Economic Cost of Injury — United States, 2019

Unintentional and violence-related injuries, including suicide, homicide, overdoses, motor vehicle crashes, and falls, were among the top 10 causes of death for all age groups in the United States and caused nearly 27 million nonfatal emergency department (ED) visits in 2019.*,† CDC estimated the economic cost of injuries that occurred in 2019 by assigning costs for medical care, work loss, value of statistical life, and quality of life losses to injury records from the CDC's Web-based Injury Statistics Query and Reporting System (WISQARS).§ In 2019, the economic cost of injury was $4.2 trillion, including $327 billion in medical care, $69 billion in work loss, and $3.8 trillion in value of statistical life and quality of life losses. More than one half of this cost ($2.4 trillion) was among working-aged adults (aged 25-64 years). Individual persons, families, organizations, communities, and policymakers can use targeted proven strategies to prevent injuries and violence. Resources for best practices for preventing injuries and violence are available online from CDC's National Center for Injury Prevention and Control.¶.

comorbidities, sex, and age, and modified to 2019 USD.** Medical costs were assigned to WISQARS records by injury outcome (fatal or nonfatal), mechanism (e.g., fall), intent (e.g., unintentional), and place of death (e.g., inpatient hospital) or ED visit disposition (treated and released or hospitalized, including transferred). Work loss costs for nonfatal injuries were assigned by injury mechanism and ED visit disposition ** U.S. Bureau of Economic Analysis. National Income and Product Accounts: to injured persons of all ages; this approach assumes injured children and older adults incur lost productivity among working-aged adult caregivers. Aggregated medical and work loss costs (e.g., combined intents by mechanism or combined mechanisms by ED visit disposition) from reference sources were assigned when specific estimates by intent or mechanism were not available. The cost of injury mortality includes value of statistical life, a monetary estimate of the collective value placed on mortality risk reduction as derived in research studies through revealed preferences (e.g., observed wage differences for dangerous occupations) or stated preferences from surveys of individual persons' willingness to pay for mortality risk reduction (3). Value of statistical life estimates were assigned by decedent age: 0-17 years, $16.9 million (4); 18-65 years, $10.7 million (3); and values descending from $6 million (aged 66 years) to $410,000 (aged ≥100 years), reflecting the estimate for persons aged 18-65 years adjusted for older adults' decreasing general life expectancy and baseline quality of life. Cost of nonfatal injury morbidity includes quality of life losses measured in terms of quality-adjusted life years (QALY; 1 QALY equals 1 year of perfect health) (5) and valued at $540,000 per QALY (3). Injury count, rate per 100,000 population, cost by type (medical, work loss, value of statistical life, and quality of life loss), and total cost are reported by intent, sex, and age group. All reported data can be queried online using WISQARS. This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy. † † In 2019, the economic cost of injury was $4.2 trillion, including $327 billion in medical care, $69 billion in work loss, and $3.8 trillion in value of statistical life and quality of life losses (Table). The economic costs were $2.2 trillion for fatal injuries and $2.0 trillion for nonfatal injuries. The number of injury deaths and associated economic cost were higher among males (169,628 and $1.6 trillion, respectively) than among females (76,413 and $607 billion, respectively). The cost of nonfatal injury was similar for males and females ($1 trillion). Except for nonfatal self-harm, the age-adjusted rate, number, and economic cost for all injury outcomes (fatal and nonfatal) and intents (unintentional, homicide or assault) were higher for males than for females.

Discussion
This report used injury incidence data to estimate the economic cost of injuries that occurred in the United States during 2019. Economic cost was highest among working-aged adults, highlighting that injuries during the most productive part of people's lives result in a high societal cost. These findings highlight the need for targeted prevention strategies to achieve long-term value, or even cost-savings, by preventing injury morbidity and mortality through addressing the causes of unintentional and violence-related injuries at the individual, family, organizational, and community levels.
The 2019 economic cost of injuries ($4.2 trillion) is more than six times as high as a comparable estimate in 2013 ($671 billion) (6,7). Even though the number of nonfatal ED injury visits in 2019 was approximately 15% lower than it was in 2013, the 2019 nonfatal injury economic cost ($2.0 trillion) is more than four times as high as the 2013 estimate ($457 billion) (6), primarily because of including the cost of diminished quality of life. The 2019 fatal injury economic cost ($2.2 trillion) is substantially higher than the similar estimate in 2013 ($214 billion) (7). This difference reflects a 28% higher number of injury deaths in 2019 and mortality cost based on value of statistical life, which represents a value that is approximately 10 times as high as the value attributed to mortality based on foregone employment compensation, which was used in the previous estimate.
The findings in this report are subject to at least five limitations. First, the economic cost of nonfatal injuries is underestimated because only injuries treated in an ED are included (injuries initially treated in urgent care or doctor's offices not included), Abbreviation: USD = U.S. dollars. * In millions of 2019 USD. † Fatal all intents estimates include injuries with legal intervention intent, undetermined intent, unknown sex, and unknown age. § Per 100,000. Age-adjusted rate is presented for "Total, " "Male, " and "Female" columns. ¶ Nonfatal injuries are an estimated number of hospital visits for injury care that start in an emergency department (with disposition treated and released, transferred, or hospitalized; visits with observed, left against medical advice, and unknown disposition were not included) based on a nationally representative probability sample of hospitals. ** Nonfatal all intents estimates include injuries with legal intervention intent, unknown sex, and unknown age. Nonfatal assault, self-harm, and legal intervention include cases that are confirmed or suspected; all other cases are considered unintentional.
other costs such as property damage and criminal justice are not included, and nonfatal costs address only the first year following an injury. The cost of nonfatal injury includes observed medical care and work loss attributable to injuries based on comparing injured patients and non-injured persons during the year following the injured patient's initial ED visit (1,2). A 1-year time horizon is appropriate for many injury types but does not address the longterm physical and mental health consequences of some injuries (e.g., traumatic brain injury and violence-related injuries). Second, although injury-related medical care and work loss have costs to specific, identifiable payers (including individual persons, health insurance payors, and employers), the highest cost elements presented here are value of statistical life and quality of life losses; these costs are not readily identifiable through financial transactions and thus not as visible to some stakeholders as are direct costs, such as medical care. Third, although this study aimed for a reasonable use of available value of statistical life data, the relationship between value of statistical life and age (particularly, value of statistical life for older adults) is likely more complex than applied here and would benefit from further direct study (8). Fourth, quality of life loss estimates might indirectly capture some work loss; therefore, the nonfatal economic cost estimate might partially double count such costs. Finally, this report provides an initial assessment of the economic cost of injury by intent based on injured person sex and age group. Estimation of injury costs by other demographic and geographic factors within the United States can provide additional meaningful information for injury prevention. Individual persons, families, organizations, communities, and policymakers can use targeted proven strategies to prevent injuries and violence. Data and resources that can assist in measuring and preventing injuries and violence, including suicide, overdoses, falls, firearm violence, motor vehicle crashes, traumatic brain injury, adverse childhood experiences, youth violence, sexual violence, and intimate partner violence,

State-Level Economic Costs of Fatal Injuries -United States, 2019
Cora Peterson, PhD 1 ; Feijun Luo, PhD 1 ; Curtis Florence, PhD 1 Unintentional and violence-related injury fatalities, including suicide, homicide, overdoses, motor vehicle crashes, and falls, were among the 10 leading causes of death for all age groups in the United States in 2019.* There were 246,041 injury deaths in 2019 (unintentional injury was the most frequent cause of death after heart disease and cancer) with an economic cost of $2.2 trillion (1). Extending a national analysis (1), CDC examined state-level economic costs of fatal injuries based on medical care costs and the value of statistical life assigned to 2019 injury records from the CDC's Web-based Injury Statistics Query and Reporting System (WISQARS). † West Virginia had the highest per capita cost ($11,274) from fatal injury, more than twice that of New York, the state with the lowest cost ($4,538). The five areas with the highest per capita total fatal injury costs were West Virginia, New Mexico, Alaska, District of Columbia (DC), and Louisiana; costs were lowest in New York, California, Minnesota, Nebraska, and Texas. All U.S. states face substantial avoidable costs from injury deaths. Individual persons, families, organizations, communities, and policymakers can use targeted proven strategies to prevent injuries and violence. Resources for best practices for preventing injuries and violence are available online from the CDC's National Center for Injury Prevention and Control. § The economic cost estimate for injuries that occurred in 2019 uses the societal perspective, including tangible and intangible costs to multiple payers. Costs are presented in 2019 U.S. dollars (USD). WISQARS fatal injury counts are from CDC's National Vital Statistics System mortality data. In 2019, approximately 70% of U.S. injury deaths (173,040) were attributable to unintentional injuries (among which 36% were related to drug poisoning, 23% to falls, and 22% to motor vehicle traffic); approximately 20% were suicides; and 8% were homicides. ¶ Medical costs were adjusted for patient characteristics (2), including comorbidities, sex, and age, and were modified to 2019 USD** and assigned to WISQARS records by injury mechanism (e.g., fall), intent (e.g., unintentional), and place of death (e.g., inpatient hospital). Aggregated medical costs (e.g., combined intents by mechanism or combined mechanisms by place of death) from reference sources were assigned when specific estimates by intent or mechanism were not available. The average medical cost among 2019 injury deaths was approximately $15,400 † † ; however, many injury deaths had lower costs because the deaths occurred outside a health care setting (2). The cost of injury mortality includes value of statistical life, a monetary estimate of the collective value that persons place on mortality risk reduction as derived in research studies through revealed preferences (e.g., observed wage differences for dangerous occupations) or stated preferences from surveys of persons' willingness to pay for mortality risk reduction (3). Value of statistical life estimates were assigned by decedent age: 0-17 years, $16.9 million (4); 18-65 years, $10.7 million (3); values descending from $6 million (aged 66 years) to $410,000 (aged ≥100 years), reflecting the estimate for persons aged 18-65 years adjusted for older adults' decreasing general life expectancy and baseline quality of life. Per capita fatal injury costs by U.S. state are presented graphically by injury intent (Figure). Injury fatality rates by state, intent, sex, and age group (0-24 years, 25-64 years, and ≥65 years) were examined to better understand the contributing circumstances in states with the highest per capita total injury costs. All reported data can be queried online using WISQARS. This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy. § § The five areas with the highest per capita total fatal injury costs were West Virginia, New Mexico, Alaska, DC, and Louisiana; those with the lowest costs were New York, California, Minnesota, Nebraska, and Texas ( Figure). ¶ ¶ West Virginia had the highest per capita cost ($11,274) from fatal injury, more than twice that of New York, the state with the lowest cost ($4,538). The five states † † Data available at https://wisqars.cdc.gov/cost and can be accessed using the following data selections: Injury  State-level injury fatality rates in 2019 by intent, sex, and age group suggest that the five areas with the highest per capita total fatal injury cost face different challenges in injury prevention. West Virginia had the highest age-adjusted unintentional injury fatality rate in the nation (including the highest for males, females, and persons aged 25-64 years) and the second highest for persons aged ≥65 years.*** New Mexico had the *** Data available at https://wisqars.cdc.gov/data/explore-data/home using the second highest age-adjusted unintentional injury fatality rate (including the second highest for males and females and third highest for persons aged 25-64 years) and was among the five states with the highest age-adjusted rate of both suicide (including for males, females, and each of the three assessed age groups) and homicide deaths (including for females and persons aged 25-64 years). Alaska had the nation's second highest age-adjusted suicide rate (including for males, females, and persons aged 25-64 years, and the highest rate for persons aged 0-24 years) and the highest age-adjusted homicide rate for females. DC had the lowest age-adjusted suicide rate among all areas but the highest homicide rate (including for males and persons aged 0-24 and 25-64 years). Louisiana was among the top three states for the highest rate of age-adjusted homicide deaths (including for males, females, and each of the three assessed age groups).

Discussion
All U.S. states face substantial avoidable costs from injury deaths. Identifying the economic cost of injuries is an essential part of the public health approach to injury and violence prevention and can support identification of cost-effective interventions. These findings highlight the need for targeted prevention strategies to achieve long-term value, or even cost savings, by preventing injury morbidity and mortality through addressing the causes of unintentional and violence-related injuries at the individual person, family, organizational, and community levels.
The number of injury deaths and a higher proportion of younger decedents had the biggest impact on each state's total fatal injury cost. The cost of medical care for 2019 injury fatalities is marginal in comparison to the value of statistical life (1), which aims to capture complex costs related to mortality. Value of statistical life represents a value that is approximately 10 times higher than the value attributed to mortality based on foregone employment compensation, which was used to estimate states' economic cost of fatal injuries in 2014 (5). At that time, similar to the results presented in this report, Alaska, Louisiana, New Mexico, Oklahoma, and West Virginia had the highest per capita total injury costs among U.S. states.
The findings in this report are subject to at least four limitations. First, although injury-related medical care incurs costs to specific, identifiable payers (including individual persons, health insurance payors, and employers), value of statistical life aims to capture costs of mortality that are not readily identifiable through financial transactions and thus are not as visible to some stakeholders. Second, available average medical

Summary
What is already known about this topic?
In 2019, injuries accounted for 246,041 U.S. deaths; the economic cost of these injuries was $2.2 trillion.
What is added by this report?
West Virginia had the highest per capita cost ($11,274) from fatal injury, more than twice that of New York, the state with the lowest cost ($4,538). The highest per capita fatal injury costs occurred in Alaska, District of Columbia, Louisiana, New Mexico, and West Virginia; the lowest occurred in California, Minnesota, Nebraska, New York, and Texas.
What are the implications for public health practice?
All states face substantial avoidable costs due to injury deaths. Resources for best practices for preventing injuries and violence are available online from CDC's National Center for Injury Prevention and Control.
costs in reference sources were not state-specific. Third, on the basis of available data, this study assigned value of statistical life by age group; however, the relationship between value of statistical life and age (and in particular, value of statistical life for older adults) is likely more complex than applied here and would benefit from further direct study (6). Finally, this report provides an initial assessment of states' economic costs of injury by intent. Observed differences in per capita total fatal injury costs likely reflect important differences in affected populations (e.g., children, youths, and young adults versus older adults) and injury mechanism (e.g., firearm, fall, or drug poisoning) that must be understood to effectively target prevention resources.
Individual persons, families, organizations, communities, and policymakers can use targeted proven strategies to prevent injuries and violence. Data and resources that can assist in measuring and preventing injuries and violence, including suicide, overdoses, falls, firearm violence, motor vehicle crashes, traumatic brain injuries, adverse childhood experiences, youth violence, sexual violence, and intimate partner violence, are available online from CDC's National Center for Injury Prevention and Control. Opportunities to investigate national and state-level injury data and costs are available online from WISQARS.

Incidence of Nonfatal Traumatic Brain Injury-Related Hospitalizations -United States, 2018
Alexis B. Peterson, PhD 1 ; Karen E. Thomas, MPH 1 Traumatic brain injury (TBI), which can disrupt normal brain function and result in short-and long-term adverse clinical outcomes, including disability and death, is preventable. To describe the 2018 incidence of nonfatal TBI-related hospitalizations in the United States by sociodemographic characteristics, injury intent, and mechanism of injury, CDC analyzed data from the Healthcare Cost and Utilization Project (HCUP) National (Nationwide) Inpatient Sample. During 2018, there were 223,050 nonfatal TBI-related hospitalizations; rates among persons aged ≥75 years were approximately three times higher than those among persons aged 65-74 years, and the age-adjusted rate among males was approximately double that among females. Unintentional falls were the most common mechanism of injury leading to nonfatal TBI-related hospitalization, followed by motor vehicle crashes. Proper and consistent use of recommended restraints (i.e., seatbelts, car seats, and booster seats) and, particularly for persons aged ≥75 years, learning about individual fall risk from health care providers are two steps the public can take to prevent the most common injuries leading to nonfatal TBIs. The findings in this report could be used by public health officials and clinicians to identify priority areas for prevention programs.
Estimates for nonfatal TBI-related hospitalizations were obtained from the 2018 HCUP National Inpatient Sample files. The National Inpatient Sample is a stratified sample of approximately 20% of hospital discharges in the United States and is sponsored by the Agency for Healthcare Research and Quality. Records were included if the primary diagnosis was an injury and a TBI-related International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) code (S02.0, S02.1-, S02.80X-S02.82X, S02.91, S04.02, S04.03-, S04.04-, S06-, S07.1, and T74.4) was present in any diagnosis field. A record could potentially include multiple external cause of injury codes; injury mechanism/intent categories were based on the first cause code found, as it was considered the first valid external cause of injury code. ICD-10-CM codes and more detailed methods are available online (1). Rates were calculated using bridged race population estimates obtained from the National Center for Health Statistics as denominators. Nonfatal hospitalizations were weighted to provide national estimates, and 95% CIs were calculated using complex survey procedures in SAS (version 9.4; SAS Institute). Age-adjusted rates were calculated using the direct method and the 2000 U.S. Census Bureau standard population. This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy.* In 2018, there were 223,050 nonfatal TBI-related hospitalizations in the United States. Among nonfatal TBI-related hospitalizations with known age, 16,480 (7.4%) occurred among infants, children, and adolescents aged 0-17 years, and 70,445 (31.6%) occurred among adults aged ≥75 years (Table 1). National rates of nonfatal TBI-related hospitalizations were highest among persons aged ≥75 years (321.4 per 100,000 population) and among males (81.3 per 100,000 population, age-adjusted). The rate of nonfatal TBI-related hospitalizations among persons aged ≥75 years was approximately three times higher than that among those aged 65-74 years (105.5 per 100,000 population), and the rate among males was approximately double that among females (44.4 per 100,000 population, age-adjusted). Age-adjusted rates of nonfatal TBI-related hospitalizations were similar among non-Hispanic White persons (59.0 per 100,000 population), non-Hispanic Black persons (60.0 per 100,000 population), and Hispanic persons (59.6 per 100,000).
In 2018, approximately 75% of nonfatal TBI-related hospitalizations were caused by either unintentional falls (51.0%) or motor vehicle crashes (23.8%) ( Table 2). Rates for nonfatal TBI-related hospitalizations attributable to unintentional falls were highest among adults aged ≥75 years (263.3 per 100,000 population), 65-74 years (69.9 per 100,000 population), and 55-64 years (33.2 per population). Among all age groups, the highest rates of motor vehicle crashes leading to a nonfatal TBIrelated hospitalization were among persons aged 15-24 years (24.6 per 100,000 population) and aged 25-34 years (21.9 per 100,000 population). Among the major examined unintentional and intentional mechanisms of injuries that contributed to a nonfatal TBI-related hospitalization (e.g., motor vehicle crashes, falls, being struck by or against an object, self-harm, and assault), higher total estimates and age-adjusted rates were observed among males compared with females for all mechanisms of injury (Table 3).

Discussion
Nationally, 223,050 nonfatal TBI-related hospitalizations occurred during 2018. Rates varied by age group, sex, principal mechanism of injury and, within each principal mechanism of injury, by age group. Consistent with findings from a previous CDC surveillance report (1), the highest estimates and rates of nonfatal TBI-related hospitalizations occurred among adults aged ≥75 years and among males, and unintentional falls and motor vehicle crashes were the most common mechanisms of nonfatal TBI-related injury. The highest rate of nonfatal TBI-related hospitalizations was among persons aged ≥75 years, the oldest age group in this study; hospitalizations among this age group can be complicated by the presence of underlying medical conditions, including hypertension, diabetes mellitus, or coronary heart disease (2). Older age is a known major risk factor for TBI (2), and following a TBI, older adults perform worse on measured cognitive abilities (e.g., naming and vocabulary) when compared with older adults without a history of TBI (3). Consistent with previous data suggesting that males are more likely than are females in the general adult population to sustain a TBI (4), this study found higher age-adjusted rates of nonfatal TBIrelated hospitalizations among males compared with females across all mechanisms of injury. Reported incidence of TBI by sex is complex and potentially affected by several factors, including differing biologic vulnerabilities to injury and sex differences in care-seeking behavior (5).
Unintentional fall was the leading mechanism of injury contributing to a nonfatal TBI diagnosis for which the patient was hospitalized. During 2018, the highest rate of nonfatal TBIrelated hospitalization attributable to falls was among adults aged ≥75 years, consistent with older age being a major risk factor for falls (6). Health care providers should evaluate older adult patients for signs and symptoms of TBI if they have fallen or had a fall-related injury, such as a hip fracture (7). Further, more older adults receive aspirin and anticoagulant therapies (e.g., warfarin [Coumadin] and non-vitamin K oral anticoagulants) as part of routine management of chronic conditions. The prevalence of anticoagulant use in this population can result in an increased likelihood of intracranial hemorrhage (8) and further complications from TBIs. Consistent with previous epidemiologic data (1), the age-adjusted rate for nonfatal TBIrelated hospitalization attributable to falls was higher among males than among females. This finding might be related to circumstances of the fall, such as a larger proportion of males falling from heights (e.g., ladders) (9), which are more likely to result in moderate to severe injuries, including TBI. Abbreviation: TBI = traumatic brain injury. * All rates are per 100,000 population. † In-hospital deaths and patients who transferred from another hospital were excluded. § Excluded falls of undetermined intent. ¶ Injuries in persons aged <10 years were excluded because determining intent in younger children can be difficult. Rates for nonfatal TBI-related hospitalizations because of intentional self-harm were age-adjusted to the population aged ≥10 years. ** Includes undetermined intent, legal intervention, war, intentional self-harm for age <10 years, and cases without information about cause of injury.
† † Entry suppressed because of data confidentiality concerns associated with unweighted case counts ≤10. § § Hospitalizations with missing/unknown age were included. ¶ ¶ Hospitalizations with missing/unknown age were excluded. Rates were age-adjusted using the direct method to the 2000 U.S. Census Bureau standard population. Abbreviation: TBI = traumatic brain injury. * Hospitalizations with missing age were excluded. Rates were age-adjusted using the direct method to the 2000 U.S. Census Bureau standard population (per 100,000 population). † In-hospital deaths and patients who transferred from another hospital were excluded. § Falls of undetermined intent were not included. ¶ Injuries in persons aged <10 years were excluded because determining intent in younger children can be difficult. Rates for nonfatal TBI-related hospitalizations because of intentional self-harm were age-adjusted to the population aged ≥10 years. ** Includes undetermined intent, legal intervention, war, intentional self-harm for those aged <10 years, and cases without information about cause of injury.
The second most common mechanism of injury among all age groups was motor vehicle crashes, with the age-adjusted rate among males being approximately double that among females. Males are involved in more motor vehicle crash fatalities when compared with females, † and data from one state suggest that this finding might be the result of a higher incidence of speeding and loss-of-control crashes among males (10). The likelihood of nonfatal TBI-related hospitalization from a motor vehicle crash can be reduced for persons of all ages, including older adults, by consistently and properly wearing a seatbelt while driving or riding in a motor vehicle, never driving under the influence of drugs or alcohol, and driving at recommended speeds. Consistently and properly buckling children into ageand weight/height-appropriate car or booster seats § can prevent pediatric TBIs caused by a motor vehicle crash. New adolescent and young adult drivers can help prevent TBIs attributed to motor vehicle crashes by engaging in graduated driving licensing systems that help build driving skills (e.g., lane merging, passing, and maintaining a safe distance while driving) and limiting driving under high-risk conditions. Motorcyclists and bicyclists can reduce the likelihood of TBI during a crash by properly and consistently wearing a helmet.
The findings in this report are subject to at least three limitations. First, persons who only sought care in the emergency department or outside the hospital setting (e.g., urgent care, primary care, and specialty care), who received a TBI diagnosis in federal, military, or Veterans Administration hospitals, or who did not seek care at all were not included. Therefore, this † https://cdan.dot.gov/query § https://www.cdc.gov/transportationsafety/child_passenger_safety/ report is not a complete accounting of all nonfatal TBIs in the United States. Second, this analysis did not differentiate nonfatal TBI cases by severity of injury. Finally, the mechanism and intent of injury were unknown for 8.4% of nonfatal TBIrelated hospitalizations, and as a result, estimates by mechanism of injury and injury intent are undercounts.
A TBI can happen to anyone at any age; during 2018, the oldest age group (aged ≥75 years) experienced the highest numbers and rates of nonfatal TBI-related hospitalizations. Among all nonfatal TBI-related hospitalizations, unintentional falls were the leading cause of injury, with half of these hospitalizations occurring among older adults, highlighting the need to intensify prevention efforts for falls, particularly among this age group. The CDC's Stopping Elderly Accidents, Deaths, and Injuries (STEADI) ¶ initiative can support health care providers in screening for fall risk, assessing modifiable risk factors, and intervening to reduce risk by updating patients' personalized fall prevention plans. Proper restraint use (i.e., seatbelts, car seats, and booster seats) is a proven strategy for reducing motor vehicle occupant injuries, including TBIs. The CDC's Motor Vehicle Prioritizing Interventions and Cost Calculator for States (MV PICCS)** can aid states in identifying strategies that could effectively reduce motor vehicle crash injuries. TBIs are preventable. The findings in this report could be used by public health officials to support identification of priority areas for TBI prevention programs and groups at increased risk for TBI. ¶ https://www.cdc.gov/steadi/ ** https://www.cdc.gov/transportationsafety/calculator/index.html

Summary
What is already known about the topic? Traumatic brain injury (TBI), an injury that can disrupt normal brain function, contributes to a substantial number of hospitalizations each year.
What is added by this report?
During 2018, there were 223,050 nonfatal TBI-related hospitalizations in the United States. Rates were highest among males and persons aged ≥75 years. Unintentional falls and motor vehicle crashes were the most common injuries leading to a nonfatal TBI-related hospitalization.
What are the implications for public health practice?
Proper and consistent restraint use (i.e., seatbelts, car seats, and booster seats) and learning about individual fall risk from health care providers are two steps the public can take to prevent the most common injuries leading to a nonfatal TBI.

Introduction
Gay, bisexual, and other men who have sex with men (MSM) have been disproportionately affected by HIV since the onset of the epidemic and have been a priority population for HIV prevention and treatment (1). Despite focused prevention efforts, approximately two thirds of new HIV infections in the United States occur in MSM (2). Advances in HIV prevention and treatment have made HIV infection increasingly preventable, but new infections have continued. Preexposure prophylaxis (PrEP) is highly effective in preventing infection, and consistent antiretroviral therapy (ART) enables persons with HIV to become virally suppressed and prevents transmission to others (3,4). By maximizing these advances, the Ending the HIV Epidemic in the U.S. (EHE) initiative aims to reduce the number of new HIV infections in the United States by 90% by 2030; 57 state and local jurisdictions began implementing the initiative in 2020.* EHE goals cannot be achieved without substantial reductions in HIV infections among MSM. CDC analyzed data from three national surveillance systems to assess HIV prevention and treatment outcomes among MSM in the United States during the years before EHE implementation and the progress needed to reach EHE and other national goals (Supplementary Box, https://stacks.cdc.gov/view/cdc/111462).

Methods
CDC assessed select outcomes related to the use of important HIV prevention services and steps in the HIV care continuum † * https://www.hiv.gov/federal-response/ending-the-hiv-epidemic/overview † https://www.cdc.gov/hiv/pdf/library/factsheets/cdc-hiv-care-continuum.pdf among MSM overall and by race/ethnicity and age group using data from the National HIV Surveillance System (NHSS), National HIV Behavioral Surveillance (NHBS), and Medical Monitoring Project (MMP).

Estimated number of new HIV infections and percentage of infections that were diagnosed.
Using NHSS data, the estimated number of new HIV infections among MSM was 25,100 in 2010 and 23,100 in 2019 (p = 0.05) (Figure). During this period, infections significantly decreased from 7,500 to 5,100 among White MSM (p<0.01) but did not decline significantly among Black or African American (Black) MSM (9,000 to 8,900; p = 0.90) and Hispanic/Latino MSM (6,800 to 7,900; p = 0.10). Infections decreased among MSM aged 13-24 years (10,400 to 5,700; p<0.01) and 45-54 years (2,900 to 2,000; p<0.01) but increased among MSM aged 25-34 years (6,700 to 10,000; p<0.01).
Among the estimated 692,900 MSM living with HIV infection in 2019, 85% had received an HIV diagnosis (Table 1). A lower percentage of Black MSM (83%) and Hispanic/ Latino MSM (80%) with HIV had received a diagnosis than did White MSM (90%). The lowest percentages of diagnosed infection were among MSM aged 13-24 years (55%) and 25-34 years (71%).  a positive HIV test >12 months before the interview. Among these 7,577 MSM, 79% were tested for HIV in the past 12 months (Table 1). Among 1,181 MSM who visited a health care provider but had not tested in the past 12 months, 78% were not offered an HIV test. Neither HIV testing in the past 12 months nor having visited a provider without testing in the past 12 months differed by race/ethnicity. MSM aged 45-54 years and ≥55 years had the lowest percentages of testing in the past 12 months; visiting providers without a test did not differ by age group. Approximately one half (52%) of HIV-negative MSM with likely PrEP indications** reported having discussed PrEP with a health care provider in the past 12 months, and approximately one third (36%) had used PrEP in the past 12 months ( MSM aged 25-34 years (65%) had the lowest percentages of viral suppression. The median HIV-related stigma score † † among MSM with diagnosed HIV infection was 29 on a scale of 0 to 100. MSM aged 18-24 years had the highest median score (39). Black MSM (33) and Hispanic/Latino MSM (32) had higher median scores than did White MSM (26).

Discussion
These findings indicate that new HIV infections among Black MSM and Hispanic/Latino MSM did not decrease during the decade before EHE implementation despite decreases or † † Participants indicated their agreement with the following statements: 1) "During the past 12 months, I have been hurt by how people reacted to learning I have HIV;" 2) "During the past 12 months, I have stopped socializing with some people because of their reactions to my HIV status;" 3) "During the past 12 months, I have lost friends by telling them I have HIV;" 4) "I am very careful who I tell that I have HIV;" 5) "I worry that people who know I have HIV will tell others;" 6) "I feel that I am not as good a person as others because I have HIV;" 7) "Having HIV makes me feel unclean;" 8) "Having HIV makes me feel that I'm a bad person;" 9) "Most people think that a person with HIV is disgusting;" and 10) "Most people with HIV are rejected when others find out." Median scores and 95% CIs were calculated on a scale of 0 (no stigma) to 100 (highest stigma). stable numbers among other MSM subgroups, and new infections increased among MSM aged 25-34 years. Use of many prevention and treatment strategies were less prevalent among Black MSM, Hispanic/Latino MSM, and younger MSM. Longstanding inequities in access to and delivery of needed services among some racial/ethnic and age groups, particularly Black MSM and Hispanic/Latino MSM, have persisted despite focused efforts to prevent HIV in these populations for decades. Efforts to reduce these and other disparities must address their root causes, including systemic racism, stigma, discrimination, homophobia, poverty, homelessness, and unequal access to care and prevention services (1).

TABLE 2. Percentage of gay, bisexual, and other men who have sex with men with a likely indication for preexposure prophylaxis who discussed preexposure prophylaxis with a health care provider in the past 12 months or used preexposure prophylaxis in the past 12 months, by race/ethnicity and age group -
Achieving the EHE goals to reduce the number of HIV infections by 90% by 2030 will require that at least 95% of infections are diagnosed and 95% of persons with diagnosed HIV infection are virally suppressed (Supplementary Box, https:// stacks.cdc.gov/view/cdc/111462); the most recent available data indicate that among MSM, only 85% of HIV infections are diagnosed and 68% of MSM with diagnosed HIV infection are virally suppressed. Approximately 20% of MSM not previously receiving a diagnosis of HIV infection had not been tested for HIV in the past year, which is inconsistent with CDC recommendations that all sexually active MSM be tested at least annually (5). Missed clinical opportunities for testing were common among MSM who had not been tested in the past year. Further, PrEP was used by only one third of MSM for whom it was likely indicated, well below the EHE target of 50% PrEP coverage (Supplementary Box, https://stacks.cdc. gov/view/cdc/111462). Median HIV-related stigma scores were nearly double the national target (6). The persistence of HIVrelated stigma might hinder access to testing, prevention, and treatment for MSM, thus potentially undermining progress toward national goals. Together, these findings suggest the need for innovative approaches that can better deliver testing, prevention, and treatment services to MSM.
Several innovative and culturally appropriate strategies have successfully reduced barriers to access of services and might help achieve national goals of improving prevention, diagnosis, and treatment of HIV infection among MSM. § § For example, HIV testing scale-up has been determined to be cost-effective across diverse local conditions (7). Some jurisdictions have successfully implemented programs that increased screening frequency among MSM (8). Numerous strategies have been implemented to deliver HIV testing services to MSM by expanding or tailoring existing clinical screening programs, enhancing community-based testing options, or providing HIV self-tests (9). HIV self-testing can be a cost-saving delivery strategy (10) with potential to mitigate HIV-related  stigma and better reach MSM (11). Multiple jurisdictions have demonstrated that HIV self-test distribution programs can successfully deliver HIV testing to racial/ethnic minority MSM and MSM not reached by other testing programs (12). CDC recently supported a national self-test distribution program designed to improve access to HIV testing for those who had not been previously reached. ¶ ¶ To improve HIV care outcomes, strategies and approaches supported by the Ryan White HIV/AIDS Program (RWHAP) can be scaled up to reach all U.S. facilities that provide HIV care. RWHAP-funded facilities deliver comprehensive care and essential support services to approximately one half of persons with diagnosed HIV infection in the United States through enhanced collaboration with local partners, community ¶ ¶ https://together.takemehome.org engagement, effective data collection, and provider training. RWHAP activities have led to recent increases in viral suppression among MSM from 84.7% in 2015 to 89.1% in 2019 (13). These activities also reduced racial/ethnic disparities by as much as one third by addressing structural factors, such as unstable housing, that impede access to HIV care and treatment (14). Other programs have used surveillance data to identify persons not receiving care and have successfully engaged them using interventions such as patient navigation to reduce barriers to access (15).
Prevention of new infections can be enhanced by ensuring that PrEP providers are available in communities most affected by HIV and by integrating PrEP services into existing clinical settings, such as sexually transmitted disease (STD) clinics. As part of the EHE initiative, CDC supports local efforts to build the capacity of STD clinics to implement innovative, locally tailored strategies to provide PrEP and other HIV prevention services to MSM at risk for acquiring HIV.*** Such clinics often function as safety nets for populations with limited access to other sources of care, thus providing crucial prevention services to underserved populations and reducing racial/ethnic disparities in care (16). Local programs have highlighted opportunities to improve rapid PrEP initiation and navigation services for STD clinic patients with ongoing risk for HIV infection (17).
Emerging interventions and delivery strategies for testing, prevention, and treatment might also reduce barriers to accessing services and reaching EHE goals. Telehealth and other novel care models can provide additional options for accessing and improving adherence to HIV treatment and PrEP (18,19). Development of long-acting HIV medications could further expand access and facilitate adherence to PrEP and ART (20). Such innovative interventions and delivery strategies should be prioritized for use among the most disproportionately affected groups, including Black and Hispanic/Latino MSM and younger MSM. Their implementation should be designed to address structural factors that often limit access to and use of these technologies. To further promote engagement in HIV services and reduce HIV-related stigma, MSM should be engaged in HIV prevention or treatment services regardless of their HIV status (i.e., a status neutral approach) (1). This approach helps persons with HIV and persons at higher risk for infection receive the services needed to prevent HIV transmission or acquisition without status-specific structures that reinforce stigma and other related barriers. *** https://www.cdc.gov/hiv/funding/announcements/ps20-2010/index.html The findings in this report are subject to at least seven limitations. First, data were collected before the onset of the COVID-19 pandemic and do not reflect disruptions in HIV testing, prevention, or treatment services. Second, MMP and NHBS data were self-reported and are subject to recall and social desirability biases. Third, NHBS behavioral measures of likely PrEP indication did not correspond directly with clinical guidelines and might have underestimated MSM with likely PrEP indications who discussed PrEP with a health care provider or used PrEP. Fourth, viral suppression measures presented here did not include data from jurisdictions without complete laboratory reporting and therefore might not be representative of all persons with diagnosed HIV infection in the United States. Fifth, the small number of MSM in some subgroups might have reduced the reliability of their estimates. Sixth, outcomes based on NHSS data for MSM aged 13-24 years were presented for a single age category, potentially obscuring differences in this developmentally diverse group. Finally, NHSS data presented by transmission category (i.e., male-to-male sexual contact) are based on sex at birth. Therefore, estimates based on NHSS data included some persons with a gender identity other than male (e.g., transgender women) who were classified as MSM based on their sex at birth.
Intensified and innovative efforts to expand access to HIV testing, prevention, and treatment services for MSM, particularly Black MSM, Hispanic/Latino MSM, and younger MSM, are required to decrease health disparities and reduce new HIV infections by 90% to reach EHE goals. Jurisdictions should identify and implement those programs and interventions most responsive to local needs and acceptable to disproportionately affected populations of MSM. All programs should implement a status neutral approach to reduce barriers to prevention, testing, and treatment by breaking down institutional barriers and reducing HIV-related stigma.
COVID-19 outbreaks have been reported in homeless shelters across the United States (1). Many persons experiencing homelessness are older adults or persons with underlying medical conditions, placing them at increased risk for severe COVID-19-associated illness. The proportion of persons experiencing homelessness who are fully vaccinated against COVID-19 in the United States is currently unknown. Many persons experiencing homelessness express a willingness to receive the COVID-19 vaccine (2,3).
Through conversations with public health and housing assistance partners, CDC identified six* urban public health jurisdictions with data on vaccination coverage among persons experiencing homelessness. These six jurisdictions reported data on COVID-19 vaccinations † administered to persons experiencing intermittent homelessness during December 13, 2020-August 31, 2021. Full vaccination status § and evidence of coverage with at least 1 COVID-19 vaccine dose ¶ among persons experiencing homelessness were obtained by performing data linkage between immunization information systems and homeless services data systems or through data collection during vaccination events at homeless service sites. Total populations of persons experiencing homelessness were estimated using either the total number of persons accessing homeless services during the study period or an annual census of persons experiencing homelessness.** Vaccination coverage and size of the general population in each jurisdiction were obtained from CDC's COVID Data Tracker † † or from local health departments. The percentage point differences in vaccination coverage between persons experiencing homelessness and the general population were calculated, along with 95% CIs. This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy. § § Full COVID-19 vaccination coverage among persons experiencing homelessness ranged from 18.6% to 44.5% in the six jurisdictions compared with 43.6% to 59.8% in the general population in each jurisdiction or corresponding area (Table). In each jurisdiction, full vaccination coverage among persons experiencing homelessness was substantially lower (11.2-37.2 percentage points) than that among the general population of the respective jurisdiction. Coverage with at least 1 COVID-19 vaccine dose across the six jurisdictions ranged from 22.0% to 52.0% among persons experiencing homelessness, and from 46.5% to 65.7% in the respective general populations.
These estimates highlight relatively low COVID-19 vaccination coverage among persons experiencing homelessness compared with coverage in the general populations in a convenience sample of six jurisdictions. Estimating vaccination coverage for persons experiencing homelessness is challenging because housing status is not routinely collected in vaccination records. In addition, because homelessness could be temporary, estimating population size is difficult. Some health departments have overcome these challenges by fostering relationships with health clinics and homeless service providers. The use of integrated data systems to link deidentified, individual-level records across housing, health care, and public health systems is an emerging potential solution.
The findings in this report are subject to at least three limitations. First, because of varying data collection methods, comparison across jurisdictions was not possible. Second, the systems used for estimating homelessness rely on use of homeless services, and not all persons experiencing homelessness access these services, particularly persons living unsheltered. Finally, because of nonrandom selection and inclusion of only six jurisdictions, these findings are not generalizable to all persons experiencing homelessness in the United States, particularly in rural areas Given low COVID-19 vaccination coverage and increased risk for infection with SARS-CoV-2, the virus that causes COVID-19, in congregate settings (4), it is important that state and local health departments continue to follow CDC guidance to plan and respond to COVID-19 among persons experiencing homelessness. ¶ ¶ Vaccine access for persons experiencing homelessness can be enhanced by using multiple strategies (5), including pop-up vaccination clinics in convenient locations, mobile clinics in partnership with trusted providers, and street outreach teams. COVID-19 vaccination coverage can be improved by strengthening partnerships across health departments, health care clinics, and homeless ¶ ¶ https://www.cdc.gov/coronavirus/2019-ncov/community/homeless-shelters/ plan-prepare-respond.html (Accessed April 1, 2021). service providers. Furthermore, including persons who have experienced homelessness in vaccination planning is critical to helping ensure approaches are tailored to the needs of persons experiencing homelessness.

Summary
What is already known about this topic?
Increases in mental health conditions have been documented among the general population and health care workers during the COVID-19 pandemic; however, data on public health workers are limited.
What is added by this report?
Among 26,174 surveyed state, tribal, local, and territorial public health workers, 52.8% reported symptoms of at least one mental health condition in the past 2 weeks. Symptoms were more prevalent among those who were unable to take time off or worked ≥41 hours per week.
What are the implications for public health practice?
Implementing prevention and control practices that eliminate, reduce, and manage factors that cause or contribute to public health workers' poor mental health might improve mental health outcomes during emergencies.
respectively ( Table 1). The highest prevalences of symptoms of a mental health condition or suicidal ideation were among respondents aged ≤29 years (range = 13.6%-47.4%), transgender or nonbinary persons of all ages (range = 30.4%-65.5%), and those who identified as multiple races (range = 12.1%-43.4%); prevalence of symptoms of PTSD was higher among respondents who had a postbaccalaureate graduate education (40.7%). Most (92.6%) respondents reported working directly on COVID-19 response activities; the majority (59.2%) worked ≥41 hours in a typical week since March 2020. The prevalences of all four mental health outcomes and the severity of symptoms of depression or PTSD increased as the percentage of work time spent directly on COVID-19 response activities and number of work hours in a typical week increased (Table 1) (Figure). Public health workers who were unable to take time off from work when they needed were nearly twice as likely to report symptoms of an adverse mental health condition (prevalence ratio range = 1.84-1.95) as were those who could take time off. Among those not able to take time off from work (8,586), the most common reasons were concern about falling behind on work (64.4%), no work coverage (60.6%), and feeling guilty (59.0%); 18.2% reported that their employer did not allow time off from work. Needing mental health counseling/services in the last 4 weeks, but not receiving these services, was reported by nearly one in five (19.6%) respondents. Employee assistance programs were available to nearly two thirds (66.1%) of respondents but were accessed by only 11.7% of those respondents; 27.3% of all respondents did not know whether their employer offered an employee assistance program.
Respondents reported experiencing traumatic events or stressors since March 2020, including feeling overwhelmed by workload or family/work balance (72.0%), receiving job-related threats because of work (11.8%), and feeling bullied, threatened or harassed because of work (23.4%); 12.6% of respondents reported having received a diagnosis of COVID-19 (Table 2). Respondents who reported traumatic events or stressors, either personal or work-related, were more likely to report symptoms of PTSD than respondents who did not experience these events or stressors.

Discussion
Among a convenience sample of 26,174 state, tribal, local, and territorial public health workers, approximately one half experienced symptoms of a mental health condition in the 2 weeks preceding the survey, with highest prevalences reported among younger respondents, and transgender or nonbinary respondents. Public health workers who reported certain workplace practices, such as long work hours and the inability to take time off, were more likely to have experienced symptoms of a mental health condition. Implementing prevention and control practices that eliminate, reduce, and manage workplace factors ¶ ¶ that cause or contribute to public health workers' adverse mental health status*** might improve mental health outcomes during this and other public health emergencies.
The overall prevalence of symptoms of mental health conditions among public health workers was higher than previously reported in the general population (approximately 40.9%) (1). Prevalences of symptoms of depression and anxiety among public health workers were similar to those in previous reports among health care workers (3); however, prevalence of PTSD symptoms among public health workers was 10%-20% higher than that previously reported among health care workers (2), frontline personnel (3), and the general public (1). Symptoms of PTSD disproportionately affected public health workers who experienced work-related traumatic stressors (e.g., felt inadequately compensated or felt unappreciated at work), particularly those factors that affect workers' personal lives (e.g., felt disconnected from family and friends because of workload). Traumatic and stressful work experiences related to the COVID-19 pandemic might have played a role in elevating the risk for experiencing symptoms of PTSD among public health workers.
Increases in adverse mental health symptoms among workers have been linked to increased absenteeism, high turnover, lower productivity, and lower morale, which could influence the effectiveness of public health organizations during emergencies (8,9). Among public health worker respondents, nearly 20% reported that their employer did not allow them to take time off; the inability to take time off had the largest impact on reporting ¶ ¶ https://www.cdc.gov/niosh/twh/guidelines.html *** https://unhealthywork.org/category/mental-health-outcomes/ Abbreviations: AI/AN = American Indian or Alaska Native; CI = confidence interval; IES-6 = 6-item Impact of Event Scale; GAD-2 = General Anxiety Disorder; NH = non-Hispanic; NH/PI = Native Hawaiian or Pacific Islander; PHQ-9 = 9-item Patient Health Questionnaire; PR = prevalence ratio; PTSD = post-traumatic stress disorder; Ref = referent group. * Symptoms of mental health conditions were scored and categorized by severity. Respondents who scored ≥10.0 out of 27 on the PHQ-9 for depression, ≥3.0 out of 6 on the GAD-2 for anxiety, or ≥1.75 out of 4 on the IES-6 for PTSD were considered symptomatic for the respective conditions. Respondents who indicated that they would be better off dead or thought of hurting themselves at any time in the past 2 weeks were categorized as experiencing suicidal ideation. † Some categories might not sum to 26,174 because of missing data. Denominators for categories are respondents who answered the questions to be scored. symptoms of mental health. Approximately one quarter of public health workers did not know whether their workplace offered an employee assistance program. Even where available, employee assistance programs were not commonly accessed. Several strategies could reduce adverse mental health symptoms among public health workers during public health emergencies. For example, expanding staffing size (e.g., recruiting surge personnel to backfill positions) and implementing flexible schedules might reduce the need for long work hours; encouraging workers to take regular breaks and time off could help avoid overwork and reduce the risk for adverse mental health outcomes. In addition, implementing, evaluating, and promoting use of employee assistance programs could improve employee resiliency and coping.
The findings in this report are subject to at least four limitations. First, the study used a nonprobability-based convenience sample of public health worker respondents, and a completion rate could not be determined. Although the participating national public health membership associations reach many public health workers, the findings might not be representative of all state, tribal, local, and territorial public health workers in risk, including those who are younger (10) or transgender or nonbinary persons. In addition, employee assistance programs could be evaluated and adjusted to be more accessible and acceptable to workers and focus more on building workplace cultures that promote wellness and destigmatize requests for mental health assistance. In 2020, 31.8% of live births were to women who had a cesarean delivery. The rate of cesarean delivery was lowest for women who were underweight before pregnancy (20.7%); the rate rose steadily as BMI increased to obesity class III (52.3%). One quarter (25.1%) of women of normal weight had a cesarean delivery.