The figure displays the incidence rates per 100 000 person-years for the current study, as well as prior US population-based studies. In this study, there was a significant increase in incidence rates with older age (P < .001). The Northern California Epidemiology of Uveitis Study found a similar significant increase in incidence with age (P < .001), while the Rochester, Minnesota, study (P = .30) and Pacific Northwest Veterans Affairs Study (P = .61) did not.
eTable 1. International Classification of Diseases, Ninth Revision (ICD-9) Diagnosis Codes Used for Initial Identification of Uveitis Cases
eTable 2. Reasons for Exclusion From the Study, Including All Records Identified Between January 1, 2006, and December 31, 2007
eTable 3. Comparison of Incidence Rate Estimates from US Population–Based Uveitis Epidemiology Studies by Age and Sex Subgroups
eTable 4. Comparison of Annual Active Prevalence Rate Estimates from US Population–Based Uveitis Epidemiology Studies by Age and Sex Subgroups
Acharya NR, Tham VM, Esterberg E, Borkar DS, Parker JV, Vinoya AC, Uchida A. Incidence and Prevalence of UveitisResults From the Pacific Ocular Inflammation Study. JAMA Ophthalmol. 2013;131(11):1405-1412. doi:10.1001/jamaophthalmol.2013.4237
Uveitis is responsible for a significant proportion of legal blindness in the United States. Currently, there are few population-based reports characterizing the epidemiology of uveitis.
To ascertain the incidence and prevalence of uveitis in a Hawaiian population and compare these estimates with those from prior population-based studies.
Retrospective, population-based cohort study conducted from January 1, 2006, to December 31, 2007.
Kaiser Permanente Hawaii, a multispecialty managed care organization serving approximately 15% of the general Hawaiian population with locations throughout the Hawaiian islands.
All patients enrolled in the Kaiser Permanente Hawaii health plan during the study (N = 217 061).
Main Outcomes and Measures
Clinical diagnosis of uveitis, either incident or prevalent, during the study determined by an initial search of the electronic medical record database of Kaiser Permanente Hawaii for uveitis-associated International Classification of Diseases, Ninth Revision diagnosis codes and subsequently confirmed through individual record review by a uveitis specialist.
Of 217 061 eligible patients, 872 were identified using International Classification of Diseases, Ninth Revision codes and 224 cases of uveitis were confirmed. The overall uveitis incidence rate was 24.9 cases per 100 000 person-years. The annual prevalence rates for 2006 and 2007 were 57.5 and 58.0 per 100 000 persons, respectively. No difference in incidence rate was found by sex (P = .63), but female patients had a higher prevalence (P = .008). Incidence and prevalence increased with older age (P < .001 for incidence and prevalence). Pacific Islanders had a lower prevalence rate than non–Pacific Islanders (2006: P = .09, 2007: P = .04), while white individuals had a higher prevalence rate than nonwhite individuals (2006: P = .07, 2007: P = .01).
Conclusions and Relevance
The incidence and prevalence of uveitis in this population were much lower than in the Northern California Epidemiology of Uveitis Study, but similar to the Northwest Veterans Affairs Study. The results of this study highlight incidence and prevalence estimates in a new population and provide novel comparisons by race. These differences by race raise questions regarding the effects of genetic and environmental influences on the pathophysiology of uveitis.
Uveitis is a set of conditions defined by intraocular inflammation and is believed to be the cause of up to 10% of legal blindness in the United States, or approximately 30 000 new cases of blindness per year.1- 4 In contrast to common age-related eye disorders, uveitis may have a strong socioeconomic impact because it often affects younger working-age patients.4 Despite the clear public health implications, to our knowledge, there are few population-based reports characterizing the epidemiology of uveitis.2,5- 10 The complexity of uveitis and its treatment has led to a predominance of published reports stemming from subspecialty tertiary care centers rather than population-based investigations. Experience in subspecialty clinics is not thought to be reflective of disease characteristics in the population owing to referral bias.11
To date, only 3 epidemiologic studies have been performed in the United States to describe the incidence and prevalence of verified uveitis cases within a general population. However, their respective estimates vary widely, ranging from 17.4 to 52.4 cases per 100 000 person-years for incidence and 69.0 to 114.5 per 100 000 persons for annual prevalence.2,5,6 Additionally, these estimates were derived from different populations and at different times.
The Hawaii region of the Kaiser Permanente (KP) health maintenance organization serves a racially diverse population, representing all age groups and more than 15% of the total population of Hawaii. The KP health care system is ideal for population-based studies since patients generally receive all aspects of their medical care within the KP network, and all patient information is recorded in an accessible electronic medical records system. The Pacific Ocular Inflammation Study is interested in the epidemiology of ocular inflammatory disease in the Hawaiian islands. This specific study aimed to estimate the incidence and prevalence of uveitis, with a focus on age, sex, race, and anatomic location subgroup comparisons, and to compare these estimates with those from prior studies.
Institutional review board approval for this study was obtained at the University of California, San Francisco, and KP Hawaii. The clinical database of KP Hawaii was searched for any visit from January 1, 2006, to December 31, 2007, that referenced an International Classification of Diseases, Ninth Revision (ICD-9) code that might be associated with a diagnosis of uveitis (eTable 1 in Supplement). The inclusive list of diagnosis codes used for this study was adapted from 2 prior uveitis incidence and prevalence studies.5,6 This list includes a broad range of codes that are not all uveitis specific to ensure that potential uveitis cases were not missed. The medical records flagged by this search were then individually reviewed to eliminate patients without a confirmed diagnosis of uveitis and to determine disease status.
All medical records were reviewed by a uveitis fellowship-trained ophthalmologist (N.R.A.). Standardization of Uveitis Nomenclature criteria were used to classify the confirmed cases according to anatomic location of inflammation and disease course using information from the entire medical record. All cases were classified into 1 of 4 categories for the year 2006 and for the year 2007: incident, active prevalent, inactive prevalent, or no visit. Incident cases were defined as patients whose first diagnosis of uveitis occurred during that year. Active prevalent cases were those who had active uveitis or were undergoing treatment for uveitis during that year, but whose first diagnosis of uveitis occurred in a previous year. Incident and prevalent cases were combined for prevalence calculations. Inactive prevalent cases, which were included in secondary analyses, were defined as those who had a visit to an ophthalmologist or optometrist during that year, but who did not have active inflammation and were not receiving treatment. Patients who were prevalent or inactive prevalent in 2007 but did not have any visits in 2006 were recorded as having no visits and were not included in any calculations for 2006.
Incidence was calculated per 100 000 person-years by dividing the confirmed incident cases by the total person-years of follow-up and multiplying by 100 000. Owing to constant fluctuations in KP membership, the total person-years of follow-up accumulated during each year of the study were calculated using quarterly membership data. Annual prevalence per 100 000 persons was calculated by dividing total incident and prevalent cases by the total KP membership at the midpoint of each year (June 30, 2006, and June 30, 2007) and multiplying by 100 000. Ninety-five percent probabilistic confidence intervals were calculated assuming a Poisson distribution.12
Incidence rates were reported for the overall study period (January 1, 2006, to December 31, 2007). Incident cases can only be counted once so data were combined over the 2-year period to allow for tighter point estimates. However, prevalence rates were calculated by year because the status could change each year (ie, changing from active in 2006 to inactive in 2007). Rates were reported for the entire population and stratified by age, sex, and anatomic location of uveitis. Age was divided into the following categories to allow comparability with prior uveitis studies: 0 to 14 years, 15 to 24 years, 25 to 44 years, 45 to 64 years, and 65 years or greater.5,6 Total person-years or persons for incidence and prevalence estimates reported by age and sex were calculated using the same method as the overall rate estimates but using the membership data from within that age and/or sex category.
Additional data, including systemic diagnoses, were collected electronically from the medical records of all patients with uveitis. Self-reported race was collected for both cases and the entire KP population. Estimates of incidence and prevalence by race were calculated using person-years and midpoint populations adjusted to reflect the distribution of available race information.
Demographic data were compared between uveitis cases and the general Kaiser population using Fisher exact test. Comparisons of age-specific, sex-specific, race-specific, and anatomical location–specific incidence and prevalence rates and comparisons with prior studies were made using Poisson regression. All analyses were performed using R statistical software (The R Foundation for Statistical Computing) and Stata version 11.0 (StataCorp).
A person-time estimate of 433 794.2 person-years was obtained for the study period. The midpoint populations for 2006 and 2007 were 219 262 and 215 669 people, respectively. From this population, 873 patient records were selected on the basis of an ICD-9 diagnosis code search (eTable 1 in Supplement). Of these, 224 patients were confirmed uveitis cases based on individual medical record review. Reasons for exclusion of cases are outlined in eTable 2 (Supplement).
Demographic characteristics between patients with uveitis and the general KP Hawaii population were compared (Table 1). Notably, patients with uveitis were significantly older than the general Kaiser population. Additionally, the racial distribution of the 2 groups differed significantly. White individuals made up a significantly higher proportion of uveitis cases compared with the general population (P = .002, Table 1), while Pacific Islanders made up a lower percentage of cases compared with the general KP Hawaii membership (P = .03, Table 1).
Table 2 describes the clinical characteristics of the 224 patients with uveitis. The most common anatomic location of inflammation was anterior (Table 2). Forty-three percent of patients had an acute disease course. Thirty-seven percent of patients had an associated diagnosis, while 8% had an eye-limited disease. The most common associated diagnosis was herpes infection.
There were 108 incident cases of uveitis during the 2-year study, resulting in an overall incidence rate of 24.9 per 100 000 person-years (95% CI, 20.4-30.1; Table 3). Anterior uveitis was the most common, with an incidence of 20.3 per 100 000 person-years (95% CI, 16.3-25; Table 3), followed by posterior/panuveitis (3.9 per 100 000 person-years; 95% CI, 2.3-6.3; Table 3), and intermediate uveitis (0.7 per 100 000 person-years; 95% CI, 0.1-2.0; Table 3). Table 3 also reports incidence rates by age and sex subgroup. There was no significant difference between the overall incidence rates for males and females (P = .63). However, incidence rates increased significantly with age (P < .001). When further broken down by anatomic location of uveitis, rates of anterior uveitis increased with age (P < .001) but rates of intermediate (P = .41) and posterior/panuveitis (P = .96) did not.
Detailed breakdowns of annual active prevalence by age, sex, and location of inflammation are displayed in Table 4. The annual active prevalence was similar between 2006 and 2007, with overall rates of 57.5 per 100 000 persons (95% CI, 47.9-68.4; Table 4) and 58.0 per 100 000 persons (95% CI, 48.2-69.1; Table 4). Active prevalence was higher among women (2006: P = .14, 2007: P = .02; Table 4) and also increased with age (2006: P < .001, 2007: P < .001; Table 4). Similar to trends in incidence rates, the active prevalence rate of anterior uveitis increased significantly with age (2006: P < .001, 2007: P < .001), but rates of intermediate (2006: P = .07, 2007: P = .25) and posterior/panuveitis (2006: P = .11, 2007: P = .15) did not.
Including inactive prevalent cases increased the overall prevalence rate to 70.7 per 100 000 persons in 2006 (95% CI, 60.0-82.7; Table 5) and 81.6 per 100 000 persons in 2007 (95% CI; 70.0-94.6; Table 5). Consistent with trends in active prevalence rates, rates were higher among women (2006: P = .04, 2007: P = .003; Table 5) and increased with age (2006: P < .001, 2007: P < .001; Table 5).
In addition to age, sex, and location of inflammation, comparisons of incidence and prevalence rates were made by racial subgroup (Table 6). Notably, Pacific Islanders had the lowest incidence of uveitis at 20.1 cases per 100 000 person-years (95% CI, 13.1-29.3); however, this was not significantly lower than the incidence rate for the remainder of the population (P = .23). Similarly, the subgroup of black patients had the highest incidence of uveitis at 39.6 cases per 100 000 person-years (95% CI, 9.3-108.5), but this was not significantly higher than the remainder of the population (P = .45). There were no significant differences in incidence rates for any other racial subgroups.
While there was no significant difference in incidence rates for Pacific Islanders compared with non–Pacific Islanders, there was a trend toward a lower prevalence rate for Pacific Islanders in 2006 and a significantly lower prevalence in 2007. This was true for both annual active prevalence (2006: P = .09, 2007: P = .04), as well as for active and inactive prevalence (2006: P = .09, 2007: P = .01). Additionally, the white subgroup had a higher active prevalence (2006: P = .07, 2007: P = .01) and active and inactive prevalence (2006: P = .04, 2007: P < .001) than the remainder of the population. There were no significant differences in prevalence rates for any other racial subgroups.
The overall incidence of uveitis in this study of 24.9 cases per 100 000 person-years was not significantly different than the incidence rate of 25.6 cases per 100 000 person-years found in the Northwest Veterans Affairs (NWVA) Study (P = .88).6 However, our estimate was significantly higher than the Rochester, Minnesota, study estimate of 17.4 cases per 100 000 person-years (P = .03) and significantly lower than the Northern California Epidemiology of Uveitis (NCEU) Study estimate of 52.4 cases per 100 000 person-years (P < .001).2,5
Similar results were found when comparing prevalence rates. The active prevalence rates in our population for both 2006 and 2007 were not significantly different than the prevalence estimate of 69.0 per 100 000 persons in the NWVA Study (2006: P = .17, 2007: P = .19).6 Compared with the prevalence estimate of 115.3 per 100 000 persons in the NCEU Study, our estimates were significantly lower (2006: P < .001, 2007: P < .001). Specific incidence and annual active prevalence estimates by age subgroup for each study are displayed in the Figure. In Supplement, eTables 3 and 4 display incidence and active prevalence, respectively, by age and sex for the current study and prior United States–based studies.
The overall uveitis incidence rate of 24.9 per 100 000 person-years found in this population was comparable with that found by the NWVA Study, but approximately half of that found by the NCEU Study.5,6 Additionally, our result was significantly higher than the incidence rate found in the Rochester, Minnesota, study more than 50 years ago.2 The use of annual prevalence periods allowed us to calculate prevalence estimates that were directly comparable with both the NCEU and NWVA studies. Our estimates of 57.5 and 58.0 per 100 000 persons were significantly lower than the estimated prevalence by the NCEU Study, but comparable with the estimate reported in the NWVA Study.
Differences in our population and previously studied populations may explain the variation in results. Although both our study and the NCEU Study were conducted using a KP population, it is possible that more patients were seeking care outside of KP Hawaii, resulting in missed uveitis cases. However, only approximately 5% of the KP Hawaii population has dual insurance that would facilitate receiving health care outside of Kaiser. Additionally, the racial demographics of our population and the NCEU Study population varied considerably. Black individuals made up a significant portion of the NCEU Study population, while only 2% of our population was black. Notably, prior international studies have found higher incidence and prevalence rates of uveitis in populations with higher proportions of Africans or South Asians.7,10 Similarly, a recent population-based study conducted in Taiwan found overall incidence and prevalence rates approximately 5 times the estimates obtained in our study.13 Asian Americans comprised 40% of our population; however, the incidence and prevalence estimates in this racial subgroup were comparable with the overall estimates in our study. Although the absence of medical record review to adjudicate uveitis cases could have led to overestimation of incidence and prevalence in this recent study from Taiwan, the higher estimates in this Chinese population suggest that environmental influences could also play a role in the development and progression of uveitis.
While our overall estimates were significantly different, trends in incidence and prevalence rates based on age and sex stratification closely matched those found in the NCEU Study. Incidence was not significantly different by sex, but increased with age. Further subanalysis by anatomic location revealed that only incidence of anterior uveitis increased with age, while intermediate and posterior/panuveitis did not show this trend. One prior US study in the elderly also found a high incidence of uveitis, particularly anterior uveitis, in patients 65 years of age and older, suggesting a high disease burden in this subgroup of the population.14 Prevalence rates showed similar significant trends with respect to increasing age, but there was also a significant difference by sex, with women having a higher prevalence rate than men. This could indicate that women develop more chronic or recurrent uveitis than men, and thus are requiring ongoing treatment and monitoring. The higher prevalence found in women is consistent with sex differences found in many systemic chronic autoimmune and inflammatory conditions such as multiple sclerosis, rheumatoid arthritis, and systemic lupus erythematosus.15- 17
Notably, the NWVA Study did not find this same difference by sex and only found a trend toward increasing prevalence with older age. However, the NWVA Study was not equipped to ascertain differences by sex since women made up less than 10% of the overall population. Additionally, there were very few people in the younger 2 age groups in the VA population used in the NWVA Study, making it harder to detect differences by age.
To our knowledge, this is the first US-based study that characterizes incidence and prevalence by race. Pacific Islanders in KP Hawaii had a lower prevalence, both active and inactive compared with non–Pacific Islanders. The active prevalence estimates found in this racial subgroup are similar to a prior population-based study in Japan, which found a prevalence of 40.4 per 100 000 persons.18 Given that many Pacific Islanders are originally of Asian, particularly Japanese, ancestry, this similarity could support the role of genetic susceptibility in the development and progression of uveitis. Although we did find increasing incidence and prevalence with age, this did not seem to be a confounding factor for the lower prevalence estimates in the subgroup of Pacific Islanders. The age distributions of Pacific Islanders and the overall population in KP Hawaii were similar.
The subgroup of white patients had a significantly higher prevalence of uveitis, both active and inactive, compared with the remainder of the population. Notably, the incidence rate of uveitis for white individuals was not significantly different than the remainder of the population. These results could possibly be explained by higher rates of certain autoimmune diseases in this racial subgroup that are associated with chronic uveitis such as multiple sclerosis.19,20 In addition, some immunological profiles, such as HLA-B27, are more common in white individuals and are associated with recurrent disease.21,22 While the subgroup of black patients in our population did have the highest incidence and prevalence rates compared with other racial groups, statistical comparison was not possible given the small number of patients in this racial group. However, this crude comparison does support the hypothesis that differences in racial demographics could explain the variation between our estimates and those found in the NCEU Study.5
There are some limitations to acknowledge. While the KP Hawaii population represents more than 15% of the Hawaiian population, differences in demographic characteristics could potentially affect the generalizability of our results to the remainder of the Hawaiian population. Based on the most recent census data, the age and sex distribution of the Kaiser population is comparable with the Hawaiian population.23 However, there is a higher proportion of Pacific Islanders in KP Hawaii. Since Pacific Islanders had lower incidence and prevalence rates of uveitis, it is possible that the overall incidence and prevalence rates found in KP Hawaii may be an underestimate of the rates in the Hawaiian population. Another limitation to consider is that the KP Hawaii membership may disproportionately comprise individuals who are healthy enough to be employed and thus have access to Kaiser insurance. If this is the case, it is possible that our results could be biased toward lower incidence and prevalence estimates of potentially disabling diseases such as uveitis.
As with the other epidemiologic studies performed in the United States, generalizability of our results to other populations is a concern. However, this study highlights incidence estimates in a unique population and ascertains differences by racial groups, such as Pacific Islanders, that may not have been possible in other populations. Additionally, our overall incidence and prevalence estimates are comparable with those of the NWVA Study population, suggesting that the overall estimates may be applicable to other populations.
Diagnostic and reporting issues should also be considered. It is possible that some patients may seek specialty care elsewhere despite enrollment in KP Hawaii, which would cause those cases to be missed; however, this is unlikely, given that only a small percentage of KP Hawaii patients have outside insurance. Also, it would be particularly uncommon for Kaiser patients to seek ophthalmologic care out of network given that the KP Hawaii ophthalmology practice has many experienced subspecialty providers including uveitis, retina, and cornea specialists. Our initial case ascertainment was based on ICD-9 codes. It is possible that if patients’ diagnoses were not coded appropriately, these cases could have been missed. However, the initial search began with an extremely broad list of diagnosis codes to the extent that almost three-quarters of cases from the initial screen were excluded after individual medical record review. Additionally, race information was not available for all patients, so estimates of incidence and prevalence by race may be biased if reported race data were not missing at random.
In summary, our study provides population-based uveitis incidence and prevalence estimates of 24.9 cases per 100 000 person-years and 57.5 to 58.0 cases per 100 000 persons, respectively, in a unique population. Furthermore, these results confirm interesting differences by age and sex that have been observed in previous studies.5,6 This study also provides novel comparisons of incidence and prevalence rates by racial subgroup and prompts questions about racial influences on the development and progression of uveitis. Further study is warranted to determine what factors, including genetic susceptibility and environmental influences, are driving these differences.
Corresponding Author: Nisha R. Acharya, MD, MS, F. I. Proctor Foundation, 513 Parnassus Ave, Rm S309, University of California–San Francisco, San Francisco, CA 94143-0412 (email@example.com).
Submitted for Publication: September 30, 2012; final revision received February 25, 2013; accepted February 26, 2013.
Published Online: September 5, 2013. doi:10.1001/jamaophthalmol.2013.4237.
Author Contributions: Dr Acharya 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: Acharya, Tham, Esterberg.
Acquisition of data: All authors.
Analysis and interpretation of data: Acharya, Esterberg, Borkar.
Drafting of the manuscript: Acharya, Esterberg, Borkar, Parker, Vinoya, Uchida.
Critical revision of the manuscript for important intellectual content: Acharya, Tham, Borkar.
Statistical analysis: Acharya, Esterberg, Borkar.
Obtained funding: Acharya, Tham, Esterberg.
Administrative, technical, or material support: Acharya, Tham, Esterberg, Parker, Vinoya, Uchida.
Study supervision: Acharya, Tham.
Conflict of Interest Disclosures: None reported.
Funding/Support: Dr Acharya’s work was supported by National Eye Institute grant K23EY017897, a Research to Prevent Blindness Career Development Award, and a University of California at San Francisco Research Evaluation and Allocation Committee Award. The University of California at San Francisco Department of Ophthalmology is supported by National Eye Institute grant EY06190, That Man May See Foundation, and an unrestricted grant from the Research to Prevent Blindness Foundation.
Role of the Sponsor: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Previous Presentation: This study was presented in part at the American Uveitis Society Meeting; April 30, 2011; Ft Lauderdale, Florida.