive interpersonal, emotional, and behavioral skills from a young age and promote healthy relationships and economic stability. Further, strategies to intervene and lessen future harms include approaches like hospital-based violence intervention programs and street outreach with community norms change. Specifically, implementing and staffing CVI or HVIP programs, particularly during high-risk periods such as those elucidated in this study, are concrete actions that hospitals can take that could help reduce rising firearm injury and mortality rates.