A firearm violence research methodologic pitfall to avoid

Firearm injuries, defined as “a gunshot wound or penetrating injury from a weapon that uses a powder charge to fire a projectile,” are a leading cause of morbidity and mortality in the United States.1, 2 Despite this, multiple barriers to firearm research exist, including minimal funding, legislation and regulations that prohibit the collection and access to important data, and open hostility toward researchers pursuing firearm violence research.3 Notwithstanding these barriers, researchers across disciplines have dedicated their efforts to develop evidence-based approaches to mitigate firearm injuries in recent decades, yet a lack of scientific publications commensurate with other injuries and diseases remains.4-6 One integral part of accurately describing firearm injuries and outcomes is ensuring accurate data. As a group of researchers dedicated to investigating firearm injuries among children and adolescents, we recently published in Academic Emergency Medicine and then self-retracted work that inadvertently included nonpowder firearm injuries in our analysis.7 This error was due to a failure to closely examine International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) codes leading to the inclusion of nonpowder injuries such as those caused by airgun, paintball, and BB guns when describing nonfatal firearm injuries. When we recognized our error and excluded nonpowder firearm injuries, our original sample decreased by approximately 45% (from 3296 children to 1821 children). Although the significant findings in the manuscript did not change, we have since revised the manuscript to only include firearm injuries and the replacement version was published in Academic Emergency Medicine in August 2021.7 ICD is a classification system developed by the World Health Organization (WHO) ICD-10-CM (Clinical Modification) is an adaption of the WHO's ICD with additional details added to the classification system for use in the United States. The National Center for Health Statistics maintains the U.S. ICD version. There is a set of ICD-10-CM codes which are frequently utilized by firearm injury researchers in publications for nonfatal firearm injuries.9-11 These codes are often provided in the methods section as ICD-10-CM categories which must be expanded to their full characters to examine individual codes. For example, they are often described as the following categories: W32–W34, X72–X74, X93–X95, and Y22–Y24. What partially caused our error, and is a potential pitfall of this strategy, however, is that when examining nonfatal firearm injuries, nonpowder injuries (e.g., airgun or paintball injuries) are embedded within these commonly described categories of codes. The majority of studies, unless otherwise specified, mean to represent handguns and long guns, which use gunpowder as the propellant, and not airgun or paintball injuries, which are not associated with fatal injuries, are almost always unintentional, and are a different focus of public policy efforts.12 Unless researchers are diligent in their case selection when investigating nonfatal firearm injuries and explicitly describe it in their methods, nonpowder injuries could be included and erroneously reported as the type of firearm injuries that the researchers hope to describe. Table 1 includes a detailed listing of ICD-10-CM codes associated with firearms with the nonpowder injury codes embedded within these codes bolded and highlighted. A detailed review of ICD codes is also necessary when reporting mortality data as well. Some ICD codes, including W34, X74, X95, and Y24, all include “airguns, BB guns and flare guns” as shown in Table 2. Researchers should consider exclusion of these specific codes because they include nonpowder injuries. The potential inclusion of nonpowder gun mechanisms in research related to both fatal and nonfatal firearm injuries has clear implications for understanding the epidemiology of firearm injuries and limit our ability to improve prevention and treatment. Therefore, it is our responsibility as firearm injury researchers to ensure precision when examining and publishing these data. Consistency across definitions and methodology in firearm research would allow for more accurate comparisons of results across publications and offer more credibility toward creating evidence-based policy. Fortunately, the CDC Web-based Injury Statistics Query and Reporting System (WISQARS) database, which provides surveillance of firearm injuries in the United States, does not include nonpowder injuries in their estimates of nonfatal firearm injuries. Firearm violence is a public health epidemic. It is of utmost importance that firearm injury researchers ensure precision of data given the implications of simultaneous inflation and dilution of the data with common and typically minor injuries caused by nonpowder firearms. Accurate information is essential to inform sound public policy and researchers should continue their diligence in the description of their data sample when utilizing code methodology. All authors have no conflicts of interest to disclose.