Purpose
To identify common causes of emergency department–treated eye injury among older adults in the United States and to characterize fall-related ocular trauma in this population.
Design: Retrospective cohort study.
Methods
Data from the National Electronic Injury Surveillance System, a nationally representative database of US emergency department–treated injuries, was used to assemble a cohort of adults ≥65 years of age with eye injuries between January 1, 2000 and December 31, 2019. Demographic information, diagnosis, disposition, injury location, and the consumer product associated with injury were collected. Narrative descriptions of all injuries were reviewed to identify eye injuries caused by falls.
Results
Four thousand nine hundred fifty-three eye injuries among older adults were reported from 2000-2019, a stratified probability sample representing approximately 238,162 injuries, with an average annual frequency of 12,000 injuries. Falls accounted for 11.5% of these injuries. Fall-related eye injuries commonly presented from home (66.5%) and were more likely to occur in the winter than eye injuries from other causes (28.1% vs 18.4%, P < .01). Risk factors for fall-related eye injury included older age (odds ratio [OR] 1.11 [95% confidence interval {CI} 1.10-1.13 per year), female sex (OR 2.3 [95% CI 1.6-3.1] vs male), black race (OR 2.4 [95% CI 1.3-4.5] vs white), and presentation from a nursing home (OR 12.7 [95% CI 4.9-32.8] vs other locations). Older adults with fall-related injuries were more likely to be hospitalized (OR 22.8 [95% CI 15.3-33.9]) and to have a ruptured globe (OR 14.1 [95% CI 6.5-30.6]) than those with fall-unrelated injury.
Conclusions
Falls are an important mechanism of ocular trauma in older adults and are associated with worse outcomes compared with eye injuries from other causes.
E E ye injuries contribute substantially toward visual impairment and blindness in the United States, , with the reported incidence of emergency department (ED)–treated eye injuries estimated at 3.15 per 1000 individuals. In 2006, 1.5 million individuals in the United States were blind because of ocular trauma and 1.7 million were visually impaired. While most eye injuries affect males in their 20s and 30s (commonly as a result of foreign bodies and projectile objects), ocular trauma can have a significant impact on visual impairment and quality of life among older adults. A study of consumer product categories associated with ocular trauma among older adults has shown that 15% of ED-treated eye injuries in this population are associated with home furniture. The authors suggest, upon evaluation of narrative descriptions of these injuries, that many are caused by falls. Older adults are at an increased risk of falling, and falls in this population contribute toward morbidity, mortality, and quality of life. , , In 2014, 28.7% of older adults in the United States reported falling, and falls in this population resulted in 7 million injuries. Notably, fall-related eye injury is more likely to lead to blindness than other causes of ocular trauma, with fall-related open globe injuries associated with more severe ocular complications and poorer visual acuity (VA) outcomes compared with globe injuries from other causes. ,
Despite the demonstrated association between eye injury and visual impairment among the elderly, and the subsequent impact on morbidity and quality of life, few studies describe causes of and risk factors for ocular trauma in the elderly. Epidemiologic characteristics of fall-related eye injuries, in particular, are not well understood. While population-adjusted incidence of eye injuries decreases with age, the steadily aging U.S. population (1 in 5 residents predicted to be of retirement age by 2030 ) makes understanding the epidemiology of eye trauma in this population crucial.
The purpose of our study is to determine the proportion of eye injuries caused by falls in adults ≥65 years of age (defined here as older adults) using a nationally representative sample of ED-treated eye injuries and to identify the risk factors, mechanisms, and outcomes associated with fall-related eye injuries. An understanding of the epidemiology of ocular trauma in older adults in the United States, with particular focus on fall-related injury, will facilitate targeted research and intervention important for injury prevention and for the preservation of vision and quality of life in this population.
Methods
Data Acquisition
Data for this retrospective cohort study were obtained from the US Consumer Product Safety Commission (USCPSC) National Electronic Injury Surveillance System (NEISS). The USPSC is a federal commission charged with monitoring and preventing threats of injury or death from consumer products in the United States and operates the NEISS database to collect data on consumer product–related injuries in the United States and its territories from a sample of 100 hospitals. These hospitals are a stratified probability sample of all US hospitals that have ≥6 beds and provide 24-hour ED services. Patient information is collected from each NEISS hospital for every emergency visit involving an injury associated with consumer products. Data collected by the NEISS include demographic information, product(s) associated with injury, body part(s) involved, injury location, diagnosis, disposition, and a short narrative description of the ED visit. Data from the NEISS database are publicly available and deidentified and may be downloaded upon request. This study was therefore exempt from institutional review board approval and patient permission was waived.
Study Design: Participants, Cohort Definition, and Cohort Description
Eye injuries among individuals ≥65 years of age between January 1, 2000 and December 31, 2019 (inclusive) were included in this study. Narrative descriptions were reviewed and subjects were excluded if they did not experience a true injury according to the diagnosis and/or narrative (eg, retinal artery occlusion, nontraumatic vitreous detachment, and acute angle closure).
More than 400 distinct products were reported as contributing to injuries in our sample. We grouped product codes into product categories defined by the USCPSC. For all individuals with a diagnosis defined as “other,” narrative diagnoses were reviewed and further categorized into injury types (eg, ruptured globe, subconjunctival hemorrhage, and hyphema). All narrative descriptions were reviewed to define the following: mechanism of injury and, if related to a fall, the fall mechanism and position before the fall. In addition, narrative descriptions were reviewed to identify injuries occurring in nursing homes, almost all of which are classified as “public” in the NEISS database. A separate “nursing home” category was created. Injuries were classified in the following categories: fall, chemical, thermal, blunt trauma, foreign body (organic), foreign body (inorganic), projectile, and assault. Fall mechanisms were classified as tripping, slipping, sleep-related (eg, “falling out of bed” and “found on the floor next to bed”), loss of balance, weakness, dizziness, syncope, or alcohol-related. Position before the fall was defined as standing, seated, lying, or climbing stairs.
Statistical Methods
Descriptive statistics were used to compare eye injuries stratified into fall-related and non–fall-related injuries and to report the frequency of injuries by year and by month. Two-sample t tests and Pearson χ 2 tests were used to assess differences between fall-related and non–fall-related injuries, where appropriate. Bar graphs were used to represent the frequency of fall-related and non–fall-related eye injuries by year and by month. A multivariable logistic regression model was used to determine risk factors associated with fall-related injuries. Separate multivariable regression analyses were used to determine the relationship between fall-related injury and ruptured globe, and that between fall-related injury and hospitalization.
In all multivariable models, covariates considered for inclusion were age, sex, race, injury location, and season and were included if they were significant at P < .1 on univariate analyses. Given the probability sampling design, we used weights provided by NEISS to provide valid estimates for all analyses. All statistical analyses were carried out using STATA software (v 16.1; StataCorp, College Station, Texas, USA). P < .05 was considered statistically significant.
Results
Overall Sample Characteristics
There was a total of 4953 eye injuries among individuals ≥65 years of age in the USCPCS NEISS database from 2000-2019, representing approximately 238,162 individuals ≥65 years of age presenting to US EDs. The mean age of individuals in this sample was 73.2 (95% CI 72.8-73.5) years, 63.1% (95% CI 60.8%-65.3%) were male, and 59.6% (95% CI 51.6%-67.1%) were white. Sample characteristics are summarized in Table 1 . The estimated annual frequencies of ED-treated eye injury among older adults increased between 2000 and 2019 ( P < .001), with an average annual frequency of 11,908 eye injuries between 2000 and 2019. Eleven and a half percent (95% CI 9.3%-14.1%) of these injuries were fall-related. Other common mechanisms of injury included chemical (21.1% [95% CI 18.6%-23.9%), projectile (19.7% [95% CI 17.5%-22.2%), blunt (12.8% [95% CI 11.7%-14.0%), and foreign body (inorganic 13.7% [95% CI 12.5%-15.0%]; organic 10% [95% CI 8.9%-11.3%]).
| Overall (N = 4953) | Fall-Related Injury (n = 730) | Non–Fall-Related Injury (n = 4223) | P Value | |
|---|---|---|---|---|
| Estimated population size a | 238,162 | 27,371 | 210,791 | — |
| Age, y (CI) | 73.2 (72.8-73.5) | 79.8 (79.0-80.5) | 72.3 (72.0-72.6) | <.001 |
| Female, % (CI) | 36.9 (34.8-39.2) | 62.0 (57.3-66.5) | 33.7 (31.4-36.0) | <.001 |
| Race, % (CI) | ||||
| White/Caucasian | 59.6 (51.6-67.1) | 48.4 (38.3-58.7) | 61.0 (52.8-68.7) | < .01 |
| Black/African American | 7.2 (4.8-10.8) | 13.1 (6.7-24.2) | 6.5 (4.4-9.5) | |
| Other/unspecified | 33.2 (25.6-41.8) | 38.4 (27.6-50.6) | 32.5 (24.7-41.4) | |
| Season, % (CI) | ||||
| Winter | 19.5 (17.4-21.8) | 28.1 (23.0-33.7) | 18.4 (16.1-20.9) | <.01 |
| Spring | 26.6 (25.1-28.1) | 24.9 (21.4-28.7) | 26.8 (25.2-28.5) | |
| Summer | 28.8 (26.1-31.6) | 23.5 (20.2-27.0) | 29.4 (26.5-32.6) | |
| Fall | 25.2 (23.7-26.8) | 23.6 (18.2-30.0) | 25.4 (23.7-27.2) | |
| Location, % (CI) | ||||
| Home | 67.0 (62.1-71.6) | 66.5 (60.0-72.3) | 67.1 (61.9-72.0) | <.001 |
| Public | 2.9 (2.3-3.7) | 7.0 (4.5-10.7) | 2.4 (1.9-3.1) | |
| Nursing home | 1.5 (.8-2.7) | 10.4 (5.8-17.8) | 0.3 (0.2-0.6) | |
| Other b | 3.8 (2.7-5.3) | 1.6 (0.8-3.2) | 4.1 (2.9-5.7) | |
| Unknown | 24.8 (20.0-30.2) | 14.7 (10.3-20.4) | 26.1 (20.1-31.9) | |
| Product category, % (CI) | ||||
| Workshop | 20.3 (18.4-22.3) | 0.4 (0.1-2.0) | 22.8 (20.7-25.1) | <.001 |
| Yard and garden | 17.5 (15.6-19.7) | 1.7 (0.8-3.9) | 19.6 (17.5-21.9) | |
| Home maintenance | 15.9 (14.2-17.8) | 0.3 (0.0-2.2) | 18.0 (16.1-19.9) | |
| Home furnishings | 12.2 (10.9-13.6) | 42.8 (37.5-48.2) | 8.3 (7.3-9.3) | |
| Home structures | 9.2 (7.7-11.0) | 34.9 (28.8-41.5) | 5.9 (5.0-6.8) | |
| Personal use items | 6.2 (5.5-7.1) | 9.1 (6.9-12.1) | 5.9 (5.1-6.8) | |
| Sports and recreation equipment | 5.7 (4.2-7.7) | 3.4 (2.0-5.6) | 6.0 (4.5-8.1) | |
| Packaging | 4.3 (3.6-5.0) | 1.6 (0.8-3.3) | 4.6 (3.9-5.5) |
a Estimated using stratified probability sampling of all US hospitals with >5 beds and 24-hour emergency department care.
b Other locations: farm, street, mobile, industry, school, or sports.
Characteristics of Fall-Related vs Unrelated Eye Injuries
There were 730 fall-related ED-treated eye injuries among older individuals in the NEISS database from 2000-2019, representing approximately 27,371 total injuries across US EDs. The annual frequency of fall-related eye injuries increased between 2000 and 2019 ( P < .001), with an estimated average annual frequency of approximately 1,369 injuries. Annual frequencies of fall-related and non–fall-related injuries are shown in Figure 1 . Characteristics of individuals with fall-related and non–fall-related eye injuries are listed in Table 1 . Among those with fall-related injuries, the mean age was 79.8 years (95% CI 79.0-80.5), 62% (95% CI 57.3%-66.5%) were female, and 48.4% (95% CI 38.3%-58.7%) were white. The mechanism of fall could not be verified from narrative descriptions in 43% (95% CI 37.2%-49.3%) of cases. Tripping caused 21.9% (95% CI 16.3%-28.8%) of falls, 14.0% (95% CI 11.3%-17.3%) were sleep-related (fell out of bed), 8.6% (95% CI 6.0%-12.0%) caused by a loss of balance, and 7.6% (95% CI 5.3%-10.8%) were caused by slipping. Most falls occurred at home (66.5% [95% CI 60.0%-72.3%) and involved home furnishings (42.8% [95% CI 37.5%-48.2%) or home structures (34.9% [95% CI 28.8%-41.5%). More than 10% (10.4% [95% CI 5.8%-17.8%]) of fall-related injuries occurred in nursing homes. In 28.1% (95% CI 23.0%-33.7%), fall-related injuries occurred in the winter months of December, January, and February, and slipping as a fall mechanism was more common in the winter compared with the summer months of June, July, and August (12.7% [95% CI 7.8%-20.0%] vs 2.4% [95% CI 0.8%-6.9%], respectively, P < .05). Monthly frequencies of fall-related and non–fall-related injuries are shown in Figure 2 .
