Customize your JAMA Network experience by selecting one or more topics from the list below.
De Lott LB, Kerber KA, Lee PP, Brown DL, Burke JF. Diplopia-Related Ambulatory and Emergency Department Visits in the United States, 2003-2012. JAMA Ophthalmol. 2017;135(12):1339–1344. doi:10.1001/jamaophthalmol.2017.4508
How common are diplopia-related visits to ambulatory and emergency department settings in the United States?
In this prespecified secondary analysis of cross-sectional, population-based survey data collected from 2003 through 2012, there were 804 647 ambulatory and 49 790 emergency department visits annually for diplopia. Most ambulatory visits occurred with ophthalmologists (70.4%), and diplopia was frequently the primary reason for ambulatory setting (48.6%) or emergency department (36.5%) visits.
Diplopia-related visits are common in the United States, suggesting that future studies defining patient outcomes and association of practice patterns with outcomes should be largely aimed at practices providing eye care.
Diplopia is believed to be a common eye-related symptom. However, to date, there are no available population-based estimates, which are necessary to understand the impact of this disabling symptom on the health care system and to identify steps to optimize patient care.
To describe diplopia presentations in US ambulatory and emergency department (ED) settings.
Design, Setting, and Participants
Ambulatory and ED visits in the United States by patients with diplopia were analyzed in this prespecified secondary analysis of National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey data collected for a 10-year period (2003-2012). Data were analyzed for the present study from October 6, 2016 to August 18, 2017.
Main Outcomes and Measures
Numbers of ambulatory and ED diplopia presentations were estimated using weighted sample data. Weighted proportions of patient and clinician (ie, ophthalmologists, general practitioners, and specialty physicians) characteristics, diagnoses, and imaging use were calculated.
In total, 804 647 (95% CI, 662 075-947 218) ambulatory and 49 790 (95% CI, 38 318-61 262) diplopia-related ED visits occurred annually; 12.3% of ambulatory visits were primarily for acute- or subacute-onset diplopia. Mean (SD) patient age was 62.1 (20.3) years for ambulatory vs 48.1 (22.3) years for diplopia-related ED visits. Most visits primarily for diplopia were by patients 50 years or older (ambulatory, 79.1% [95% CI, 72.9%-84.2%]; ED, 51.8% [95% CI, 41.0%-62.4%]) who were white (ambulatory, 81.7% [95% CI, 74.8%-87.0%]; ED, 86.1% [95% CI, 77.8%-91.6%]) women (ambulatory, 51.1% [95% CI, 44.1-58.1]; ED, 52.8% [95% CI, 41.6%-63.7%]). Most diplopia-related ambulatory visits were conducted by ophthalmologists (70.4% [95% CI, 62.2%-77.5%]) even when symptoms were acute or subacute (89.0% [95% CI, 81.0%-93.9%]). The most common diagnosis in both settings was diplopia (International Classification of Diseases, Ninth Revision, Clinical Modification code 368.2). None of the 10 most frequent diagnoses was life threatening in the ambulatory setting, but approximately 16% of diplopia-related ED visits resulted in a stroke or transient ischemic attack diagnosis. Computed tomography or magnetic resonance imaging was ordered in 6.2% (95% CI, 2.8%-12.9%) of ambulatory and 59.7% (95% CI, 38.6%-77.7%) of ED visits, primarily for diplopia.
Conclusions and Relevance
Approximately 850 000 diplopia visits occur in the United States annually; 95% were outpatient visits, and diagnoses were rarely serious in the ambulatory setting but potentially life threatening in 16% of diplopia-related ED visits. Given the low probability of a serious neurologic diagnosis in the ambulatory setting and higher probability in an ED, future cohort studies are needed to define the association of various diagnostic practice patterns, such as imaging, with patient outcomes.
Diplopia is a symptom in which the patient sees 2 images instead of 1. It can arise from a single eye (monocular diplopia) or may occur when the eyes are misaligned (binocular diplopia). Compared with patients without diplopia, those with diplopia report more negative feelings about their overall quality of life and feelings of general disability.1,2 However, little information exists regarding the frequency of diplopia presentations and patterns of care for patients experiencing this disabling symptom. Prior studies of patients with diplopia have been narrowly focused on diplopia following a surgical procedure involving the eye or orbit,3-5 specific diplopia-related diagnoses,6-9 or diplopia presentations in ophthalmology clinics or ophthalmic emergency departments (EDs).10,11 Studying diplopia from a broader perspective is important to provide real-world information regarding the association of this symptom with health care use and the care patients receive.
The purpose of the present study was to ascertain information from nationally representative data regarding the presentation of diplopia in both ambulatory and ED settings across the United States. Specifically, we sought to determine the frequency of diplopia presentations in these settings and to describe common diagnoses, imaging tests, and patient presentation as well as clinician (ie, ophthalmologists, general practitioners, and specialty physicians) characteristics. By providing the scope of diplopia presentations in the United States and their influence on the US health care system, we aim to offer information that will help direct future research efforts and identify steps to optimize patient care.
Our study was a prespecified secondary analysis of deidentified, cross-sectional data collected for the National Ambulatory Medical Care Survey (NAMCS) and National Hospital Ambulatory Medical Care Survey (NHAMCS) in the 10-year period from 2003 to 2012.12 A 10-year period was used to maximize the accuracy and precision of the estimates. Both the NAMCS and NHAMCS are annual, nationally representative surveys conducted in the United States by the National Center for Health Statistics (NCHS), Centers for Disease Control and Prevention. The NAMCS uses a 3-stage sampling design (geographic region, physician practices stratified within specialties, and patient visits within practices) to enable a nationally representative characterization of nonfederal, outpatient, office-based patient care. The NHAMCS uses a 4-stage probability sampling design (geographic region, hospitals within a region, emergency service locations selected from these hospitals, and patient visits to these facilities) of nonfederal, general, and short-stay hospitals with EDs in the United States. The present study was deemed exempt from review by the University of Michigan Institutional Review Board, which also waived the need for informed patient consent.
The inclusion criterion was a diplopia-related visit in an ambulatory or ED setting. Diplopia-related visits were identified using the variable “patient’s reason for visit” (RFV) or the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code for “diplopia” (368.2). Each visit was allowed up to 3 RFVs and 3 ICD-9 diagnosis codes. The first RFV listed was identified as the patient’s chief concern. Patients were selected with an RFV of “double vision” (ie, diplopia) or a diagnosis of diplopia (368.2) in any of the 3 RFV or diagnosis fields. The diplopia RFV did not include abnormal eye movements (“crossed eyes” or “deviation”) or other types of visual dysfunction captured by the survey (eg, blurred vision, extraneous vision [spots or fuzzy vision, including “floaters”], and diminished vision [eg, weak eyes, scotoma]). These RFVs were not included in the diplopia definition because patients with abnormal eye movements may not have had diplopia, and the remaining RFVs were nonspecific. Instead, by identifying patients using the ICD-9-CM code for diplopia in addition to the diplopia-specific RFV code, we sought to capture patients who may have had difficulty describing diplopia or used other nonspecific terms to explain their symptoms (eg, “blurred vision”) but were ultimately determined by the clinician to have diplopia.
To determine the diagnoses that resulted from diplopia-related visits, we used the ICD-9-CM codes recorded for each visit. For each visit, up to 3 ICD-9-CM diagnoses are recorded, with the first diagnosis identified by the physician as the “primary diagnosis.”
Physician specialty was abstracted for outpatient visits only. Of note, nonphysician health care professionals (eg, optometrists) were excluded from the NAMCS.
Imaging data were abstracted to the survey instruments when specific boxes on the forms were selected for magnetic resonance imaging (MRI) or computed tomography (CT). However, the part of the body imaged was not specified. For ED visits, data were available from MRI and CT scans as a combined variable from 2003 through 2012 and as separate variables from 2005 through 2012. For ambulatory visits, separate MRI and CT variables were available from 2007 through 2012.
Data were analyzed for the present study from October 6, 2016 to August 18, 2017. The total number of outpatient and ED visits for diplopia were estimated with descriptive statistics using weighted survey data. Weighted proportions described age (<18, 18-49, or ≥50 years), sex, race, geographic region (physician practice location), and metropolitan statistical area ([MSA] for physician practice; eliminated from the NHAMCS in 2012). The imputed race variable calculated by the NCHS for each survey accounted for missing race data. Weighted proportions described accompanying RFVs, physician specialty (ambulatory visits only), imaging use, and physician diagnoses using ICD-9-CM codes. The association between demographic variables (continuous and categorical) and imaging use was assessed with bivariate logistic regression. Multiple logistic regression modeling tested the odds of whether physicians in certain specialties were more likely to order imaging after adjusting for patient age, sex, and race and the geographic area of the medical practice. Analyses were performed using the svy suite of survey data commands in Stata, version 13 (StataCorp LP).
From 2003 through 2012, there were approximately 850 000 diplopia-related visits annually in the United States: 804 647 (95% CI, 662 075-947 218) were ambulatory and 49 790 (95% CI, 38 318-61 262) occurred in the ED. Of the 804 647 diplopia-related ambulatory visits reported every year, approximately 12.3% occurred primarily for acute- or subacute-onset diplopia (<3 months as defined by the NAMCS, 99 058 visits [95% CI, 65 474-135 642]; data not available for ED visits). Diplopia was a patient-reported RFV in 83.6% (95% CI, 77.0%-88.6%) of all diplopia-related outpatient visits and 74.7% (95% CI, 63.3%-83.5%) of ED diplopia visits. Diplopia was identified as the patient’s primary RFV in 48.6% (95% CI, 42.0%-55.3%) of all diplopia-related outpatient visits and 36.5% (95% CI, 26.8%-47.5%) of diplopia-related ED visits. Diplopia was the only RFV in 23.8% (95% CI, 17.9%-30.9%) of diplopia-related ambulatory and 8.9% (95% CI, 5.2%-14.7%) of ED diplopia visits. Patient mean (SD) age was 62.1 (20.3) years for diplopia-related ambulatory visits and 48.1 (22.3) years for diplopia-related ED visits. Most visits primarily for diplopia were by patients 50 years or older (ambulatory, 79.1% [95% CI, 72.9%-84.2%]; ED, 51.8% [95% CI, 41.0%-62.4%]) who were white (ambulatory, 81.7% [95% CI, 74.8-87.0]; ED, 86.1% [95% CI, 77.8%-91.6%]) women (ambulatory, 51.1% [95% CI, 44.1%-58.1%]; ED, 52.8% [95% CI, 41.6%-63.7%]). Additional demographic characteristics are provided in Table 1.
For diplopia-related ambulatory visits, the 5 most common additional RFVs were diminished vision (17.0% [95% CI, 12.0%-23.7%]), eye examination (8.7% [95% CI, 5.2%-14.2%]), headache (6.3% [95% CI, 3.3%-11.7%]), visual dysfunction (5.8% [95% CI, 2.7%-12.1%]), and progress visit, not otherwise specified (5.4% [95% CI, 2.6%-10.7%]). For diplopia-related ED visits, the 5 most common RFVs were headache (20.1% [95% CI, 12.1%-31.4%]), vertigo or dizziness (15.6% [95% CI, 9.1%-25.3%]), diminished vision (10.8% [95% CI, 5.0%-21.5%]), nausea (6.5% [95% CI, 2.8%-14.3%]), and loss of feeling (6.0% [95% CI, 2.9%-12.0%]).
Most diplopia-related ambulatory visits occurred with ophthalmologists (70.4% [95% CI, 62.2%-77.5%]). The second most common specialty of the physicians conducting diplopia-related visits was general practitioner (14.5% [95% CI, 8.8%-23.1%]), which included general medicine, family practice, internal medicine, and pediatric physicians. Neurologists conducted 10.5% [95% CI, 7.5%-14.4%) of diplopia-related visits. All other specialty physicians rarely conducted diplopia-related ambulatory visits (4.6% [95% CI, 2.1%-9.5%]). Of the visits primarily for diplopia conducted by nonophthalmologists, 26.4% (95% CI, 13.2%-45.9%) resulted in a referral to another physician compared with 15.3% (95% CI, 8.0%-27.6%) for ophthalmologists. For patients with diplopia as the primary RFV, 58.7% (95% CI, 49.4%-67.4%) of the visits were for acute or subacute care, 89.0% (95% CI, 81.0%-93.9%) of which occurred in ophthalmology offices. By comparison, ophthalmologists were responsible for 5.7% (95% CI, 5.2%-6.2%), general practitioners for 50.6% (95% CI, 48.9%-52.3%), and neurologists for 1.5% (95% CI, 1.3%-1.6%) of all ambulatory visits in the United States during the study period, and diplopia was a listed RFV in 1.0% (95% CI, 0.9%-1.3%) of all ophthalmology and 0.02% (95% CI, 0.02%-0.03%) of all other physician visits.
The most common primary diagnosis recorded for diplopia-related visits when diplopia was the primary RFV was diplopia (ICD-9-CM code 368.2) in both ambulatory (16.2% [95% CI, 10.9-23.4]) and ED (15.2% [95% CI, 7.3%-29.0%]) settings. In 8.7% (95% CI, 5.2%-14.3%) of ambulatory and 10.4% (95% CI, 4.4%-22.7%) of ED visits primarily for diplopia, diplopia was the only diagnosis. Table 2 summarizes the top 10 primary ICD-9-CM diagnoses in both ambulatory and ED settings for patients presenting with diplopia as the primary RFV.13-15
An MRI or CT scan was ordered in an estimated 7.3% (95% CI, 3.7%-14.2%) of all diplopia-related ambulatory visits and 6.2% (95% CI, 2.8%-12.9%) of ambulatory visits primarily for diplopia. Age, race, geographic region, and MSA were not associated with having an imaging study ordered at the time of a diplopia-related visit. Imaging was ordered in 7.7% (95% CI, 3.0%-18.2%) of visits in which the primary RFV was acute or subacute diplopia. In the ambulatory setting, MRI was the most frequent imaging study ordered (81.4% [95% CI, 45.8%-95.8%] when diplopia was the primary RFV compared with 67.4% [95% CI, 27.5%-91.8%] when diplopia was just one of many reasons for the visit). Although ophthalmologists were responsible for most of the imaging studies ordered at diplopia-related visits (ophthalmologists, 58.7% [95% CI, 21.7%-87.9%] vs other physician specialties, 41.3% [95% CI, 12.1%-78.3%]) when diplopia was the primary RFV and it was acute or subacute, ophthalmologists had significantly lower odds (odds ratio, 0.15 [95% CI, 0.02-0.9]; P = .04, 2-sided [P < .05 considered statistically significant]) of ordering neuroimaging than all other types of physicians. For diplopia-related ED visits, CT or MRI was performed in 58.9% (95% CI, 47.6%-69.2%) of visits and 59.7% (95% CI, 38.6%-77.7%) of visits primarily for diplopia. Unlike in the ambulatory setting, most imaging procedures ordered in the ED setting were CT (78.2% [95% CI, 51.0%-92.5%]) when diplopia was the primary RFV, although 44.0% (95% CI, 25.7%-64.1%) of patients who presented with diplopia as the primary RFV underwent an MRI. As in the ambulatory setting, age, race, geographic region, and MSA were not associated with having an imaging study ordered at the time of the diplopia-related visit in the ED.
These prespecified secondary analyses of cross-sectional, population-based survey data provide nationally representative information regarding diplopia-related visits in the United States. From 2003 through 2012, diplopia-related visits constituted more than 850 000 visits per year, with approximately 95% of these occurring in an ambulatory setting. In nearly half of all diplopia-related ambulatory visits and one-third of all diplopia-related ED visits, patients identified diplopia as the primary reason for seeking care. Our study expands on previous work by considering diplopia-related visits for any reason to any physician in an ambulatory or ED setting across the United States to determine how health care is accessed and what diagnostic practices are used for this common symptom.
Most diplopia-related visits were to ophthalmologists, and ophthalmologists also had the largest proportion of their visits as diplopia related. Diplopia-related visits occurred in ophthalmology offices 70% of the time and nearly 90% of the time when diplopia was acute and the primary reason for seeking care. Therefore, it is imperative that all ophthalmologists are skilled in the evaluation of diplopia and knowledgeable regarding the broad differential diagnosis, which can include potentially life-threatening conditions, such as stroke or aneurysm, when diplopia is binocular. The assessment and evaluation of patients with binocular diplopia is a requirement for ophthalmology residents in the United States, but there are no studies that have assessed skill retention or clinician confidence in assessing patients with diplopia.16 This skill is particularly important because assessments of ocular motility and alignment are often not performed as part of the routine examination, potentially leading to more rapid skill degradation. Research aimed at understanding how clinicians assess patients with diplopia or at interventions to improve patient evaluations should be largely aimed at ophthalmology practices.
Overall, imaging studies were infrequently ordered during ambulatory visits for diplopia. Yet, ophthalmologists order more imaging studies than all other physicians, likely because ophthalmologists conducted the majority of diplopia-related visits in the United States. Compared with the other providers described in the present study, the odds of undergoing an imaging study were lower when a patient with acute- or subacute-onset diplopia presented to an ophthalmologist than to another physician. None of the top 10 diagnoses in the ambulatory setting that we identified when the visit was primarily for diplopia necessarily warranted imaging. Furthermore, diplopia was an isolated symptom in the ambulatory setting approximately 25% of the time and rarely had accompanying neurologic symptoms. Assuming high diagnostic accuracy, these findings suggest that the low imaging frequency is appropriate. A previous study assessing causes for diplopia in patients presenting to an ophthalmology practice found that potentially life-threating neurologic causes (stroke, neoplasm, or aneurysm) were infrequent, occurring in 13 of 226 cases (5.8%).11 The diplopia was described as nonisolated in all but 1 of the 13 cases (aneurysmal cranial nerve III palsy). An additional 3.1% of patients with diplopia had a diagnosis of multiple sclerosis (all recovered from diplopia without treatment).
We found a lower-than-expected frequency of imaging in ED visits that were primarily for diplopia. Although approximately 60% of all ED visits primarily for diplopia included CT, MRI, or both, patients in approximately 16% of visits primarily for diplopia received a serious neurologic diagnosis of ischemic stroke or transient ischemic attack. In contrast to the ambulatory setting, diplopia was an isolated symptom in less than 10% of the ED visits, and the most common accompanying symptoms were potentially neurologic, that is, headache, dizziness or vertigo, nausea, and loss of feeling. As in the ambulatory setting, there are few reported studies of diplopia presentations in the ED. A prospective study of patients with binocular diplopia in a general ED found that approximately 35% had a secondary cause (defined as stroke, aneurysm, tumor, multiple sclerosis, or other systemic diseases) to explain the diplopia. Stroke was most common, occurring in 14.6% of patients.17 Although our study was notably different because it was not limited to patients with binocular diplopia, we likewise found that a large percentage of patients with diplopia visiting the ED received a potentially disabling neurologic diagnosis. Two other studies of diplopia in an ED setting have reported that only 6 of 171 patients (3.5%)18 and 15 of 275 patients (5.5%)10 with diplopia have a life-threatening or neurologic cause (other than an isolated cranial nerve palsy); however, both of those studies were conducted at an eye hospital dedicated to ophthalmologic emergencies. The lower-than-expected use of imaging in the ED found in our study might be explained by a stroke or transient ischemic attack diagnosis related to a history rather than an acute event, if a nonspecific visual symptom (eg, “trouble seeing”) was reported by a patient as double vision, or if a symptom was mistakenly recorded as diplopia.
Because we did not assess long-term data for each diplopia-related visit and could not ensure diagnosis accuracy, our data should not be directly used to make imaging recommendations. Instead, our data provide prior probabilities of encountering potentially serious diagnoses in specific clinical settings and therefore inform medical decision making. For example, stroke is a rare diagnosis in the ambulatory setting in a patient with diplopia. Because the probability of encountering stroke is very low, obtaining an MRI or CT scan to rule out stroke should not be routinely performed unless guided by a specific examination finding or accompanied by other neurologic signs or symptoms. In cases of uncertainty, consultation with a neurologist or neuro-ophthalmologist is likely more appropriate. Therefore, our findings support the recommendation by the American Academy of Pediatric Ophthalmology and Strabismus Choosing Wisely that, because most causes of diplopia do not have a neurologic basis, patients should not undergo routine imaging prior to an eye examination.19 We add the caveat that this recommendation is most appropriate for diplopia in the outpatient setting and only when unaccompanied by neurologic signs or symptoms. In contrast to the ambulatory setting, at least 1 in 6 patients presenting to the ED primarily for diplopia received a potentially life-threatening or urgent neurologic diagnosis, and diplopia was rarely isolated. Therefore, our study results support the practice of having a low threshold for ordering neuroimaging in patients with binocular diplopia presenting to the ED. This information may be helpful for ophthalmologists who provide remote or telephone consultations for EDs. Furthermore, incorporating this type of information into a robust diplopia decision support tool may assist nonexperts, both eye care and non–eye care physicians, in making appropriate triage and imaging decisions.
There are several important limitations to our study. The RFV was first recorded in the survey form using the patient’s words but was then coded in the data set using established and rigorous methodology from the NCHS.12,20 However, errors in data abstraction may have led to inaccurate coding. Similarly, we could not ensure the ICD-9-CM coding accuracy. It is likely for 2 reasons that our results underestimated the number of diplopia-related visits in the United States. First, visits to optometrists were not included in the NAMCS data set. Second, we defined patients with diplopia as those with an RFV of diplopia or an ICD-9-CM diagnosis of diplopia, but patients may have experienced eye misalignment as visual blur rather than frank diplopia, and diplopia may not have been coded by the clinician if a more specific diagnosis was identified. In addition, given the limited data each year, only aggregate estimates could be provided and trends over time could not be determined. Because we did not have access to the full case record for each visit, we could not comment on the quality of care delivered and could only speculate about how it might have influenced patient outcomes. Furthermore, when imaging was ordered, we could not be certain the study was of the head rather than another part of the body.
More than 850 000 diplopia-related visits occurred annually in the United States from 2003 through 2012, and most were in ophthalmology practices even when symptoms were acute or subacute. Diagnoses were rarely life threatening in the ambulatory setting but were potentially life threatening in 16% of diplopia-related ED visits. Future cohort studies are needed to define the association of practice patterns, such as urgent or delayed imaging, with outcomes. Because most patients with diplopia were examined by ophthalmologists, studies or interventions with the goals of optimizing the evaluation and management of diplopia should be largely aimed at practices providing eye care.
Corresponding Author: Lindsey B. De Lott, MD, MS, Department of Ophthalmology and Visual Sciences, University of Michigan, 1000 Wall St, Ann Arbor, MI 48105 (email@example.com).
Accepted for Publication: September 11, 2017.
Published Online: October 26, 2017. doi:10.1001/jamaophthalmol.2017.4508
Author Contributions: Dr De Lott had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: De Lott, Kerber, Brown, Burke.
Acquisition, analysis, or interpretation of data: De Lott, Kerber, Lee.
Drafting of the manuscript: De Lott.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: De Lott, Kerber, Lee.
Study supervision: Burke.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Kerber reported receiving grant support outside of the submitted work from the National Institutes of Health (NIH). Dr Lee reported serving as a paid consultant for the Alcon Research Institute and the Centers for Disease Control and Prevention, receiving grants from the Kellogg Foundation and Research to Prevent Blindness, and owning stock in GlaxoSmithKline, Medco Health Solutions, Merck, and Pfizer. Dr Brown reported receiving grant support and funding from the NIH, travel reimbursement from the American Academy of Sleep Medicine, and research support from the Blue Cross Blue Shield of Michigan Foundation, the Michigan Department of Community Health, and the University of Michigan for stroke-related research. Dr Burke reported receiving grant support from the NIH. No other disclosures were reported.
Funding/Support: Dr De Lott is supported by a Clinical Scientist Institutional Career Development Award (K12EY022299) from the National Eye Institute and was previously funded by Ruth L. Kirschstein Institutional National Research Service Awards (T32 NS007222-32 and 33) from the National Institute of Neurologic Disorders and Stroke (NINDS). Dr Burke is supported by a Mentored Clinical Scientist Research Career Development Award (K08NS082597) from the NINDS.
Role of the Funder/Support: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Create a personal account or sign in to: