Prevalence and hospital charges from firearm injuries treated in US emergency departments from 2006 to 2016

Abstract

Background: Age-and intent-related differences in the burden and costs of firearm injury treated in emergency departments are not well-documented. Methods: We performed a serial cross-sectional study of the Healthcare Cost and Utilization Program Nationwide Emergency Department Survey from 2006 to 2016. We used International Classification of Diseases diagnoses codes revisions 9 and 10 to identify firearm injuries. We calculated survey-weighted counts, proportions, means, and rates and confidence intervals of national, age-specific (0e4, 5e9, 10e14, 15e17, 18e44, 45e64, 65e84, >84) and intent-specific (assault, unintentional, suicide, undetermined) emergency department discharges for firearm injuries. We used survey-weighted regression to assess temporal trends. Results: There was a total of 868,483 (25.5 per 100,000) emergency department visits for firearm injuries from 2006 to 2016, and 7.8% died in the emergency department. Overall, firearm injury rates remained steady (P ¼ .78). The largest burden was among those 25 to 44 years of age, but their rates remained stable (10.8 per 100,000). Overall assault injuries declined from 39.7% to 36.4%, and overall unintentional injuries increased from 46.4% to 54.7%. Legal-intervention injuries declined from 0.6 to 0.3 per 100,000.

Pediatric firearm injury related emergency department visits and hospitalizations: a population-based study in the United States

“…The medical and social impact of FI on children and adolescents is profound, conferring long-lasting physical and mental disabilities that require intense medical interventions. 15 The burden of FI on healthcare costs, hospital admissions, and emergency department (ED) visits in the US has been reported. 12 , 14 , 15 , 16 , 17 While nationally representative data in the US pertaining to FI- related ED visits are available through 2016, 14 , 18 , 19 contemporary data for ED presentations or hospitalizations that are specific to children and adolescents are lacking.…”

“… 15 The burden of FI on healthcare costs, hospital admissions, and emergency department (ED) visits in the US has been reported. 12 , 14 , 15 , 16 , 17 While nationally representative data in the US pertaining to FI- related ED visits are available through 2016, 14 , 18 , 19 contemporary data for ED presentations or hospitalizations that are specific to children and adolescents are lacking. Recent reports pertaining to FI in adults suggest that there are marked disparities by gender, race and ethnicity, which have increased in recent years.…”

Assessment of the Accuracy of Firearm Injury Intent Coding at 3 US Hospitals

“…The findings of this cross-sectional study underscore questions raised by prior work about the accuracy of discharge data as a source of firearm injury intent, and by extension, the validity of findings from studies that have used ICD-coded discharge data to describe aspects of intent-specific firearm injuries, such as how many have occurred, the demographic distribution of the people injured, and the medical costs they engender. [15][16][17][18][19][20][21][22][23][24][25] Author Contributions: Drs Miller and Rowhani-Rahbar had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.…”

“…The findings of this cross-sectional study underscore questions raised by prior work about the accuracy of discharge data as a source of firearm injury intent, and by extension, the validity of findings from studies that have used ICD -coded discharge data to describe aspects of intent-specific firearm injuries, such as how many have occurred, the demographic distribution of the people injured, and the medical costs they engender …”

Classifying Firearm Injury Intent in Electronic Hospital Records Using Natural Language Processing

“…[2][3][4][5] As a result, peer-reviewed studies that have used International Classification of Diseases (ICD)-coded discharge data to assess firearm injuries by intent, including over a dozen published in high-impact journals over the past 6 years, inaccurately characterize the burden of and trends in intent-specific firearm injuries. [6][7][8][9][10][11][12][13][14][15][16] Natural language processing (NLP) and machine learning (ML) have been used to efficiently identify information contained in free-text narratives, including clinical notes and reports in electronic health records (EHRs). 17 Because these approaches have the potential to accurately assign firearm injury intent and thus result in more accurate surveillance of these injuries, we developed a hybrid model using NLP and ML to classify intent from free-text narratives in the EHR.…”