Nash EA, Margo CE. Patterns of Emergency Department Visits for Disorders of the Eye and Ocular Adnexa. Arch Ophthalmol. 1998;116(9):1222-1226. doi:10.1001/archopht.116.9.1222
To characterize the magnitude and patterns of visits to the emergency department (ED) for problems related to the eye and ocular adnexa.
The National Hospital Ambulatory Medical Care Survey was used to obtain information on ED visits in the United States for conditions of the eye and ocular adnexa in 1993. Patients were identified by International Classification of Diseases, Ninth Revision, Clinical Modification, codes. National projections were based on a staged probability design.
There were 2.32 million projected ED visits for problems of the eye and ocular adnexa in 1993. Forty-nine percent of visits were for injuries, two thirds of which occurred in males. Thirty-five percent of injuries occurred in the home and 18% occurred in the workplace. Only 3% of patients required hospitalization. Most patients had private insurance, but substantial variations in coverage existed for patients who used the ED for injury- vs non–injury-related care.
Emergency departments in the United States provide a large amount of eye care, much of which is for conditions other than trauma. Differences in insurance coverage for injury- and non–injury-related eye care indicate that factors other than medical urgency are involved in the decision to use ED services. Further studies are needed to determine the cost-effectiveness and quality of ocular-related ED visits.
THERE IS limited information on the role emergency departments (EDs) play in the provision of eye care in the United States. Most epidemiological studies of ocular-related ED visits concentrate on ocular injury.1- 5 Severe ocular trauma has a strong correlation with poor visual outcome, and any meaningful approach to the prevention of accidental injury must be based on an understanding of its pattern of occurrence. Eye care in the ED, however, involves much more than ocular trauma. We used the National Hospital Ambulatory Medical Care Survey (NHAMCS) database to quantify and characterize the spectrum and magnitude of ED visits for eye problems in the United States.
The NHAMCS was conducted by the Ambulatory Care Statistics Branch of the National Center for Health Statistics, Centers for Disease Control and Prevention, Atlanta, Ga. The 1993 database was obtained from the Department of Health and Human Services, National Center for Health Statistics, Hyattsville, Md. A detailed description of the study design and of the methods of analysis is available elsewhere.6,7 Briefly, the survey contains information collected by the NHAMCS from December 1992 through December 1993. The data collected represent a national probability sample of visits to the EDs and outpatient departments of noninstitutional general and acute care hospitals in the 50 states and the District of Columbia. The survey does not include federal, military, and Department of Veterans Affairs hospitals. The NHAMCS used a staged probability design with 4 levels of sampling: primary sampling units (ie, a county or a group of counties), hospitals within primary sampling units, clinics within hospitals, and patient visits within clinics. The primary sampling unit sample was selected from approximately 1900 geographically defined units. The primary sampling units were stratified according to socioeconomic variables and were selected with a probability proportional to their size. Eligible hospitals included general medical, surgical, and children's hospitals listed in the 1991 SMG Marketing Group Inc Hospital Market Database. Of the 5582 eligible hospitals, 89% had EDs and 90% had outpatient departments. Hospitals were stratified by region, size, ownership, and hospital class (eg, presence or absence of an ED and an outpatient department). Hospitals were selected and randomly divided into 16 subsets of nearly equal size. Each subset was assigned a 4-week reporting period for 1 year beginning December 28, 1992, and ending December 26, 1993. A total of 445 hospitals were eligible to participate in the survey, and 395 EDs provided data.
The data collection procedure began with extensive field training and a hospital induction period before actual data collection took place. Field representatives trained hospital staff to correctly complete the Patient Record form. Place of injury was part of the record form. Physician judgment was used to determine whether the patient's principal presenting problem was alcohol related, drug related, both, neither, or unknown. Field quality control was performed throughout the study. Field representatives reviewed hospital records to determine if cases were missed and edited completed forms for missing data. Attempts were made to retrieve all missing data. Identifying information, eg, names, was not removed from the hospital records. Assurance of confidentiality was provided to hospitals (section 308[d] of the Public Health Service Act).
Visits to the ED because of injury to the eye and ocular adnexa were identified from International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes (Table 1).8 The principal diagnosis reflects the best judgment of the physician at the visit and represents the reason the patient sought ED care. Up to 3 diagnoses could be listed and up to 3 external causes of injury could be coded according to the ICD-9-CM system for supplementary classification of external causes of injury and poisoning. An eye and ocular adnexal injury code was included if it was among the 3 reasons for the ED visit. The place of injury was obtained from the patient record. Emergency department visits for eye and ocular adnexal problems unrelated to injury were identified by ICD-9-CM diagnostic codes 360.0 though 379.9. Several unique identifiers were created for this survey to identify such situations as an illegible written diagnosis, a patient leaving before being seen, or a patient being transferred to another facility.
The probability sample design of the NHAMCS allows the sample to be weighted and used to calculate a national estimate. Details concerning the statistical methods used to weight data are available elsewhere.6 The relative SE (RSE) of a projected estimate is obtained by dividing the SE by the estimate itself. The RSE is expressed as a percentage. Estimates with RSEs in excess of 30% are not considered reliable. Annual incidences are calculated using 1990 US census data extrapolated to the study year—1993—and are expressed as the number of patient visits per 100000 population.
A single source of expected payment was identified for each patient visit. In persons with 2 sources of coverage (eg, Medicare and commercial supplemental health insurance), only the primary source was used for analytic purposes.
Information on other aspects of the 1993 NHAMCS has been published elsewhere.9,10
Seven hundred seventy-two visits for problems related to the eye and ocular adnexa were identified in the 1993 NHAMCS, which projects to 2323695 annual ED visits (RSE=7.6%). There were 1140265 visits (RSE=8.6%) for injury to the eye and ocular adnexa, which represents 1.3% of all ED visits for that year; 40.3% of patients with injury-related visits were between the ages of 25 and 44 years (Table 2), and 63.1% of injury-related visits were made by males (Table 3). The annual rate of ED visits for ocular and adnexal injury in males was 581.4 per 100000 population and in females was 322.0 per 100000 population. Males sustained more injuries in all diagnostic categories. Superficial injury of the eye and adnexa was the most common diagnostic category, with 597396 annual visits (RSE=18.4%). Annual visits exceeded 100000 for 3 other diagnostic categories: foreign body on the external eye, contusion of the eye and adnexa, and open wound of the adnexa (Table 3). Of all injuries, 34.7% occurred at home, 17.9% took place at work, and 33.4% happened in unspecified and unknown locations (Table 4). Six percent of injuries were related to or were associated with alcohol or drug use. The rate of ED visits for injury was 16.3% higher for blacks than whites (532.3 vs 445.8 per 100000 population; Table 5). There were 34200 projected hospital admissions (RSE=38.6%) for eye and ocular adnexal injury, which was 3% of eye injury visits. The sample size for hospital admissions was too small for analysis of diagnostic categories and patient characteistics.
The source (ie, cause or circumstance) of injury varied according to sex (Table 6) and where the injury took place (Table 7). Males were more likely than females to be injured by a foreign body, in a fight, in a motor vehicle accident, and while playing a sport. Females were injured more often than males by falls and from assaults (Table 6). Foreign body injuries accounted for 66.8% of injuries at work and 30.9% of injuries at home. Of all school-related injuries, 42.2% were caused by accidentally being struck by an object or person and 16.8% were related to sports.
There were 1183430 non–injury-related visits for conditions of the eye and ocular adnexa (RSE=8.6%). More than 900000 of these were for inflammatory conditions of the conjunctiva and eyelids, with more than a half million for unspecified conjunctivitis alone (Table 8). There were 11125 hospital admissions for non–injury-related disorders, which amounts to 0.9% of all non–injury-related eye visits. The rate of visit by age strata showed that ED use for non–injury-related causes was greatest for persons younger than 15 years (Table 2). Blacks used the ED more often than whites for non–injury-related conditions (1017.1 vs 385.7 per 100000 population) (Table 5).
Analysis of the sources of payment for ocular-related ED visits revealed that private insurance companies were the largest source of payment, but there were substantial differences in coverage between injury- and non–injury-related visits (Table 9). Forty-two percent of persons with injuries had private insurance compared with 34.5% of those with non–injury-related visits. The greatest difference was found among Medicaid beneficiaries, who showed nearly a 3-fold increase in visits for non–injury-related care compared with injury-related care (26.3% vs 9.2%) (Table 9).
Emergency departments play a major role in our ambulatory health care system. In 1996, there were an estimated 90.3 million visits to hospital EDs, or about 34 visits per 100 persons in the United States.11 Thirteen percent of all ED visits in 1993 resulted in hospital admission.10 Nonurgent problems, however, account for as much as 40% to 55% of all ED visits.12,13 Emergency departments have experienced a 22% increase in annual visits during the past decade, with the greatest increase in use by the elderly and the poor.14 A disproportionate amount of ED visits are made by persons who are uninsured or who have Medicaid or Medicare coverage.15- 17 There are no standards describing the appropriate use of an ED. The subject of ED use is complex and controversial. Use is related to many variables, including patient socioeconomic status, perceived access to primary care, and perceptions of quality and convenience of ED care.18,19 The 1993 NHAMCS is a large, prospective, population-based study. The sampling method was strategically designed to give national and regional projections for annual use of ambulatory care in the United States. Data collection was carefully enacted and frequently monitored. The national estimates are considered reliable predictions when the RSE is 30% or less.6 The SE is a measure of the sampling variability that occurs by chance. The chances are 5 of 100 that an estimate from the NHAMCS sample differs from the value that would be obtained from a complete census when ED projections are greater than 57382 visits per 100000 population.6 Because of its careful design and planning, the NHAMCS is a valuable database for national ED eye care information.
The NHAMCS estimate of 2.32 million ocular-related ED visits in 1993 corresponds to 911 visits per 100000 persons. Forty-nine percent of these visits were for injuries, and only 3% required hospitalization. Eye and ocular adnexal injuries accounted for 1.3% of all ED visits in 1993. Minor ocular injury occurs frequently and has been estimated to compose up to 5% of some ophthalmic practices.20 In Wisconsin, there were 4.2 hospital-treated eye injuries for every 1000 population,2 which is close to the NHAMCS rate of 4.5 per 1000 population. The rate of minor ocular injury in New England, when measured by a telephone survey, was nearly 10.0 per 1000 adults.21
From a public health perspective, it makes more sense to study the sources of serious ocular trauma because these injuries have the greatest likelihood of causing permanent vision loss. The National Health Interview Survey estimated that ocular trauma may be responsible for nearly 5% of all visual impairment.22 The NHAMCS gives limited information on the sources of serious ocular trauma because of limitations of ICD-9-CM coding. Hospital-based epidemiological studies that use admission or discharge records are designed to identify serious injuries. Although admission diagnoses are usually verified by ophthalmologists in these studies, they are subject to bias caused by local and regional referral patterns. Ocular trauma was responsible for 13.2 hospital admissions per 100000 population in Maryland from 1979 through 1988.4 The national discharge rate for ocular trauma based on hospital discharge abstracts from 1984 through 1987 showed a rate of 13.2 per 100000 population.3 These figures are in close agreement with the NHAMCS hospital admission rate of 13.4 per 100000 population, which suggests that there has been no appreciable decline in the rate of hospitalization for ocular and adnexal injuries since the late 1970s.3,4 The sample size for hospital admission for eye-related causes in the NHAMCS, however, was small (RSE=38.6%) and must be interpreted with caution.
Six percent of ocular injuries were alcohol or drug related. It has been recognized that the method used to ascertain whether the ED visit was alcohol or drug related in this study underestimates the relationship between alcohol and drug use and injury.10
There are several reasons that the total burden of ocular and adnexal injuries is underestimated in this survey. First, EDs in eye hospitals and other specialty hospitals were not included. Although there are few EDs dedicated exclusively to eye care, these centers may take care of a disproportionate number of eye emergencies in some areas. Second, only 3 diagnoses per patient were coded in this study. It is possible that some ocular injuries went unrecognized in patients with multiple injuries because they were not included among the top 3 diagnoses.
The fiscal impact of 2.32 million annual ED visits for ocular-related injuries is substantial. Federal and state sources were primarily responsible for payment of nearly 20% of these visits. Although the proportion of persons with private insurance who sought ED care for noninjuries was 7% less than that for injuries, the opposite trend was observed for persons with Medicaid and Medicare. The increase in non–injury-related visits was particularly striking for Medicaid beneficiaries, whose proportional use of services increased 285%. Increased use of the ED by Medicaid enrollees has been shown to be inversely related to access to primary care physicians, but other reasons for ED use by Medicaid beneficiaries exist.23- 27 The inappropriate use of the ED has been targeted as an important reason for the escalating cost of health care, but the controversy of cost-effective ED use is more complex than this single issue suggests. The ED may be a more cost-effective source for some types of nonurgent care than other ambulatory care settings.24,26
Emergency departments provide a large amount of ambulatory eye care each year in the United States. Although ED visits for nonurgent care may be more costly than comparable care delivered in other ambulatory settings, this observation may not be generalizable to ocular-related ED visits. More detailed studies concerning the cost-effectiveness and the quality of ED care for ocular injuries are needed.
Accepted for publication June 17, 1998.
Corresponding author: Curtis E. Margo, MD, MPH, Department of Ophthalmology, Watson Clinic, 1600 Lakeland Hills Blvd, Lakeland, FL 33805.