Good vision and eye health are essential components to a person’s quality of life.1–5 As well, the economic burden of vision loss has been shown to be significant.6,7 Comprehensive routine eye examinations (REE) are believed to play a preventative role in vision loss by screening for asymptomatic eye disease.8,9 The prevalence of asymptomatic eye disease has been found to be 14 to 26% of patients10–12 and Quigley13 reports that less than 50% of patients with glaucoma know they have it.
For symptomatic patients, few would argue against the need for an eye examination. Similarly, for patients considered at high risk for ocular disorders, e.g. diabetics,14–16 there is literature rationalizing eye examination frequency.17 However, the ideal REE frequency in asymptomatic patients is unknown. Recommendations should be age dependent as visual outcomes8,18–20 and costs associated with vision deficits6 vary with age.
Existing professional guidelines17,21–23 for REE frequency (Table 1) are based on expert opinion9 or eye disorder prevalence,8 so it is not surprising that they vary by profession and patient age. Ophthalmology tends to recommend REEs less frequently than optometry. This difference could arise from ophthalmology recommendations being based more on disease detection whereas optometry may also consider non-disease conditions such as refractive error. Generally, it is recommended that children and older adults have more frequent REEs than young and middle age adults. A conflict of interest may be perceived for guidelines developed by experts whose profession benefits from a high REE frequency.9 Given the basis of current recommendations, the discrepancies between them, and the potential for conflict of interest, empirical evidence on how REE frequency influences eye disorder detection is needed.
Some research is available to support annual REEs in patients >65 years.8,11,18 Older adults having frequent REEs are less likely to experience vision loss.9,24 For individuals <65 years, the related literature is more sparse. Werner25 reviewed the files of 25- to 35-year-old optometric patients and found ≥38% had refractive changes, 4% had undiagnosed disease, and 15.8% had binocular vision disorders, when examined within 2 years. Assessment intervals of 2 to 5 years resulted in higher numbers of significant findings (newly diagnosed disease, ≥0.50D change in refractive error, failure to meet Sheard’s Criterion for vergence disorders, and decreased amplitude of accommodation for age). A recent study found that a previous-year eye examination was associated with better vision status in 40- to 65-year-old patients.26 Fraser et al.27 found a 25% increased odds of patients (>40 years) first presenting with advanced glaucomatous visual field loss for each year since the last visit to an optometrist. In patients >40 years of age, with normal baseline results, Taylor et al.28 reported that over a 5-year period, 2.39% had a loss in visual acuity to poorer than 20/40. Thirty-seven percent of those patients did not notice a change in vision.
The present study is interested in determining the percentage of asymptomatic patients for which REEs result in spectacle prescription change, new critical diagnosis, or new management of existing conditions for six age groups. We also report the median time intervals between REEs for patients in the different age groups and compare them to currently recommended guidelines.
The Waterloo Eye Study (WatES) is a retrospective cross-sectional database of patients who presented at the University of Waterloo Optometry Clinic during a 1-year period from January 2007 to January 2008. The methods and repeatability of WatES data abstraction and population representation have been outlined earlier.29 The study was approved by the Office of Research Ethics at the University of Waterloo. Data were extracted for all patients whose reason for presenting was to have a routine eye examination as reported in the case history (including those presenting for employment purposes, to obtain contact lenses, or to replace spectacles). Those who did not report any eye-related symptoms, even during case history questioning (e.g. headaches, diplopia, blurred vision, or flashes and floaters), are referred to hereafter as asymptomatic REE patients. There were some patients who initially presented for a REE but reported symptoms when specifically questioned. These patients were excluded from the main analysis, but their overall percentage of significant change is reported for comparison.
A spectacle prescription change was considered to be significant if in at least one eye, the sphere, cylinder, or any reading addition changed by >0.5D from the entering to the exiting spectacle prescription, or if the cylinder axis changed as follows: >15 degrees if the absolute value of the final cylinder value was <1D, >10 degrees if the cylinder was ≥1D but <2D, or >5 degrees if the cylinder was ≥2D.30 A critical diagnosis was considered new if it was not reported in the clinic file case history or at previous examinations. For patients whose most recent previous assessment was not performed at the University of Waterloo (n = 742), diagnoses were considered to be new if they were not recorded in the case history. New diagnoses were classified as critical if the disorder or abnormal finding resulted in vision loss, could progress to vision loss or physical discomfort, or had a systemic implication (see Table 2). A management (not including prescription change) was considered new if it was not initiated at a previous visit or if there was a change compared to the last available information. New managements included referrals, new treatment, or changes in monitoring schedule.
The numbers of asymptomatic REE patients with significant change (defined as one or more of a spectacle prescription change, new critical diagnosis, or new management) since their last assessment were determined for each of the following age groups: <4 years, 4 to <7 years, 7 to <20 years, 20 to <40 years, 40 to <65 years, and ≥65 years. The time between assessments (assessment interval) was determined from the number of years between the study assessment and the patient’s previous eye examination. Assessment intervals were calculated for all patients whose last eye examination was ≥1 year before the study assessment. The median assessment interval was calculated for each age group. The numbers of patients with a spectacle prescription change, new critical diagnosis, new management, or any of these (significant change) were determined for patients presenting for a first ever assessment as well as for the following assessment intervals: 1 to <2, 2 to <3, 3 to <5, 5 to <10, and 10+ years since their previous assessment. Odds ratios (OR) were calculated for patient age, patient sex, assessment interval, and significant change using multivariable logistic regression in SPSS. For this analysis, patient ages and assessment intervals were modeled as continuous variables.
The WatES database contains 6397 patients, 3913 (61%) of which presented for a REE. Of these, 1257 patients reported symptoms when questioned. Thus, there were 2656 asymptomatic REE patients (42% of all patients), and this group is the focus of the study. This group had a median age of 38.5 years, an age range from 0.4 to 93.9 years, with 48% males; comparable to the entire WatES clinic population.29
Overall, there were 1078 (41%) asymptomatic REE patients with a spectacle prescription change, 434 (16%) with a new critical diagnosis, and 809 (31%) with a new management. In total, 1535 (58%) patients had one or more of these outcomes. As age increased, so did the likelihood of a patient having a significant change (Fig. 1). Patients <4 years had the lowest prevalence of significant change (8%), and patients ≥65 had the highest (78%). The percentage of symptomatic REE patients with a significant change was higher (77%) than that for the asymptomatic REE patients (58%).
Previous assessment dates were available for 2606 (98.1%) of the asymptomatic REE patients. There were 142 patients having their first ever eye examination. Of these, 93% were <20 years of age, 69% <7 years, and 46% <4 years. Breakdown by age group and category of change can be found in Table 3.
The assessment interval was <1 year for 119 patients, and these patients along with the first eye examination patients were excluded from the subsequent analysis. Median assessment interval values for the remaining 2345 patients, grouped by age, are shown in Figure 2. Median assessment intervals were greatest for patients in the 20 to <40 and 40 to <65 age groups, at 2.8 and 2.9 years between assessments, respectively. The median assessment interval for patients 7 to <20 years was just over 1.5 years and between 1 and 1.5 years in patients <7 and ≥65 years.
Figure 3 shows the effects of age and assessment interval on the prevalence of having a significant change (N = 2345). Breakdown by category of change can be found in Table 3. For an assessment interval 1 to <2 years, ≥38% of patients in age groups ≥7 had a significant change. All previously examined patients 4 to <7 years of age had an assessment interval of <5 years, with 86% of this age group seen at 1 to <2 years from their last assessment; 11% of these had a significant change. As necessitated by their age, patients <4 years had an assessment interval of <3 years. Approximately 92% of these were seen within 1 to <2 years of their last assessment; close to 5% had a significant change. Through logistic regression analysis, increasing age was found to be significantly associated with increased odds of having a significant change (OR = 1.03, 95% CI 1.03–1.04) while controlling for assessment interval and sex. For every 1-year increase in age, there was a 3% increase in the odds of having a significant change. Similarly, when controlling for age and sex, an increase in assessment interval was associated with increased likelihood of having a significant change (OR = 1.06 per year, 95% CI 1.02–1.11). Patient sex was not significantly associated with having a significant change (OR = 1.07 for females, 95% CI 0.90–1.29).
Figure 4 demonstrates that across all assessment intervals except ≥10 years, the most frequent contributor to a significant change for the 2487 asymptomatic REE patients was a spectacle prescription change, followed by a change in management and, lastly, a new critical diagnosis. The prevalence of any significant change as well as a change in each category appears to increase with longer assessment intervals (Fig. 4, Table 3). Exceptions include spectacle prescription change and any significant change at >10 years, and new management for 5 to ≤10 year interval. At 1 to <2 years, nearly 15% of patients had a new critical diagnosis which increased to 27% for an interval ≥10 years. Similarly, prevalence of management change ranged from a minimum of 28% between 1 and <2 years to a maximum of 43% at ≥10 years. Prescription changes were lowest at 40% for 1 to <2 years and highest at 52% for 5 to <10 years. The frequency of patients with any significant change increased from a minimum of 56% between 1 and <2 years to a maximum of 70% between 5 and ≤10 years.
More than half of the asymptomatic patients (58%) who presented for a REE had a change in ocular status or care compared with 77% of symptomatic REE patients. In asymptomatic patients, age was a strong predictor of having a significant change. This was true regardless of the assessment interval and corresponds well with known age-related ocular changes such as presbyopia and increasing prevalence of eye disease.8,18–20 It also makes sense that as the assessment interval increased, the odds of having a significant change increased. The longer a patient waits for their next assessment, the older they will be at presentation, increasing the risk of age-related conditions. However, we also found an association between assessment interval and detection of a significant change when controlling for age, so although the association between age and assessment interval does play a role, it is not the sole explanation. Greater assessment intervals would allow more time for a disease or condition to develop, irrespective of age.
Many factors influence the assessment interval for individual patients including patient age, cost of examination, insurance coverage, recommendations given by practitioners or professional bodies, practice recalls (not applicable at this clinic), as well as patients’ perceived risk of visual impairment and their understanding of the consequences of not seeking eye care. The observed median assessment interval for the various asymptomatic REE age groups (how often patients have an REE, Fig. 2) matches more closely the recommended optometric guidelines21,22 than the age-related trend in significant change outcomes found in the current study (Fig. 1). Presumably, this is because this is how they are instructed by practitioners. REEs for patients aged 20 to 64 years were not publicly funded except for 8 defined medical conditions. Annual REEs for patients <20 and >64 years were publically funded. Jin et al.31 showed that de-insurance of eye examinations reduced the uptake of eye examinations for people in lower income levels. The WatES clinic is in an area of slightly above average socio-economic status compared to the province of Ontario overall.32 Patients in funded age groups had smaller time intervals between assessments than non-funded age groups (Fig. 2), suggesting that those with above average socio-economic status may also be affected by a lack of insurance or public funding. Furthermore, based on data from the Canadian Longitudinal National Population Health Survey, Chan et al.33 found that for patients 65 years or older, provinces where REEs were not funded had reduced patient awareness of glaucoma and cataracts, and increased vision loss.
Comparing our data to the existing literature is difficult because the age ranges, study populations, conditions evaluated, criteria for change, exclusion criteria, and dates of studies (which may reflect scope of practice changes) all vary.12,25,27 Fraser27 studied patients with newly diagnosed glaucoma and found the adjusted odds ratio of first presenting with advanced glaucoma increased by 1.25 times per year since their last visit to an optometrist. Although Werner’s25 data were derived from an academic optometric population, the data covered a more limited age range, potentially a different patient demographic, had different exclusion criteria, and was conducted 25 years earlier than WatES. The closest comparison is our 20 to <40 years age group and 1 to <2 years assessment interval to their 25- to 35-year-old patients and <24 months assessment interval. Their finding of 15% of asymptomatic patients having refractive, disease, or binocular vision problems is considerably lower than our 58% of asymptomatic patients with a significant change. Wang et al.12 studied patients who were ≥40 years of age from a primary care clinic in a teaching hospital. Their population was predominately African American, had a high prevalence of chronic diseases, and had low socioeconomic status. They found a prevalence of unknown eye disease of 43% for a 1 to <2 year assessment interval. WatES new critical diagnosis values for the only comparable assessment interval were lower, 13 and 28% for 40 to <65 and ≥65 years, respectively. Similar to Wang et al.12, we found an increase in new critical diagnoses for patients 40 to <65 and ≥65 years when the assessment interval increased. In contrast to Wang et al.12, we did not find that persons 40 to <65 were more likely to have a critical diagnosis of which they were unaware than those ≥65 years.
In the WatES clinic population, the detection of significant change for patients ≥7 years of age is >38% for an assessment interval of 1 to <2 years (Fig. 3). This is substantial and although we have not done an economic analysis, it could be argued that patients >7 years should have an eye examination every year. In asymptomatic REE patients <7 years of age, the detection of significant change was much lower. Given the potential inability of young children to communicate symptoms and the impact of disorders on visual development (e.g. amblyopia34), policy makers may wish to choose a lower threshold of detection for this age group when determining recommended eye examination frequency.
Limitations of the current study include those inherent in a cross-sectional study design. Although a clinical population may not represent the general population, it is representative of those who are actually seeking care. The WatES clinic population does compare favorably in terms of patient age and sex distribution to a nationwide survey of Canadian optometric practices.11,29 Robinson11 found that 32.6% of patients that presented to optometric practices came for a REE and expressed no concerns compared with 41% of the WatES clinic patient population classified as asymptomatic REE patients. Both of these values contradict Michaud and Forcier’s10 claim that asymptomatic patients are rarely seen in an optometry clinic. It is possible that some patient symptoms and/or findings were not recorded despite routine questioning of every patient. Missed examination findings would result in conservative estimates whereas missed symptoms would overestimate significant changes if persons who were not truly asymptomatic were inadvertently included.
In general, assessment intervals for the various asymptomatic REE patient age groups closely match the Canadian Optometric guidelines. Given an overall >50% detection of significant change, routine eye examinations do appear to be productive in asymptomatic patients, and this appears to increase with age.
Elizabeth L. Irving
University of Waterloo
School of Optometry and Vision Science
200 University Avenue West
Waterloo, ON, N2L 3G1
None of the authors have any financial interests or relationships to disclose. This work was supported by Canada Research Chair #950-202761 and VSP Vision Care for Life.
Received July 31, 2015; accepted December 30, 2015.
1. Brown MM, Brown GC, Sharma S, Kistler J, Brown H. Utility values associated with blindness in an adult population. Br J Ophthalmol 2001;85:327–31.
2. Lee PP, Spritzer K, Hays RD. The impact of blurred vision on functioning and well-being. Ophthalmology 1997;104:390–6.
3. Tahhan N, Papas E, Fricke TR, Frick KD, Holden BA. Utility and uncorrected refractive error. Ophthalmology 2013;120:1736–44.
4. West CG, Gildengorin G, Haegerstrom-Portnoy G, Schneck ME, Lott L, Brabyn JA. Is vision function related to physical functional ability in older adults? J Am Geriatr Soc 2002;50:136–45.
5. West SK, Rubin GS, Broman AT, Munoz B, Bandeen-Roche K, Turano K. How does visual impairment affect performance on tasks of everyday life? The SEE Project. Salisbury eye evaluation. Arch Ophthalmol 2002;120:774–80.
7. Wittenborn JS, Zhang X, Feagan CW, Crouse WL, Shrestha S, Kemper AR, et al. The economic burden of vision loss and eye disorders among the United States population younger than 40 years. Ophthalmology 2013;120:1728–35.
8. Robinson BE, Mairs K, Glenny C, Stolee P. An evidence-based guideline for the frequency of optometric eye examinations. Primary Health Care. 2012; 2(121): doi: 10.4172/2167-1079.1000121.
9. Picone G, Brown D, Sloan F, Lee P. Do routine eye exams improve vision? Int J Health Care Finance Econ 2004;4:43–63.
10. Michaud L, Forcier P. Prevalence of asymptomatic ocular conditions in subjects with refractive-based symptoms. J Optom 2014;7:153–60.
11. Robinson BE. Prevalence of asymptomatic eye disease. Can J Optom 2003;65:175–80.
12. Wang F, Ford D, Tielsch JM, Quigley HA, Whelton PK. Undetected eye disease in a primary care clinic population. Arch Intern Med 1994;154:1821–8.
13. Quigley HA. Number of people with glaucoma worldwide. Br J Ophthalmol 1996;80:389–93.
14. Kristinsson JK, Hauksdottir H, Stefansson E, Jonasson F, Gislason I. Active prevention in diabetic eye disease. A 4-year follow-up. Acta Ophthalmol Scand 1997;75:249–54.
15. Klein R, Klein BE, Moss SE, Davis MD, DeMets DL. The Wisconsin epidemiologic study of diabetic retinopathy. II. Prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years. Arch Ophthalmol 1984;102:520–6.
16. Sprafka JM, Fritsche TL, Baker R, Kurth D, Whipple D. Prevalence of undiagnosed eye disease in high-risk diabetic individuals. Arch Intern Med 1990;150:857–61.
18. Klaver CC, Wolfs RC, Vingerling JR, Hofman A, de Jong PT. Age-specific prevalence and causes of blindness and visual impairment in an older population: the Rotterdam study. Arch Ophthalmol 1998;116:653–8.
19. Attebo K, Mitchell P, Smith W. Visual acuity and the causes of visual loss in Australia. The Blue Mountains Eye Study. Ophthalmology 1996;103:357–64.
20. Kahn HA, Leibowitz HM, Ganley JP, Kini MM, Colton T, Nickerson RS, et al. The Framingham Eye Study. I. Outline and major prevalence findings. Am J Epidemiol 1977;106:17–32.
23. Clinical Practice Guideline Expert Committee. Canadian Ophthalmological Society evidence-based clinical practice guidelines for the periodic eye examination in adults in Canada. Can J Ophthalmol 2007;42:39–45, 158–63.
24. Sloan FA, Picone G, Brown DS, Lee PP. Longitudinal analysis of the relationship between regular eye examinations and changes in visual and functional status. J Am Geriatr Soc 2005;53:1867–74.
25. Werner DL. The routine eye examination for the asymptomatic patient age 25–35. J Am Optom Assoc 1981;52:899–903.
26. 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;131:499–506.
27. Fraser S, Bunce C, Wormald R, Brunner E. Deprivation and late presentation of glaucoma: case–control study. BMJ 2001;322:639–43.
28. Taylor HR, Vu HT, McCarty CA, Keeffe JE. The need for routine eye examinations. Invest Ophthalmol Vis Sci 2004;45:2539–42.
29. Machan CM, Hrynchak PK, Irving EL. Waterloo eye study: data abstraction and population representation. Optom Vis Sci 2011;88:613–20.
30. Smith G. Refraction and visual acuity measurements: what are their measurement uncertainties? Clin Exp Optom 2006;89:66–72.
31. Jin YP, Buys YM, Hatch W, Trope GE. De-insurance in Ontario has reduced use of eye care services by the socially disadvantaged. Can J Ophthalmol 2012;47:203–10.
33. Chan CH, Trope GE, Badley EM, Buys YM, Jin YP. The impact of lack of government-insured routine eye examinations on the incidence of self-reported glaucoma, cataracts, and vision loss. Invest Ophthalmol Vis Sci 2014;55:8544–9.
34. Eibschitz-Tsimhoni M, Friedman T, Naor J, Eibschitz N, Friedman Z. Early screening for amblyogenic risk factors lowers the prevalence and severity of amblyopia. J AAPOS 2000;4:194–9.