eTable. ICD-9-CM Codes for Eye Trauma
eFigure. Age-specific incidence of hospitalizations for eye trauma as primary or secondary admitting diagnosis in the United States, 2001-2014
Customize your JAMA Network experience by selecting one or more topics from the list below.
Iftikhar M, Latif A, Farid UZ, Usmani B, Canner JK, Shah SMA. Changes in the Incidence of Eye Trauma Hospitalizations in the United States From 2001 Through 2014. JAMA Ophthalmol. 2019;137(1):48–56. doi:10.1001/jamaophthalmol.2018.4685
What changes have occurred in the incidence of hospitalizations for eye trauma in the United States from 2001 through 2014?
In a cohort study using the National Inpatient Sample (a representative sample of all US community hospital discharges), the incidence of eye trauma as a primary admitting diagnosis remained constant over time and was more common in men, children, adolescents, nonwhite individuals, and individuals with lower socioeconomic status. Eye trauma as a secondary diagnosis increased because of an increased incidence of falls in elderly individuals.
In light of these findings, it is important to increase awareness and develop prevention strategies to limit eye trauma.
Eye trauma is a common cause of vision loss and a substantial public health problem.
To determine the changes in the incidence of eye trauma hospitalizations in the United States and compare the demographics of affected patients and outcomes of eye trauma as a primary or secondary admitting diagnosis.
Design, Setting, and Participants
This retrospective longitudinal cohort study used the National Inpatient Sample, a representative sample of all US community hospitals, to determine the incidence, characteristics, and causes of primary and secondary inpatient eye trauma admissions from 2001 through 2014. All inpatients with relevant diagnoses were included. Linear regression was used to estimate changes in incidence. Logistic regression was used to compare demographics and outcomes between primary and secondary diagnoses, including age, sex, race, income, primary payer, region, year of admission, length of stay, cost, and disposition at discharge.
Main Outcomes and Measures
Incidence and characteristics of inpatient primary and secondary eye trauma.
From 2001 to 2014, there were an estimated 939 608 inpatient admissions (of whom 556 886 were male patients [59.3%]; overall mean [SD] age, 49.4 [25.2] years) in the United States because of eye trauma diagnoses, with 778 967 of these (82.9%) as a secondary diagnosis. The incidence of primary eye trauma decreased from 3.9 to 3.0 per 100 000 population (difference, 0.9 [95% CI, 0.2-1.6] per 100 000 population; P = .001). The incidence of eye trauma as a secondary admitting diagnosis increased from 14.5 to 19.0 per 100 000 population (difference, 4.5 [95% CI, 1.9-7.2] per 100 000 population; P = .004). This was largely attributed to an increasing number of falls in individuals older than 65 years. The most frequent diagnosis was orbital fracture (64 149 [39.9%]) for primary trauma and contusion of eye and adnexa (19 301 [37.8%]) for secondary trauma. Primary trauma was more common in children (adjusted odds ratio [aOR], 2.21 [95% CI, 2.09-2.32]) and adolescents (aOR, 1.25 [95% CI, 1.19-1.32]) than adults (reference), African American individuals (aOR, 1.89 [95% CI, 1.81-1.97]) and Hispanic individuals (aOR, 1.52 [95% CI, 1.45-1.59]) than white individuals, and uninsured patients (aOR, 1.14 [95% CI, 1.07-1.22]) and those receiving Medicaid (aOR, 1.12 [95% CI, 1.05-1.19]) than Medicare beneficiaries. Patients with a primary diagnosis were more likely to have a stay of less than 3 days (patients with a primary diagnosis: 101 796 [63.4%]; secondary diagnosis: 274 538 [35.2%]), more likely to have costs in the lowest quartile (patients with a primary diagnosis: 51 212 [31.9%]; secondary diagnosis: 166 260 [21.3%]), and less likely to die (patients with a primary diagnosis: 526 [0.3%]; secondary diagnosis: 20 929 [2.7%]).
Conclusions and Relevance
These findings suggest that the increasing number of falls in individuals older than 65 years and the high risk of primary eye trauma in populations such as children and adolescents warrant the development and implementation of effective preventive strategies. Many of these patients are seen in ophthalmology practices where proactive risk assessment and counseling can play a critical role.
Eye trauma is an important public health problem and one of the most common causes of preventable visual impairment worldwide.1 In the United States alone, an estimated 2.0 to 2.4 million cases of eye trauma occur each year, and nearly 1 million individuals have permanent significant visual impairment because of injury, with more than 75% of these becoming monocularly blind.1-3 In fact, it is estimated that no less than 27% of serious eye injuries in the United States lead to legal blindness (as defined by a visual acuity worse than 20/200).4 Consistently, trauma is considered a leading cause of visual impairment throughout the industrialized world.5-10
Given its severity and potential to cause permanent impairment, trauma resulting in hospitalization is of the highest concern. Indeed, as ophthalmology has transitioned into a primarily outpatient service, trauma has become the second most common cause of ophthalmic inpatient admissions.11 However, there is no recent information regarding the incidence and characteristics of hospitalization for eye trauma on a national level, because previous studies on the topic were either performed more than a decade ago or focused on specific subpopulations.12-16
Epidemiologically, it is challenging to characterize hospitalization for eye trauma for a number of reasons. In most multiple-trauma cases, the identification of eye injuries is difficult because of the presence of life-threatening injuries that receive higher priority. Eye trauma is generally addressed only after the patient has been stabilized, and even then, it may be an incidental finding that would not have warranted an admission by itself. This is supported by the fact that it is not uncommon for certain injuries to go undetected, even during secondary evaluation.10,17 Eye trauma in particular can be difficult to detect because of periorbital swelling that complicates examination of the eye and/or unconsciousness in the patient, which creates the inability to assess visual acuity.18 Excluding secondary eye trauma would therefore underestimate the true health care burden of eye injuries.
The purpose of this study was to estimate the changes in incidence and characteristics of all eye trauma hospitalizations in the United States from 2001 through 2014 and to compare the demographics of affected patients and outcomes between eye trauma as a primary or secondary admitting diagnosis.
The National Inpatient Sample (NIS) is the largest publicly available all-payer inpatient database in the United States and is sponsored by the government as a part of the Healthcare Cost and Utilization Project (HCUP). It contains discharge-level data drawn from all states participating in the HCUP, and it includes as much as 96% of the country’s population.19,20 Prior to 2012, the database comprised a 20% stratified systematic sample of all US community hospitals, with all discharges retained from those hospitals. In 2012, it was redesigned to improve national estimates, and it now approximates a 20% stratified sample of discharges from all hospitals in the HCUP.
We conducted a retrospective longitudinal study using the NIS from 2001 through 2014. The NIS provides up to 25 diagnoses and 15 procedures for each hospitalization record, coded using the standard International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) codes. The first diagnosis in the database is referred to as the primary diagnosis and is considered the principal reason for admission to the hospital. Secondary diagnoses are additional diagnoses recorded during the admission. We used ICD-9-CM diagnosis codes to identify admissions with a primary or secondary diagnosis of eye trauma, in line with previous studies (eTable in the Supplement).14-16,21,22 A total of 195 690 cases were captured for analysis and subsequently reviewed.
The Johns Hopkins institutional review board approved this study. Because the database does not contain any direct patient identifiers, informed consent was not required.
The NIS permits calculation of projected national estimates for each hospital admission using discharge weights.19 The HCUP also provides elements termed trend weights to account for the change in sampling methodology in 2012 and allow longitudinal analysis.23 Population data from the US Census Bureau were then used to calculate total and age-specific incidence of admission.24,25 This methodology has been described and validated by the HCUP and several recent studies.26,27
The most frequent primary and secondary diagnoses were tabulated along with the external causes of injury and poisoning codes (E-codes), which were used to determine the mechanism of injury. The most frequent ophthalmic procedures (as described by ICD-9-CM procedure codes 08-16) were also tabulated.
Descriptive statistics were calculated for all relevant characteristics, including age, sex, race, household income quartile of the patient’s residential zip code, primary payer, hospital region, hospital location, hospital teaching status, hospital bed size, admission type, admission source, total number of diagnoses recorded, total number of chronic conditions, total number of procedures performed, length of hospital stay, cost, and disposition. Age was categorized in accordance with known standards as infants (0-1 year), children (1-12 years), adolescents (13-19 years), adults (20-44 years), middle-aged individuals (45-64 years), and elderly individuals (≥65 years). Length of stay was categorized as short (<3 days), intermediate (3-6 days), or prolonged (>6 days), which is in line with what is clinically accepted and similar to what has been defined previously.21
Cost was categorized in quartiles. Costs are the actual expenses incurred by hospital in the provision of patient care and were calculated using hospital charges and standardized hospital-specific ratios derived from accounting reports collected by the Centers for Medicare and Medicaid Services.28 All costs were inflation-adjusted based on the 2014 US dollar value using the Consumer Price Index for Hospital Services from the US Bureau of Labor Statistics.29
Linear regression was used to estimate the changes in annual incidence of eye trauma, as derived from the national estimates and population data. Multivariable logistic regression was used to compare the demographics and outcomes between eye trauma as a primary or secondary diagnosis. Covariates included age, sex, race, household income quartile, primary payer, hospital region, year of admission, length of stay, cost, and disposition. The variables were considered for inclusion based on clinical importance and were only included in the final model if the likelihood ratio was significant on univariate regression. All variables mentioned above had a P value less than .05 and were thus kept in the final model. Stata version 14 (StataCorp) was used for all statistical analysis.
During 2001 through 2014, an estimated 939 608 patients with eye trauma were hospitalized in the United States, of whom 160 641 (17.1%) had eye trauma as a primary diagnosis and 778 967 (82.9%) had it as a secondary diagnosis. Thus, eye trauma was 4.8 times (95% CI, 4.6-5.2 times) more likely to be a secondary than primary inpatient diagnosis during the study period.
The incidence of total eye trauma increased from 18.3 per 100 000 population in 2001 to 22.0 per 100 000 population in 2014 (difference, 3.7 [95% CI, 0.5-6.8] per 100 000 population). This was because of the increase in eye trauma as a secondary diagnosis, which rose 31% from 14.5 to 19.0 per 100 000 population (difference, 4.5 [95% CI, 1.9-7.2] per 100 000 population; P = .004]) in the study period. In contrast, eye trauma as a primary diagnosis decreased, although only marginally, from 3.9 to 3.0 per 100 000 population (difference, 0.9 [95% CI, 0.2-1.6] per 100 000 population; P = .001]) in the study period.
Figure 1 show the changes in incidence of eye trauma. Elderly individuals had the highest incidence of all age groups for both primary and secondary eye trauma (eFigure in the Supplement).
For eye trauma as a primary diagnosis, the most frequent diagnoses were orbital fracture (in 64 149 patients [39.9%]), ocular laceration (34 167 [21.3%]), and eyelid laceration (12 327 [7.7%]). For eye trauma as a secondary diagnosis, the most frequent diagnoses were contusion of eye and adnexa (294 483 [37.8%]), orbital floor fracture (188 620 [24.2%]), and superficial injury of the eye and adnexa (103 084 [13.2%]; Table 1). The most frequent primary diagnoses in patients with secondary eye trauma were fractures of the base of the skull (82 726 [10.6%]), fractures of face bones (81 480 [10.5%]), and intracranial hemorrhages (46 115 [5.9%]).
For 868 104 (92.4%) of all records, E-codes were available. Figure 2 demonstrates the distribution of causes by age and sex for primary and secondary trauma in the study period. The major causes of injury for eye trauma as a primary diagnosis were falls (40 710 [25.3%]) and assault (39 751 [24.7%]), while those for eye trauma as a secondary diagnosis were falls (274 369 [35.2%]) and motor vehicle crashes (207 207 [26.6%]). Falls mostly occurred in elderly individuals (212 102 [67.3%]). Unintentional injuries caused by objects and machinery were the largest contributor of eye trauma as a primary diagnosis in children (5487 [30.1%]). Motor vehicle crashes were the largest contributor for eye trauma as a secondary diagnosis in adolescents (29 692 [58.6%]).
Figure 3 demonstrates the incidence of the major causes of trauma in the study period. Falls increased steadily over the years, from 4.5 to 8.2 per 100 000 population (difference, 3.7 [95% CI, 3.0-4.4] per 100 000 population; P = .001). Assault, motor vehicle crashes, and unintentional injuries caused by objects and machinery remained relatively constant over time.
An ophthalmic procedure was performed in 79 112 of patients (49.2%) with primary eye trauma diagnoses and 127 650 of patients (16.4%) with secondary eye trauma diagnoses. Repair of the eyelid or an eyebrow laceration was the most common procedure for both patients with primary eye trauma (24 588 [15.3%]) and patients with secondary eye trauma (100 013 [12.8%]).
The mean (SD) age of all patients was 49.4 (25.2) years, with elderly individuals composing the largest age group overall (299 599 [31.9]). Most were male (556 886 [59.3%]), white (526 415 [56.0%]), in the lowest income quartile (249 470 [26.6%]), publicly insured through Medicare or Medicaid (435 211 [46.3%]), and from the South (344 916 [36.7%]).
Eye trauma was more likely to be a primary diagnosis in children (adjusted OR [aOR], 2.21 [95% CI, 2.09-2.23]) and adolescents (aOR, 1.25 [95% CI, 1.19-1.32]) than adults, male individuals (aOR, 1.21 [95% CI, 1.17–1.24]) than female individuals, African-American individuals (aOR, 1.89 [95% CI, 1.81-1.97]) and Hispanic individuals (aOR, 1.52 [95% CI, 1.45-1.59]) than white individuals, uninsured patients (aOR, 1.14; [95% CI, 1.07-1.22]) and Medicaid enrollees (aOR, 1.12 [95% CI, 1.05-1.19]) than those receiving Medicare, and the Northeast (aOR, 1.42 [95% CI, 1.36-1.47]) than the South (Table 2).
The median (interquartile range [IQR]) length of hospital stay was 2 (1-3) days for primary diagnoses, 4 (2-7) days for secondary diagnoses, and 3 (2-6) days overall. The median (IQR) inflation-adjusted cost per patient was approximately $8000 (IQR, $5000-$13 000) for primary diagnoses, $12 000 ($7000-$24 000) for secondary diagnoses, and $11 000 ($6000-$21 000) overall. The mean annual inflation-adjusted cost was $122 million for primary diagnoses and $1.2 billion for secondary diagnoses. Most patients had a routine discharge (618 227 [65.8%]) and mortality was 0.3% for primary diagnoses (n = 526), 2.7% for secondary diagnoses (n = 20 929), and 2.3% overall (n = 21 455).
Patients with eye trauma as a primary diagnosis were less likely to have a prolonged stay (defined as >6 days; OR, 0.32 [95% CI, 0.30-0.34]), have a high cost (in the fourth quartile of cost; OR, 0.53 [95% CI, 0.50-0.56]) or die (OR, 0.14 [95% CI, 0.11-0.17]) than those with eye trauma as a secondary diagnosis.
Eye trauma can have an enormous negative influence on a person’s quality of life and economic productivity, especially if the injury results in permanent impairment or loss of vision.14,22,30 The exact incidence of eye trauma is difficult to determine, because it varies greatly depending on the study design and population. Estimates include 490 to 975 cases per 100 000 population of self-reported injuries treated in any health care setting,31,32 447 to 698 cases per 100 000 population of injuries treated in an outpatient setting,33,34 and 212 to 280 cases per 100 000 population of injuries treated in the emergency department.35
McGwin et al2 reported an incidence of 18.0 cases per 100 000 population for total inpatient eye trauma in 2001, which matches closely with this study. However, that is much less than the numbers that have been reported in previous decades. Tielsch et al13 reported an incidence of hospitalization of 13.2 per 100 000 population of primary eye trauma and 27.3 cases per 100 000 population of total eye trauma in Maryland from 1979 through 1986. Klopfer et al14 reported similar incidences for the whole country from 1984 through 1987: 13.2 per 100 000 population of primary eye trauma and 29.1 cases per 100 000 population of total eye trauma. It is important to note that Tielsch et al13 did report a statistically significant decrease in incidence over the course of their study period, which is what we observed in our study as well. Taken together, this could account for a definitive decline in eye trauma as a primary cause of admission over the years. Possible reasons include an actual decline in the incidence of eye trauma, improvements in imaging and diagnostic modalities, or a change in indications for hospitalization with a shift in management to outpatient settings.
However, for eye trauma as a secondary diagnosis, even though our figures are less than those reported earlier, we observed an overall increase in incidence instead. A possible explanation to reconcile these results is that the interval period from 1984 through 2001 saw a decrease in total eye trauma because of improvements in work safety and vehicle safety regulations.36,37 The ensuing increase could then be explained by the increased rate of falls in an aging population, which is supported by the evidence, in that 82% of the increase was attributed to falls.
This may pose a major cause for concern for the health care community. Falls have long been known as the leading cause of morbidity and mortality in elderly individuals, to the extent that they have been declared a never event in the health care system.38-40 Unfortunately, the rate of falls in the United States has increased steadily over the past decade and is likely to continue to rise as life expectancy increases and elderly individuals become more mobile.41-43 While preventing falls is difficult and complex, it is not impossible. Several strategies have shown success in reducing the risk of falls, and it is imperative that all health care professionals incorporate fall prevention as a routine part of clinical care.39 Elderly individuals with vision loss (eg, from macular degeneration or prior eye injury) may be more susceptible to falls, and, as such, ophthalmologists should pay special attention to assessing and educating these patients. Effective, widespread implementation of targeted interventions will be crucial to decreasing the future health care burden of this problem.
Another area that might benefit from further research is orbital fracture, which is a leading primary and secondary diagnosis for eye trauma. Previous studies in different subpopulations and different countries have also reported a high incidence of orbital fracture, as well as a high likelihood of it causing significant visual impairment.10,12,16 Most orbital fractures can be successfully managed in the emergency department or outpatient setting, unless there is muscle entrapment, globe rupture, or optic nerve injury. A thorough clinical examination by a trained ophthalmologist supplemented with adequate radiological imaging can effectively exclude such cases. It may be worthwhile to develop guidelines or protocols to evaluate such patients, analyze the need for admission, establish indications and timelines for surgical intervention, and improve interdisciplinary communication and management. This could allow health care systems to optimize patient care and improve patient outcomes.
Stratification of eye trauma as a primary or secondary diagnosis demonstrated notable differences between the 2 in terms of both demographics and outcomes. For instance, children were more likely to have eye trauma as a primary diagnosis, which was mostly caused by unintentional injury by objects and machinery. This makes clear the importance for educating parents and children about the potential for eye injuries at home and calls for increased promotion for the use of appropriate protective eyewear during high-risk activities. Men were also more likely to have eye trauma as a primary diagnosis, a finding that aligns with previous studies.8,13-16 This may be attributed to a higher likelihood of men engaging in aggressive or violent behavior (as suggested by an increased prevalence of assault) or working hazardous jobs that could lead to occupational injury (as suggested by an increased prevalence of unintentional injuries by objects and machinery).5,7,43 African American individuals, Hispanic individuals, and populations likely to be of lower socioeconomic status, such as uninsured individuals and Medicaid patients, were also more likely to have eye trauma as a primary diagnosis. Racial disparities in rates of eye trauma have been reported before9,13,32 and may be associated with the same reasons mentioned above, namely, a higher likelihood of experiencing violence and working at jobs that could cause occupational injury. It is also possible that such differences are not just because of a higher incidence of injury in these populations but also a higher rate of complications that would necessitate inpatient admission. This could be because of factors such as a lack of adequate access to outpatient or ambulatory facilities, variations in referral practices, or delays in seeking care. Further research is necessary to understand the reasons for these disparities, so that they can be properly addressed.
In terms of outcomes, patients with eye trauma as a primary diagnosis had shorter hospital stays, lower costs, and decreased mortality. This may be because patients with eye trauma as a secondary diagnosis are likely to have extensive injuries that led to longer and more complicated hospital stays. However, it is important to note that we are unable to comment on any differences in visual outcomes, such as visual acuity.
The strengths of this study include a national sample that is representative of the general population and a longitudinal design that that allows calculation of incidence and trend analysis. In addition, it includes both primary and secondary diagnoses, which permits stratification.
This study is subject to certain limitations, most of which are inherent to large administrative database studies. First, hospital discharge information on outcomes is limited to general measures, such as length of stay, total charges, and disposition at discharge. As such, there is no specific or detailed clinical information regarding the injury, its severity, and its subsequent visual outcome. Second, the discharge summaries are prepared by coding specialists, which means they are susceptible to oversights, such as missing data and errors in documentation. Third, because the NIS contains discharge-level data rather than patient-level data, it is possible that a patient with recurrent admissions may be represented more than once, which may artificially inflate certain admission diagnoses. Last, there are certain limitations on sampling, and the frame is restricted by the availability of data from sources currently participating in the HCUP.20
In conclusion, this study presents the incidence of hospitalizations because of eye trauma in the United States, as well as the causes and outcomes of these admissions and the demographic characteristics of the patients. We are hopeful that these data will serve as a good foundation to develop guidelines and quality improvement projects that test the efficacy of various interventions to reduce the burden and improve the outcomes of eye trauma hospitalizations.
Corresponding Author: Syed M. A. Shah, MD, University of Pittsburgh Medical Center, 203 Lothrop St, Pittsburgh, PA 15213 (email@example.com).
Accepted for Publication: July 24, 2018.
Published Online: October 4, 2018. doi:10.1001/jamaophthalmol.2018.4685
Author Contributions: Drs Iftikhar and Shah had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Iftikhar, Latif, Farid, Usmani, Shah.
Acquisition, analysis, or interpretation of data: Iftikhar, Farid, Canner, Shah.
Drafting of the manuscript: Iftikhar, Latif, Farid, Usmani, Shah.
Critical revision of the manuscript for important intellectual content: Iftikhar, Latif, Farid, Canner, Shah.
Statistical analysis: Iftikhar, Latif, Farid, Usmani.
Administrative, technical, or material support: Latif, Canner.
Supervision: Latif, Shah.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. No disclosures were reported.