Key Points
Question
What is the association between the geographic distribution of eye care clinicians and visual impairment in California?
Findings
In this cross-sectional study including more than 30 million residents of California, a higher number of ophthalmologists and optometrists was associated with decreased prevalence of visual impairment. For the increase of every 1 eye care clinician per 100 000 residents, there was an adjusted statistically significant decrease of 3.90 persons with visual impairment per 100 000 residents.
Meaning
Higher number of eye care clinicians is potentially associated with lower prevalence of visual impairment in California.
This cross-sectional study examines associations between the geographic distribution of eye care clinicians and visual impairment in California.
Abstract
Importance
The association between availability of eye care clinicians and visual impairment, a condition presenting with increased morbidity and health care costs, has not been thoroughly studied.
Objective
To examine associations between the geographic distribution of eye care clinicians and visual impairment in California.
Design, Setting, and Participants
This survey-based cross-sectional study included ophthalmologists and optometrists licensed in California in 2018 and 2020 as well as respondents to the 2014 to 2018 American Community Survey (ACS) by California counties and Medical Service Study Areas (MSSAs). Data were analyzed from August 2020 to December 2021.
Main Outcomes and Measures
Prevalence of visual impairment by county and MSSA.
Exposures
The number of eye care clinicians was determined based on the number of member ophthalmologists of the American Academy of Ophthalmology in 2018 and optometrists listed in the 2020 Blue Book of Optometrists in California. The prevalence of visual impairment was determined using questionnaire data from the American Community Survey. Linear regression was used to assess multivariable associations between number of eye care clinicians and visual impairment by MSSA.
Results
A total of 30 068 581 California residents were included; 15 253 655 (50.7%) were female, and 5 314 389 (17.7%) were 65 years and older. The overall number of eye care clinicians was 22.18 clinicians per 100 000 residents. The overall prevalence of visual impairment was 2411.07 residents with visual impairment per 100 000 residents. San Francisco County had the highest number of eye care clinicians per 100 000 residents (39.24 clinicians per 100 000 residents). Four counties had no eye care clinicians (Alpine, Mariposa, Inyo, and Sierra counties). For every increase of 1 eye care clinician per 100 000 residents, there was a mean (SE) decrease of 3.90 (1.39) persons with visual impairment per 100 000 residents in adjusted analyses.
Conclusions and Relevance
In this cross-sectional study, a higher number of eye care clinicians was potentially associated with lower prevalence of visual impairment in California. Additional studies are needed to assess eye care clinician availability on a national and global scale and strategies to improve access to eye care.
Introduction
Visual impairment is a prevalent condition worldwide and can be defined by multiple measures. According to the International Classification of Diseases, visual impairment encompasses visual acuity cutoffs less than 6/12, while blindness is defined as a visual acuity of less than 3/60 with the best possible correction.1 In the World Health Organization VISION 2020 Action Plan, visual impairment includes the above visual acuity cutoffs as well as impairment secondary to visual field loss less than 20°.2 Approximately 4.24 million people in the US were visually impaired in 2015, and this number is projected to double by 2050.3 Visual impairment is a national health concern for reasons including increased systemic morbidity4 and mortality5 for individuals with poor vision. Moreover, the cost of caring for those with visual impairment is estimated to be $5.5 billion annually.6
Risk factors for low vision and visual impairment have been previously reported, including older age, female sex, lower education level, and rural residence.7,8,9 One potential risk factor that has not been well studied is the availability of eye care clinicians. Gibson10 examined the geographic distribution of eye care clinicians in the US and reported that 24% of US counties had no ophthalmologists or optometrists. McGwin et al11 estimated that one-third of participants in the Behavioral Risk Factor Surveillance System survey did not visit an eye care clinician in a 1-year period. These studies suggest a need to further investigate associations between availability of eye care clinicians and rates of visual impairment in the population. Medical Service Study Areas (MSSAs) are subcity and subcounty geographic units designated by the California Office for Statewide Health Planning and Development (OSHPD) and are recognized as “rational service areas”12 for the purposes of designating health care professional shortages and medically underserved areas. This cross-sectional study aimed to describe the geographic distribution of eye care clinicians in California and to explore associations between the number of eye care clinicians and prevalence of visual impairment in California on the county and MSSA levels. Because of the development of MSSAs within California, this study focused on associations within California rather than on the national or global level.
Methods
This study included data from the 2014 to 2018 American Community Survey (ACS),13 California Health and Human Services Open Data Portal,12 American Academy of Ophthalmology (AAO) membership data, and 2020 Blue Book of Optometrists.14 The variables of interest were the number of ophthalmologists and optometrists per 100 000 residents and prevalence of visual impairment by county and MSSA geographic units. County data were obtained from the 2014 to 2018 ACS. Approximately 3 million addresses are surveyed each year, and the self-reported survey data are pooled to produce estimates for 50 US states and Puerto Rico. Five-year estimates are the most comprehensive of the ACS period estimates. The California Health and Human Services Open Data Portal was used to obtain population characteristics by MSSA, based on information obtained from the US Census. We obtained institutional review board exemption from the University of California, Los Angeles, Institutional Review Board for this study, as it uses public use data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
The number of ophthalmologists in each California county was obtained from the 2018 year-end membership data from the AAO, of which an estimated 92% of US-based practicing ophthalmologists are members. Physician characteristics analyzed included age, sex, and practice type. The number of ophthalmologists in each county and MSSA was determined using practice zip codes from the AAO membership data with geographic information system proportioning. The number of optometrists was obtained from the 2020 Blue Book of Optometrists and divided into counties and MSSAs in the same manner as above.
Visual impairment was assessed from participant questionnaires in the ACS. The ACS defined vision difficulty as “blind or having serious difficulty seeing, even when wearing glasses”15; participants who responded positively were considered to have visual impairment for the purposes of this study. Prevalence of visual impairment by county was determined by matching census tracts to the county they comprised, using the geography crosswalk file provided by the California OSHPD.16 MSSA data were determined in a similar manner using data from the OSHPD. Distribution of visual impairment for each county and MSSA was calculated as prevalence per 100 000 survey respondents 18 years and older.
Covariates assessed as potential confounders included age, sex, race and ethnicity, education, income, lack of health insurance, and residence type. Data for these covariates were obtained from ACS participant questionnaires. Age was analyzed as the percentage of respondents 65 years and older. Sex was analyzed as the percentage of male and female respondents, and the term sex was used to match ACS wording. Race and ethnicity were classified into Asian, Black, Latino, White, and other, based on self-identification from options predetermined by the ACS. Education was assessed as the percentage with college education or above. Income was assessed as percentage earning 150% or less of the federal poverty line. Residence type was classified into urban, rural, and frontier using MSSA definitions based on population density per square mile.12
Descriptive statistics were used to examine baseline characteristics of residents and ophthalmologists in California. Baseline characteristics of optometrists were not analyzed, as this information was not available. Distribution of eye care clinicians for each county and MSSA was calculated as number per 100 000 residents of both ophthalmologists and optometrists combined. Quantum Geographic Information System (QGIS) version 3.1617 was used to make geographic information system maps for the number of eye care clinicians per 100 000 individuals and prevalence of visual impairment by MSSA. Distribution of visual impairment for each county and MSSA was calculated as prevalence of visual impairment per 100 000 residents.
At the MSSA and county levels, Spearman correlation coefficients and Evans effect size18 were used to evaluate the correlation between number of eye care clinicians and prevalence of visual impairment per 100 000 residents. Multivariable associations between number of eye care clinicians and prevalence of visual impairment per 100 000 residents were performed using linear regression, adjusting for all study covariates. To improve validity and model fit, linear regression models were performed among MSSAs with at least 1 eye care clinician owing to the large number of MSSAs without any eye care clinicians (124 [22.9%]). In addition, 1 MSSA was deemed as an outlier owing to its extremely large number of ophthalmologists (783 per 100 000 residents) and optometrists (356 per 100 000 residents) and was excluded from regression models. Linear regression models were performed among the remaining 417 MSSAs. SAS version 9.4 (SAS Institute) was used for all statistical analyses. All reported P values are 2-sided, and significance was set at P < .05; no adjustments were made for multiple comparisons.
Results
In this study, there were 58 counties and 542 MSSAs in California. Based on data from the 2014 to 2018 ACS, the total population in California 18 years and older was 30 068 581. A total of 5 524 602 respondents (14.1%) identified as Asian, 2 163 910 (5.5%) as Black, 15 220 875 (38.9%) as Latino, 14 690 682 (37.5%) as White, and 1 540 451 (3.9%) as other race. A total of 14 814 926 participants (49.3%) identified as male, while 15 253 655 (50.7%) identified as female. There were 2 919 430 residents (10.0%) with no health insurance, 9 035 853 residents (23.1%) earning 150% or less of the federal poverty line, 8 710 305 (33.2%) with college education or above, and 25 921 503 (87.6%) who lived in an urban area. Additional characteristics of the California population are outlined in Table 1.
Table 1. Characteristics of California Population 18 Years and Older From the 2014 to 2018 American Community Survey 5-Year Data by the US Census Bureau.
| Characteristic | No. (%) |
|---|---|
| Total population, No. | 30 068 581 |
| Age ≥65 y | 5 314 389 (17.7) |
| Sex | |
| Male | 14 814 926 (49.3) |
| Female | 15 253 655 (50.7) |
| Race and ethnicitya | |
| Asian | 5 524 602 (14.1) |
| Black | 2 163 910 (5.5) |
| Latino | 15 220 875 (38.9) |
| White | 14 690 682 (37.5) |
| Otherb | 1 540 451 (3.9) |
| College education or abovec | 8 710 305 (33.2) |
| Income ≤150% federal poverty level | 9 035 853 (23.1) |
| No health insuranced | 2 919 430 (10.0) |
| Residence | |
| Frontier | 181 824 (0.6) |
| Rural | 3 491 251 (11.8) |
| Urban area | 25 921 503 (87.6) |
Race and ethnicity were classified into Asian, Black, Latino, White, and other, based on self-identification from options predetermined by the American Community Survey.
Other race includes non-Hispanic American Indian and Alaska Native, non-Hispanic Native Hawaiian and Other Pacific Islander alone, non-Hispanic other race alone, and non-Hispanic 2 or more races.
Among those 25 years and older.
Among those 19 years and older.
Table 2 describes characteristics of ophthalmologists in California from the 2018 AAO membership data. There were 2378 ophthalmologists, most of whom were aged 40 to 59 years (965 [40.6%]) and male (1655 [69.6%]). There were 4186 optometrists in California listed in the 2020 Blue Book of Optometrists.
Table 2. Characteristics of Ophthalmologists in California From the American Academy of Ophthalmology Membership Dataa.
| Characteristic | No. (%) |
|---|---|
| Total, No. | 2378 |
| Age, y | |
| <40 | 564 (23.7) |
| 40-59 | 965 (40.6) |
| 60-79 | 666 (28.0) |
| ≥80 | 70 (2.9) |
| Unknown | 113 (4.8) |
| Sex | |
| Male | 1655 (69.6) |
| Female | 699 (29.4) |
| Unknown | 24 (1.0) |
| Practice type | |
| Academic institution | 284 (11.9) |
| Hospital or health care system | 196 (8.2) |
| Multispecialty or ophthalmology group practice | 833 (35.0) |
| Solo practice | 478 (20.1) |
| Unknown | 491 (20.7) |
| Otherb | 96 (4.0) |
Taken from 2018 year-end American Academy of Ophthalmology membership data.
Includes administration, consultant, government or military, international ophthalmology, medical missionary or humanitarian, and research.
Table 3 describes the crude number of eye care clinicians and visual impairment prevalence by county in California in 2018 to 2020. Overall, there were 22.18 eye care clinicians per 100 000 residents. Four counties had no eye care clinicians (Alpine, Mariposa, Inyo, and Sierra counties). The 3 counties with the highest number of eye care clinicians per 100 000 residents were San Francisco County (39.24), Mono County (35.18), and San Luis Obispo County (31.19). The 3 counties with the highest prevalence of visual impairment per 100 000 residents were Tehama County (5127.94), Calaveras County (4855.73), and Merced County (4676.09), and the 3 counties with the lowest prevalence were Sierra County (1096.67), Marin County (1647.15), and San Mateo County (1705.68).
Table 3. Number of Eye Care Clinicians and Prevalence of Visual Impairment per 100 000 Residents 18 Years and Older in California in 2018 by Countya,b.
| County | No. per 100 000 | Prevalence of visual impairment per 100 000 (n = 713 511) | ||
|---|---|---|---|---|
| Ophthalmologists (n = 2378) | Optometrists (n = 4186) | Eye care clinicians (n = 6564) | ||
| All counties | 8.04 | 14.15 | 22.18 | 2411.07 |
| Tehama | 4.19 | 12.58 | 16.78 | 5127.94 |
| Calaveras | 2.69 | 5.38 | 8.07 | 4855.73 |
| Merced | 2.15 | 7.51 | 9.66 | 4676.09 |
| Modoc | 0 | 14.26 | 14.26 | 4477.4 |
| Trinity | 0 | 9.50 | 9.50 | 4200.32 |
| Plumas | 0 | 25.84 | 25.84 | 4024.29 |
| Fresno | 5.82 | 12.95 | 18.77 | 3966.29 |
| Shasta | 7.19 | 17.25 | 24.43 | 3910.78 |
| Alpine | 0 | 0 | 0 | 3858.88 |
| Glenn | 0 | 9.93 | 9.93 | 3849.02 |
| Madera | 1.92 | 9.59 | 11.51 | 3836.75 |
| Del Norte | 10.85 | 10.85 | 21.7 | 3813.81 |
| Mariposa | 0 | 0 | 0 | 3797.38 |
| Lake | 1.99 | 7.97 | 9.96 | 3650.93 |
| Yuba | 0 | 7.62 | 7.62 | 3650.29 |
| Tulare | 3.83 | 12.13 | 15.96 | 3607.47 |
| Imperial | 3.36 | 7.56 | 10.92 | 3603.13 |
| Butte | 11.18 | 15.09 | 26.27 | 3597.00 |
| Stanislaus | 3.08 | 13.86 | 16.94 | 3504.62 |
| Kings | 3.25 | 18.43 | 21.68 | 3208.85 |
| Humboldt | 3.68 | 13.81 | 17.49 | 3165.40 |
| Tuolumne | 4.77 | 21.48 | 26.25 | 3159.30 |
| Nevada | 4.94 | 18.53 | 23.47 | 3124.58 |
| Siskiyou | 0 | 11.55 | 11.55 | 3111.83 |
| Sutter | 2.88 | 20.14 | 23.01 | 3089.40 |
| Amador | 0 | 7.02 | 7.02 | 3078.06 |
| San Joaquin | 4.41 | 10.92 | 15.33 | 2991.06 |
| Kern | 2.01 | 4.35 | 6.36 | 2953.20 |
| Mono | 8.80 | 26.39 | 35.18 | 2937.55 |
| Riverside | 3.49 | 6.29 | 9.78 | 2791.88 |
| Lassen | 0 | 18.20 | 18.20 | 2766.15 |
| San Bernardino | 3.90 | 9.39 | 13.29 | 2705.85 |
| Colusa | 0 | 6.49 | 6.49 | 2595.04 |
| Mendocino | 5.90 | 22.14 | 28.05 | 2569.97 |
| Sacramento | 8.82 | 13.41 | 22.24 | 2561.38 |
| Santa Cruz | 5.49 | 13.72 | 19.21 | 2443.71 |
| Ventura | 7.63 | 17.13 | 24.76 | 2441.60 |
| Los Angeles | 8.59 | 13.06 | 21.65 | 2353.81 |
| Napa | 0 | 2.79 | 2.79 | 2323.83 |
| Contra Costa | 6.69 | 15.68 | 22.37 | 2313.34 |
| Inyo | 0 | 0 | 0 | 2249.75 |
| Sonoma | 7.82 | 16.90 | 24.71 | 2225.10 |
| San Francisco | 16.02 | 23.23 | 39.24 | 2223.98 |
| Solano | 3.36 | 13.43 | 16.79 | 2216.93 |
| El Dorado | 4.07 | 16.28 | 20.35 | 2200.36 |
| San Luis Obispo | 12.92 | 18.27 | 31.19 | 2134.85 |
| Monterey | 5.61 | 14.53 | 20.14 | 2130.59 |
| San Diego | 8.83 | 14.11 | 22.95 | 2066.47 |
| Alameda | 7.05 | 15.97 | 23.02 | 1999.76 |
| Placer | 5.14 | 15.43 | 20.57 | 1955.93 |
| Yolo | 5.96 | 15.50 | 21.46 | 1898.36 |
| Santa Barbara | 6.81 | 14.21 | 21.02 | 1882.19 |
| Orange | 10.22 | 17.68 | 27.90 | 1849.19 |
| San Benito | 2.29 | 13.74 | 16.03 | 1809.11 |
| Santa Clara | 11.35 | 17.43 | 28.78 | 1716.79 |
| San Mateo | 12.14 | 16.47 | 28.61 | 1705.68 |
| Marin | 8.88 | 21.70 | 30.58 | 1647.15 |
| Sierra | 0 | 0 | 0 | 1096.67 |
Taken from 2018 year-end American Academy of Ophthalmology membership data.
Taken from the 2020 Blue Book of Optometrists.14
The distribution of visual impairment and eye care clinicians by MSSA is illustrated in the Figure, A and B. There was a negative association between the number of eye care clinicians and prevalence of visual impairment in the population at both the MSSA level (r, −0.21; 95% CI, −0.30 to −0.12; P < .001; weak by Evans; Figure, C) and county level (r, −0.40; 95% CI, −0.59 to −0.15; P < .001; moderate by Evans). There were 288 MSSAs (53.1%) that contained both ophthalmologists and optometrists, 124 MSSAs (22.9%) without any eye care clinicians, 9 MSSAs (1.7%) with ophthalmologists only, and 121 MSSAs (22.3%) with optometrists only. The mean (SD) prevalence of visual impairment was 3170.53 (1331.78) residents with visual impairment per 100 000 residents in MSSAs with optometrists only and 2365.07 (926.31) per 100 000 residents in MSSAs with both ophthalmologists and optometrists (P < .001). The prevalence of visual impairment was not compared in MSSAs with optometrists only vs ophthalmologists only owing to the small number of MSSAs with ophthalmologists only.
Figure. Distribution of and Correlation Between Eye Care Clinicians and Visual Impairment.
A, Prevalence of visual impairment per 100 000 residents in California by Medical Service Study Area (MSSA), using data from the 2014 to 2018 American Community Survey 5-year estimate. B, Number of eye care clinicians, including both ophthalmologists and optometrists, per 100 000 residents in California by MSSA, using data from the 2018 American Academy of Ophthalmology and 2020 Blue Book of Optometrists. The gray scales of the maps were determined by the default Jenks natural breaks classification method in the geographic information system software (QGIS version 3.16). This method reduces the variance within each subgroup while maximizing the variance between subgroups. C, A scatterplot between the number of eye care clinicians per 100 000 residents and prevalence of visual impairment per 100 000 residents in California shows a negative correlation (r, −0.21; 95% CI, −0.30 to −0.12; P < .001).
Table 4 summarizes multivariable associations between the number of eye care clinicians (optometrists and ophthalmologists combined) and prevalence of visual impairment on the MSSA level. At the MSSA level, the percentage of individuals with no health insurance was highly correlated with the percentage of individuals earning 150% or less of the federal poverty level (Pearson r = 0.70; 95% CI, 0.66 to 0.74; P < .001; strong by Evans), and federal poverty level had a higher correlation with the rate of visual impairment than lack of health insurance (federal poverty level: r, 0.49; 95% CI, 0.43 to 0.55; P < .001; moderate by Evans; lack of health insurance: r, 0.21; 95% CI, 0.13 to 0.29; P < .001; weak by Evans). Therefore, lack of health insurance was excluded from adjusted analysis, as the federal poverty level was included. For the increase of every 1 eye care clinician per 100 000 residents, there was a mean (SE) decrease of 3.32 (1.71) residents with visual impairment per 100 000 residents in the unadjusted model. When adjusting for all covariates, there was a mean (SE) decrease of 3.90 (1.39) residents with visual impairment per 100 000 residents for the increase of every 1 eye care clinician per 100 000 residents. In adjusted analyses, additional factors significantly associated with increased visual impairment included increased age, male sex, and income of 150% or less of the federal poverty level, while factors associated with decreased visual impairment included Latino ethnicity and college education or above.
Table 4. Risk Factors for Visual Impairment From Linear Regression Model at the Level of Medical Service Study Area Among California Residents 18 Years and Older.
| Risk factor | Change in No. of residents with visual impairment per 100 000 residents, Estimate (SE) | |||
|---|---|---|---|---|
| Unadjusted | P Value | Adjusteda | P Value | |
| Ophthalmologists and optometrists (change per increase in 1 eye care clinician per 100 000 residents) | −3.32 (1.71) | .05 | −3.90 (1.39) | .005 |
| Age ≥65 y (change per 1% increase) | 14.35 (10.46) | 0.17 | 36.42 (10.80) | <.001 |
| Male sex (change per 1% increase) | 89.63 (21.47) | <.001 | 54.24 (16.64) | .001 |
| Race and ethnicity (change per 1% increase)b | ||||
| Asian/otherc | −30.56 (4.02) | <.001 | −4.91 (3.62) | .18 |
| Black | 26.64 (8.67) | .002 | 14.36 (7.49) | .06 |
| Latino | 1.71 (2.40) | .48 | −15.52 (3.21) | <.001 |
| White | 1 [Reference] | NA | 1 [Reference] | NA |
| Income ≤150% of federal poverty line (change per 1% increase) | 56.74 (4.10) | <.001 | 46.89 (5.46) | <.001 |
| College education or above (change per 1% increase) | −37.0 (2.60) | <.001 | −23.71 (4.29) | <.001 |
| Residence | ||||
| Urban | 1 [Reference] | NA | 1 [Reference] | NA |
| Frontier | 1758.31 (273.83) | <.001 | 478.55 (269.87) | .08 |
| Rural | 722.52 (115.38) | <.001 | 72.56 (115.22) | .53 |
Abbreviation: NA, not applicable.
Adjusted model includes number of ophthalmologists and optometrists, age, sex, race and ethnicity, poverty, college education or above, and urban residence.
Race and ethnicity were classified into Asian, Black, Latino, White, and other, based on self-identification from options predetermined by the American Community Survey.
Other race includes non-Hispanic American Indian and Alaska Native, non-Hispanic Native Hawaiian and Other Pacific Islander alone, non-Hispanic other race alone, and non-Hispanic 2 or more races. The number of individuals of other race at the Medical Service Study Area level was too small to be included alone for the validity and stability of regression model fit, so the other race category was combined with Asian race as one covariate in the regression models owing to technical issues of statistical analysis.
Discussion
This study examined associations between the number of eye care clinicians and prevalence of visual impairment per 100 000 residents by county and MSSA in California in 2018 to 2020. On the county and MSSA levels, a higher number of eye care clinicians was associated with decreased prevalence of visual impairment in adjusted analyses.
Compared with a nationwide study of eye care clinician availability from 2011,10 findings from this study suggest that California may have more eye care clinician availability compared with the US overall. The nationwide study found that per 100 000 US residents, there were 2.1 ophthalmologists, 11.2 optometrists, and 13.2 total eye care clinicians, all of which were lower than the present study. Additionally, they reported that 24.0% of US counties had no ophthalmologists or optometrists, which is higher than the present study. A 1994 study19 estimated the ratio of ophthalmologists at the time to be around 5.6 ophthalmologists per 100 000 individuals, and another study at the time20 found a possible imbalance between eye care clinician supply and requirements. However, further investigation is needed to determine optimal ratios of eye care clinicians to population size and to create and implement effective interventions for improved eye care access for regions where clinician availability is low or absent.
This study found that on the MSSA and county levels, increased number of eye care clinicians per 100 000 residents was associated with decreased prevalence of visual impairment. While a direct causal relationship cannot be inferred, studies have shown that greater ophthalmologist availability was positively associated with increased usage of eye care services and better visual health outcomes among populations with diabetic retinopathy21 and that increased eye care clinician availability increased the odds that patients with age-related macular degeneration received an eye examination.22 An association specifically between ophthalmologist availability and decreased prevalence of visual impairment has been observed in other studies. Gibson23 found that the population with the highest ophthalmologist availability quartile was less likely to have vision-threatening diabetic retinopathy and more likely to be aware that they had diabetic retinopathy, whereas the optometrist availability quartiles were not significantly related to these outcomes. The present study found decreased prevalence of visual impairment in MSSAs with ophthalmologists and optometrists compared with MSSAs with optometrists only. Our findings suggest that overall increased eye care clinician availability, especially that of ophthalmologists, is associated with decreased prevalence of visual impairment.
Multiple demographic factors were positively or negatively associated with visual impairment in this study, including race and ethnicity, sex, residence type, and income. Multiple previous studies have provided support for these observations. One study found men were more likely to have glaucoma after adjusting for demographic factors.24 Other studies demonstrated that lower socioeconomic status was associated with worse visual outcomes25 and that children from families earning less than the federal poverty level had nearly double the likelihood of visual impairment.26 A potential confounding factor in all of these studies is access to eye care, and further studies of correlations between socioeconomic status and access to eye care are warranted, supported by our finding that prevalence of visual impairment was associated with earning less than 150% of the federal poverty level.
Limitations
This study has several limitations. First, MSSAs are defined in relation to primary care clinicians, and this study borrows the MSSA designation for eye care clinicians. Further study is needed to determine if an alternate classification should be used to geographically categorize ophthalmologists and optometrists. Second, data regarding number of ophthalmologists was from 2018, while data regarding number of optometrists was from 2020, and the prevalence of visual impairment was calculated using 5-year estimate data from 2014 to 2018. Since these data were collected in different years, it may not reflect the exact association between eye care clinicians and visual impairment at a given time point. We anticipate that the number of eye care clinicians per capita did not change greatly within this time, although future studies using eye care clinician information from the same year would be more informative. There is the potential for uncontrolled confounding by unmeasured factors, and the associations that were found in this study cannot be inferred as causal factors. Additionally, there is potential for misclassification of the geographic area of practice of eye care clinicians. The California county with the lowest prevalence of visual impairment (Sierra County) was reported as having 0 eye care clinicians between 2018 and 2020. This may be because the geographic location of ophthalmologists and optometrists in the AAO records and Blue Book of Optometrists is self-reported and can be either residential or office locations. It would also be possible that an eye care clinician could have additional addresses for secondary practices that were not reported. Additionally, the AAO membership data do not include nonmember ophthalmologists, and there is a possibility of ophthalmologists present in areas that had been noted as having 0 ophthalmologists. Another consideration is that with recent advances in telemedicine, eye care clinicians may reach patients outside of their practice zip codes. In addition, the new artificial intelligence devices being used to detect diabetic retinopathy may decrease the association of the prevalence of visual impairment and eye care clinicians, since these devices increase access to diabetic retinopathy screening. Furthermore, this study analyzed data solely within the state of California owing to the development of the MSSA designation within California. Availability of the MSSA designation on the national or global level would allow for further generalizability of findings. With ACS data, population data were based on weighted estimates, which may deviate from true population values. Additionally, the data used to calculate visual impairment does not offer specific information on laterality of visual impairment, visual acuity, and etiology of visual impairment, and the aggregate nature of ACS data may also limit its estimation abilities and ability to capture individual vs contextual data. Future studies could be conducted using data sets providing more comprehensive ophthalmology, optometry, and visual impairment information. Factors to consider with other databases would include availability of clinical data, geographic reach, and payer mix.
This cross-sectional analysis examined data from the 2014 to 2018 period. A year-by-year trend analysis was not feasible, as ACS data are generally analyzed in aggregate form over multiple years to produce sufficient sample sizes for stable estimates, and year-by-year optometry and ophthalmology data were not available for this study beyond the years analyzed. Future studies examining trends in associations between eye care clinician availability and visual impairment over time would be beneficial to inform needs in workforce availability and access to care. Additionally, future studies specifically examining the association between poverty level and visual impairment would be of value, as poverty is explored in this study only as a potential confounding factor. Additional points of future study could include further analyses of associations by type of eye care clinician and expansion of studies related to access to care, such as associations between distance traveled to the clinician and visual outcomes.27
Conclusions
In summary, this study found associations between increased number of eye care clinicians and decreased prevalence of visual impairment in California after adjusting for demographic and socioeconomic factors. Further studies are needed to investigate reasons for these associations, as well as strategies to improve access to eye care and reduce visual impairment in the population and determine whether these results can be generalized to areas outside of California.
References
- 1.World Health Organization . 9D90 Vision impairment including blindness. Accessed November 1, 2021. https://icd.who.int/browse11/l-m/en#/http%3a%2f%2fid.who.int%2ficd%2fentity%2f1103667651
- 2.International Agency for the Prevention of Blindness . VISION 2020. action plan 2006-2011. Accessed September 1, 2021. https://www.iapb.org/learn/resources/vision-2020-action-plan-2006-2011/
- 3.Varma R, Vajaranant TS, Burkemper B, et al. Visual impairment and blindness in adults in the United States: demographic and geographic variations from 2015 to 2050. JAMA Ophthalmol. 2016;134(7):802-809. doi: 10.1001/jamaophthalmol.2016.1284 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lee DJ, Gómez-Marín O, Lam BL, Zheng DD, Caban A. Visual impairment and morbidity in community-residing adults: the National Health Interview Survey 1986-1996. Ophthalmic Epidemiol. 2005;12(1):13-17. doi: 10.1080/09286580490907751 [DOI] [PubMed] [Google Scholar]
- 5.Thompson JR, Gibson JM, Jagger C. The association between visual impairment and mortality in elderly people. Age Ageing. 1989;18(2):83-88. doi: 10.1093/ageing/18.2.83 [DOI] [PubMed] [Google Scholar]
- 6.Frick KD, Gower EW, Kempen JH, Wolff JL. Economic impact of visual impairment and blindness in the United States. Arch Ophthalmol. 2007;125(4):544-550. doi: 10.1001/archopht.125.4.544 [DOI] [PubMed] [Google Scholar]
- 7.Zhang X, Andersen R, Saaddine JB, Beckles GL, Duenas MR, Lee PP. Measuring access to eye care: a public health perspective. Ophthalmic Epidemiol. 2008;15(6):418-425. doi: 10.1080/09286580802399102 [DOI] [PubMed] [Google Scholar]
- 8.Rim TH, Nam JS, Choi M, Lee SC, Lee CS. Prevalence and risk factors of visual impairment and blindness in Korea: the Fourth Korea National Health and Nutrition Examination Survey in 2008-2010. Acta Ophthalmol. 2014;92(4):e317-e325. doi: 10.1111/aos.12355 [DOI] [PubMed] [Google Scholar]
- 9.Srinivasan S, Swaminathan G, Kulothungan V, Raman R, Sharma T; Medscape . Prevalence and the risk factors for visual impairment in age-related macular degeneration. Eye (Lond). 2017;31(6):846-855. doi: 10.1038/eye.2017.72 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Gibson DM. The geographic distribution of eye care providers in the United States: implications for a national strategy to improve vision health. Prev Med. 2015;73:30-36. doi: 10.1016/j.ypmed.2015.01.008 [DOI] [PubMed] [Google Scholar]
- 11.McGwin G, Khoury R, Cross J, Owsley C. Vision impairment and eye care utilization among Americans 50 and older. Curr Eye Res. 2010;35(6):451-458. doi: 10.3109/02713681003664931 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.California Department of Technology . MSSA 2010c1 public. Accessed April 2, 2022. https://gis.data.ca.gov/datasets/CDPHDATA:mssa-2010c1-public/about
- 13.US Census Bureau . American Community Survey (ACS): methodology. Accessed October 6, 2019. https://www.census.gov/programs-surveys/acs/methodology.html
- 14.Jobson Optical Research . The Blue Book of Optometrists: 2020. Federation of American Hospitals; 2020. [Google Scholar]
- 15.US Census Bureau . How disability data are collected from the American Community Survey. Accessed April 2, 2022. https://www.census.gov/topics/health/disability/guidance/data-collection-acs.html
- 16.California Office of Statewide Health Planning and Development . Geography crosswalk file. Accessed October 6, 2019. https://data.chhs.ca.gov/dataset/4545ca2c-000b-4c1b-b264-466cb6d66c33/resource/d045eb9c-2484-45bb-bb28-355501379158/download/geography-crosswalk.csv
- 17.QGIS . A free and open source geographic information system. Accessed October 6, 2019. https://qgis.org/en/site/
- 18.Evans RH. An analysis of criterion variable reliability in conjoint analysis. Percept Mot Skills. 1996;82:988-990. doi: 10.2466/pms.1996.82.3.988 [DOI] [Google Scholar]
- 19.Weiner JP. Forecasting the effects of health reform on US physician workforce requirement. evidence from HMO staffing patterns. JAMA. 1994;272(3):222-230. doi: 10.1001/jama.1994.03520030064030 [DOI] [PubMed] [Google Scholar]
- 20.Lee PP, Jackson CA, Relles DA. Estimating eye care workforce supply and requirements. Ophthalmology. 1995;102(12):1964-1971. doi: 10.1016/S0161-6420(95)30767-1 [DOI] [PubMed] [Google Scholar]
- 21.Wang F, Javitt JC. Eye care for elderly Americans with diabetes mellitus. failure to meet current guidelines. Ophthalmology. 1996;103(11):1744-1750. doi: 10.1016/S0161-6420(96)30432-6 [DOI] [PubMed] [Google Scholar]
- 22.Sloan FA, Brown DS, Carlisle ES, Picone GA, Lee PP. Monitoring visual status: why patients do or do not comply with practice guidelines. Health Serv Res. 2004;39(5):1429-1448. doi: 10.1111/j.1475-6773.2004.00297.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Gibson DM. Eye care availability and access among individuals with diabetes, diabetic retinopathy, or age-related macular degeneration. JAMA Ophthalmol. 2014;132(4):471-477. doi: 10.1001/jamaophthalmol.2013.7682 [DOI] [PubMed] [Google Scholar]
- 24.Rudnicka AR, Mt-Isa S, Owen CG, Cook DG, Ashby D. Variations in primary open-angle glaucoma prevalence by age, gender, and race: a Bayesian meta-analysis. Invest Ophthalmol Vis Sci. 2006;47(10):4254-4261. doi: 10.1167/iovs.06-0299 [DOI] [PubMed] [Google Scholar]
- 25.Li YJ, Xirasagar S, Pumkam C, Krishnaswamy M, Bennett CL. Vision insurance, eye care visits, and vision impairment among working-age adults in the United States. JAMA Ophthalmol. 2013;131(4):499-506. doi: 10.1001/jamaophthalmol.2013.1165 [DOI] [PubMed] [Google Scholar]
- 26.Centers for Disease Control and Prevention (CDC) . Visual impairment and use of eye-care services and protective eyewear among children—United States, 2002. MMWR Morb Mortal Wkly Rep. 2005;54(17):425-429. [PubMed] [Google Scholar]
- 27.Lee CS, Morris A, Van Gelder RN, Lee AY. Evaluating access to eye care in the contiguous United States by calculated driving time in the United States Medicare population. Ophthalmology. 2016;123(12):2456-2461. doi: 10.1016/j.ophtha.2016.08.015 [DOI] [PMC free article] [PubMed] [Google Scholar]