Abstract
Background:
Cataract surgery substantially improves patient quality of life. Despite the rising prevalence of dementia in the US, little is known about use of cataract surgery among this group.
Objective:
To evaluate the relationship between dementia status and cataract surgery.
Methods:
Using administrative insurance claims for a representative sample of 1,125,387 US Medicare beneficiaries who received eye care between 2006 and 2015, we compared cataract surgery rates between patients with and without dementia via multivariable regression models to adjust for patient characteristics. Main outcome measures were annual rates of cataract surgery and hazard ratio and 95% confidence interval (CI) for receiving cataract surgery.
Results:
Cataract surgery was performed in 457,128 patients, 23,331 with a prior diagnosis of dementia. 16.7% of dementia patients underwent cataract surgery, compared to 43.8% of patients without dementia. 59 cataract surgeries were performed per 1000 dementia patients annually, versus 105 surgeries per 1000 nondementia patients. After adjusting for patient characteristics, dementia patients were approximately half as likely to receive cataract surgery compared to nondementia patients (adjusted HR = 0.53, 95% CI 0.53–0.54). Among the subset of patients who received a first cataract surgery, those with dementia were also less likely to receive second-eye cataract surgery (adjusted HR = 0.87, 95% CI 0.86–0.88).
Conclusion:
US Medicare patients with dementia are less likely to undergo cataract surgery than those without dementia. This finding has implications for quality of care and dementia progression. More information is necessary to understand why rates of cataract surgery are lower for these patients, and to identify conditions where benefits of surgery may outweigh risks.
INTRODUCTION
Alzheimer’s disease and related dementias affect over 10% of adults aged 70 years and over, and, as the population ages, the prevalence of these conditions is expected to increase [1]. Cataracts, the most common cause of reversible vision loss, are nearly ubiquitous among older adults [2]. Thus, with an aging population, there will be increasing numbers of patients with both cataracts and dementia [3]. Among older adults, cataracts are associated with injury risk, reduced quality of life, and possibly cognitive decline [4–11].
In most cases, cataracts can be safely and successfully treated with a 10–30 min outpatient surgery, resulting in better vision and quality of life and lower risk of injuries such as hip fractures [9, 12–14]. Decision-making around offering cataract surgery for patients with dementia, however, is complex. Cataract surgery for dementia patients may be more difficult and require extra resources—such as longer surgeries requiring more anesthesia, overnight hospital observation instead of immediate discharge home after surgery, or surgery in a hospital instead of freestanding ambulatory surgery center. Patients with dementia may also be at higher risk of complications if they become disoriented or cannot cooperate during awake surgery [15]. Surgeons and anesthesiologists may be more reluctant to operate on these patients due to safety risks and logistical challenges. Further, the benefits of surgery may be lower among patients who are less active, although the association of poor vision with worse cognitive function suggests that improving vision may be beneficial [5]. Despite these considerations, a single center trial demonstrated improved vision and vision-related quality of life and suggested that cataract surgery may even reduce cognitive decline among patients with dementia. And cataract surgery in a patient’s second eye has been shown to improve visual and cognitive function to a level better than after first-eye surgery alone [16–18].
Despite the rising number of dementia patients who have cataracts, relatively little is known about rates of surgery among these patients, compared to others without dementia. In this study, we used national Medicare administrative claims data to examine first- and second-eye cataract surgery rates in a real world-setting, comparing Medicare beneficiaries with and without dementia.
MATERIALS AND METHODS
Data source
Data were obtained from insurance claims from a random, nationally-representative 20% sample of Medicare beneficiaries with Parts A, B, and D coverage between January 1, 2006, and December 31, 2015. Medicare master beneficiary files were used to verify Medicare eligibility and obtain demographics and other patient characteristics. Carrier files (Physician/Supplier Part B claims files) were used to identify all visits, diagnoses, and procedures. Institutional Review Board approval was obtained prior to the study.
Sample selection
The study sample was limited to beneficiaries aged 65 years and older, who had a minimum of four years of continuous coverage in traditional fee-for-service Medicare. Patients enrolled in Medicare Advantage plans and those residing outside the United States were excluded, as we did not have all claims for these enrollees. The first three years of observed plan enrollment were treated as a “lookback period” for each enrollee and used to define predictor variables for subsequent analyses. We selected a three-year lookback period to identify dementia and other comorbidities in order to be consistent with methods published by CMS [19, 20]. We included only patients with at least one visit to an eye care provider (optometrist or ophthalmologist) and a cataract diagnosis during their lookback period (Fig. 1). The characteristics of patients not seen by an eye care provider are shown in Supplementary Table 1. We excluded patients with claims for cataract surgery in their lookback period, based on International Classification of Diseases, 9th edition (ICD-9) and Current Procedural Terminology (CPT) codes for cataract surgery, pseudophakia, or aphakia (Supplementary Table 2).
Sample Selection Strategy. 0coverage ends at the date of deaths if beneficiary died; 1presence of a procedure code for cataract surgery and/or a diagnosis code for pseudophakia or aphakia present in the first 3 years of a given beneficiary’s enrollment period (“lookback period”); 2based on provider type classification for visits during a beneficiary’s lookback period. in the first 3 years of a given beneficiary’s enrollment period; 3presence of a dementia diagnosis recorded during a beneficiary’s lookback period; 4presence of a new dementia diagnosis first recorded after the lookback period but prior to any cataract surgery; 5presence of a new dementia diagnosis first recorded after the lookback period but prior to end of observed enrollment; 6the combined number of patients with incident dementia before second-eye cataract surgery and is higher than the number of incident dementia patients before first-eye cataract surgery because some of these patients were diagnosed with dementia before second-eye surgery but after first-eye cataract surgery (thus counted only for the 2nd analysis).
Predictor variables
Patients with dementia, our predictor of interest, were identified using ICD-9 diagnosis codes (Supplementary Table 2). We defined prevalent dementia based on a dementia diagnosis recorded during the lookback period, and incident dementia based on a new dementia diagnosis noted (1) after the lookback period and (2) before cataract surgery. Methods were consistent with the construction of the Medicare Chronic Condition Warehouse dementia flag and shown in the literature to have good reliability in identifying patients with clinical dementia [19, 20]. Remaining patients—those with no record of dementia or with a dementia diagnosis first noted after cataract surgery—were classified as not having dementia.
Other comorbidities included as model covariates, including depression and common ocular conditions (diabetic retinopathy, age-related macular degeneration, and glaucoma) were also identified by ICD-9 codes (Supplementary Table 2) during patients’ lookback periods. We computed Charlson Comorbidity Index using ICD-9 codes as previously described (omitting dementia) to capture general systemic health (Supplementary Table 2) [21]. Geographic location of residence was evaluated using Rural Urban Commuting Area codes for rural and urban settings based on first observed residence [22]. We also evaluated whether patients were seen at least once by optometrist(s), ophthalmologist(s), or both during their lookback period.
Outcome
Our primary outcome was incident cataract surgery (Supplementary Table 2). Recognizing that cataract surgery may be performed in one or both eyes, we also evaluated second-eye cataract surgery among those patients who underwent first-eye surgery. Second-eye cataract surgery was identified by cataract surgery procedure codes occurring twice for the same patient during the study period.
Statistical analysis
All analyses were performed using SAS software, version 9.4 (SAS Institute, Cary, NC). Enrollee characteristics were summarized using frequencies and percentages. To evaluate initial cataract surgery, we defined the index date as the end of the lookback period, which was three years after initial observed plan enrollment, and followed patients for at least one full year to identify cataract surgery events. We selected this index date to ensure three full years of baseline claims data for each patient; had we instead selected the date of cataract diagnosis as the index date, for example, patients who were diagnosed in their first or second year of observed Medicare enrollment would have had an incomplete three-year lookback period. For second-eye cataract surgery, the index date was defined as the date of first-eye cataract surgery. Cox proportional hazards models with time-varying covariates (including dementia as a time-varying predictor) were constructed to evaluate likelihood of receiving cataract surgery after the index date. For patients who did not undergo cataract surgery during our analysis, the follow-up period was treated as non-informative censored time.
We used separate multivariable regression models to estimate hazard ratios for (1) undergoing any cataract surgery and (2) undergoing second-eye cataract surgery (among those who had a first surgery). Regression models were adjusted for potential confounding variables, including age, race/ethnicity, urban/rural residence, year of plan entry, Charlson Index, depression, and ocular comorbidities, as well as any interaction effect between age and dementia or between sex and dementia [23]. The second-eye cataract surgery model was additionally adjusted for the complexity of a patient’s first-eye cataract surgery based on CPT codes. For our primary analysis, we did not differentiate between incident and prevalent dementia, collectively considered as a time-varying dementia variable. In sensitivity analyses, we separately evaluated the likelihood of cataract surgery among patients with prevalent dementia, compared to those without dementia at baseline, and, after excluding patients with prevalent dementia, evaluated the likelihood of cataract surgery among patients with incident dementia compared to those without dementia.
RESULTS
Study sample
A total of 1,125,387 patients met our study inclusion criteria (Fig. 1), and 70% had at least 5 years of follow-up after the lookback period. Dementia was present in 6.5% of patients (n = 73,455), 5.5% (n = 61,560) with prevalent dementia, and 1.1% (n = 11,895) with incident dementia, excluding those for whom dementia was first diagnosed after cataract surgery. Dementia patients were more likely to be older, female, non-white, living in urban settings, and to have other ocular or systemic comorbidities (Table 1).
Baseline Demographics and Clinical Characteristics among Medicare Beneficiaries based on Dementia Status, 2009-20151 (N = 1,125,387)
12006–2009 data were used to define the lookback period and establish baseline beneficiary characteristics; 2015 data was only applicable for regression analysis since beneficiaries were required to have at least one year of enrollment after the end of the “lookback period”. 2Prevalent dementia was recorded during the initial 3-year lookback period; incident dementia was a dementia diagnosis that was first recorded during the subsequent follow-up period and prior to initial cataract surgery. 3Chi-square test statistic. 4Based on rural urban commuting area codes, in which rural areas represent small urban clusters with a population of 49,000 or fewer persons. 4 Urban patients were overall slightly older, more likely to be seen by an ophthalmologist at baseline, and had higher prevalence of comorbid ocular and systemic diseases (Supplemental Table 8). 5Based on receipt of eye care services during beneficiaries’ “lookback period” (first three years of observed plan enrollment). 6We omitted the dementia variable from computation of Charlson Index, such that the lowest permitted Charlson Index score was 2. 7Because we limited our analysis to patients who had at least one visit with an eye care provider, the prevalence of ophthalmic comorbidities including age-related macular degeneration and glaucoma was higher than would be expected in the general population.
Cataract surgery was performed on 457,128 patients (40.6%) in our sample. Only 16.7% of dementia patients (n = 23,332) underwent cataract surgery during the follow-up period, compared to 43.8% (n = 433,796) of patients without dementia (Fig. 1). The annual rate of cataract surgery among patients with dementia was 59 surgeries per 1000 patients (95% CI 57.9–59.4), compared to 105 surgeries per 1000 patients without dementia (95% CI 104.3–104.9).
Second-eye cataract surgery was performed in 73.5% (n = 335,792) of patients who had first-eye surgery—49.5% (n = 16,269) of dementia patients and 75.3% (n = 319,523) of patients without dementia. Compared to patients who only underwent surgery in one eye, second-eye patients were younger and more likely to be female, white, living in rural settings, and to have a low Charlson Index (Supplementary Table 3). Ocular comorbidities were less likely, and many of these patients were seen by an optometrist only (or optometrist and ophthalmologist) during the lookback period. Patients who underwent second-eye surgery also had a lower proportion of first-eye surgery cases that were complex.
Relationship between dementia and likelihood of receiving any cataract surgery
After adjusting for potential confounders, patients with incident or prevalent dementia were 47% less likely to undergo cataract surgery, compared to non-dementia counterparts (adjusted HR = 0.53, 95% CI: 0.53–0.54) (Table 2).
Adjusted Multivariable Cox Regression Models for Likelihood of Undergoing One or More Cataract Surgeries
1Adjusted for age, sex, race/ethnicity, geographic region, year of plan entry, Charlson comorbidity index, diabetes, depression, diabetic retinopathy, (eye care provider type seen during the “lookback period”). 2Among the subset of beneficiaries who received cataract surgery on their first eye.
We found that the reduction in likelihood of cataract surgery for patients with dementia increased with age. The statistically significant interaction between age and dementia status indicates that dementia patients experience an 18% decrease in likelihood of cataract surgery with every 5-year increase in age. By contrast, patients without dementia were slightly more likely to undergo surgery when older (6% greater likelihood with every 5-year increase in age). Women with dementia were 11.3% less likely than men with dementia to undergo cataract surgery, whereas women without dementia were 15.7% more likely to undergo surgery than men (Supplementary Table 4).
We also observed significant, though less marked, associations between other patient factors and cataract surgery. Patients aged 85 and above or with non-white race had lower likelihood of surgery. Female and rural patients were slightly more likely to undergo surgery, as were patients who saw ophthalmologists (versus just optometrists) during their lookback period. More systemic health problems and presence of comorbid ocular disease were associated with higher likelihood of surgery, whereas comorbid depression was associated with lower likelihood (Supplementary Table 5).
Effect of dementia on likelihood of second-eye cataract surgery
Patients with dementia were 13% less likely to undergo cataract surgery on their second eye (adjusted HR = 0.87, 95% CI: 0.86–0.88) than patients without dementia (Table 2). There was a significant interaction between age and dementia. There was a lower likelihood of second-eye surgery with increasing age among both dementia and non-dementia patients. However, the effect of age was more marked in dementia (15% reduced likelihood of second-eye surgery with every 5-year increase in age among dementia patients, compared to 10% reduced likelihood among non-dementia patients). Sex did not significantly affect hazard of second-eye surgery among dementia patients (Supplementary Table 4).
The direction of effect for most other comorbidities was similar for second-eye surgery as for first-eye surgery. However, rural patients were less likely to have second-eye surgery (Supplementary Table 6). Also, although all patients who underwent cataract surgery saw an ophthalmologist for the operation, those who only saw optometrists during their lookback period were more likely to have second-eye surgery than those who only saw ophthalmologists. Having more systemic health problems, depression, or other ocular diseases were each associated with lower likelihood of second-eye surgery. Having first-eye surgery coded as “complex” was also associated with lower likelihood of cataract surgery in the second eye.
Sensitivity analyses
When we exclusively considered prevalent dementia diagnosed during the lookback period, we found similar results (HR = 0.50 for any cataract surgery among prevalent dementia patients compared to those without dementia, 95% CI: 0.49–0.51, Supplementary Table 7). We also separately analyzed the likelihood of cataract surgery among the subset of patients with incident dementia, compared to those with no dementia, allowing us to estimate the effect of dementia duration considered as time since dementia diagnosis. Cataract surgery remained less likely among incident dementia patients compared to those without dementia (HR = 0.77 for any cataract surgery, 95% CI: 0.75–0.79, and HR = 0.86 for second-eye cataract surgery, 95% CI: 0.83–0.89). Time since dementia diagnosis also affected likelihood of undergoing initial cataract surgery—14% lower likelihood for any/first eye cataract surgery with every additional year since dementia diagnosis—but there was no significant change in likelihood of second-eye surgery based on time since dementia diagnosis (Supplementary Tables 8 and 9).
DISCUSSION
Using a large nationally-representative sample of Medicare enrollees, we compared cataract surgery rates in patients with and without dementia. We found that dementia patients are approximately half as likely to receive cataract surgery compared to patients without dementia—even after adjusting for potential confounders including demographics, geographic location of residence, eye care provider (optometrist and/or ophthalmologist), and clinical comorbidities. 59 cataract surgeries were performed per 1000 patients with dementia annually, compared to 105 surgeries per 1000 patients without dementia. This difference was more pronounced among older dementia patients, who were increasingly less likely to undergo cataract surgery with advancing age, compared to their non-dementia counterparts. Among patients with incident dementia during our analysis, we observed 14% lower likelihood of cataract surgery with every additional year since dementia diagnosis. These findings are important as we seek to ensure public health and access to eye care for a growing dementia population, since cataract surgery has established benefits for vision-related quality of life and injury prevention, and may improve cognitive outcomes in dementia.
Our results are consistent with previous studies that have evaluated the influence of patient characteristics and systemic conditions on odds of cataract surgery, as well as general risk factors for complications [24–29]. Previous findings include a recent analysis of cataract surgery patient characteristics in US Medicare and Veterans’ Affairs populations, where concurrent dementia was associated with reduced odds of receiving cataract surgery within one or five years from initial cataract diagnosis, and an analysis of self-reported survey data to which found lower rates of self-reported cataract surgery among patients with dementia [24, 25].
Our results build on these prior studies by providing longitudinal analysis of the adjusted effect of a dementia diagnosis on subsequent cataract surgery over up to 7 years’ followup. We specifically evaluated the effect of dementia on likelihood of any cataract surgery, as well as likelihood of second-eye surgery among the subset of patients who underwent initial (first-eye) surgery. Our results are robust and timely, based on a 20% representative sample of Medicare data through December 31, 2015.
In the general population, cataract surgery leads to improved vision and contrast sensitivity, and has been shown to improve quality of life (general health, mobility, self-care, activity, pain, and anxiety), reduce depressive symptoms, and decrease rates of fall-related hip fracture [9, 31]. Cataract surgery also has potential to help preserve cognitive function among dementia patients. A retrospective analysis found reduced risk of new-onset dementia among older adults who underwent cataract surgery. Preliminary data from a small randomized clinical trial showed better cognitive outcomes after 6 months among dementia patients who underwent cataract surgery versus those who did not [32–34]. These findings are supported by other prospective observational studies with cognitive function testing [35–38], and collectively suggest that the risk-benefit ratio may favor cataract surgery among dementia patients.
However, dementia patients may be less likely to undergo cataract surgery if risks are perceived too high and benefits too low—due to need for a higher-acuity surgical setting, postoperative hospitalization monitoring, and/or deeper sedation or general anesthesia, with associated risks including postoperative cognitive dysfunction or respiratory complications and difficult extubation which may be more likely in older adults [39–41]. And patients or caregivers may delay surgery until cataracts severely impair vision, at which point surgery may take longer with potentially higher risk of complications. To date there is limited data regarding surgical complexity, complication rates, or outcomes (including quality of life, cognitive trajectory, delirium, falls, or fractures) to inform cataract surgery decision-making.
Although the greatest benefit to quality of life from cataract surgery is achieved after the first-eye operation for most patients, surgery on a patient’s second eye is still often beneficial, reflecting improved depth perception and visual function. Prospective studies have demonstrated further improvement in vision-related quality of life scores following second-eye cataract surgery [17, 18]. However, we observed lower likelihood of second-eye surgery for dementia patients compared to non-dementia patients in this analysis. The effect was less marked than for first-eye surgery (13% versus 47% decrease in hazard), possibly because the subset of eligible patients had fewer systemic or ocular comorbidities and fewer complex first-eye cataract surgeries—that is, already selected as good surgical candidates.
Access to health care and frequency of contact with the health care system may also influence rates of cataract surgery. One possible explanation for lower cataract surgery rates among dementia patients is that they are less likely to be seen by an eye doctor, thus may have eye diseases that are undiagnosed and untreated. Lengthy travel distance to the nearest ophthalmologist, as in rural settings, may also limit patients’ ability to receive surgery. Although we found paradoxically higher cataract surgery rates in rural locations, these patients were generally healthier; sick patients (with dementia and other chronic conditions) were more likely to be in urban settings (Supplementary Table 10) [42].
Our findings also suggest that patients with more severe dementia may be least likely to undergo surgery. Compared to those without dementia, dementia patients are even less likely to undergo cataract surgery at older ages, and the likelihood of cataract surgery declines as the time since initial dementia diagnosis increases. This may reflect logistical factors (challenges getting to the eye doctor or to the surgery center, difficulty lying still during surgery, higher risk for general anesthesia, etc.) and/or reduced functional status with less precise visual needs. Interestingly, the association between dementia and cataract surgery also differs based on sex. Women are less likely than men to undergo cataract surgery if they have dementia, and more likely to undergo cataract surgery if they do not. Reasons for this are likely multifactorial. For example, we speculate that older men with dementia are more likely to have a younger spouse who is able to take on the role of caregiver, including transportation to the ophthalmologist and to the operating room for surgery, pre- and postoperative care, etc., whereas older women with dementia may outlive an older spouse and lack readily-available caregivers. Based on census and survey data, 32% of older women reported living alone in 2014, compared to 18% of men, and older women were substantially more likely to be widowed (37%) than men (13%) [43].
Strengths and limitations
Our study is limited by several factors, including intrinsic shortcomings of administrative claims, such as coding imprecision or inaccuracies [44]. We are also unable to account for unobserved care obtained outside of Medicare; it is possible that some patients may have received cataract surgery or a new dementia diagnosis under different health care plans. However, most patients use Medicare as their primary insurer, and we excluded Medicare Advantage patients from our analysis to minimize unobserved events. Further, our findings are restricted to patients who visited an eye specialist during the 3-year lookback period. We also recognize that some patients in our non-dementia group may have had dementia, since diagnoses are not always reliably recorded and patients may demonstrate early signs of dementia before first documented diagnosis, and up to 40% of patients meeting clinical criteria for probable dementia are undiagnosed [45]. However, such misclassification and underdetection would be expected to bias our results toward inability to detect a difference based on dementia diagnosis. The fact that we still see a difference in dementia patients suggests that the association is potentially even more compelling.
Conclusions and implications
Cataracts are an important reversible cause of visual impairment and reduced quality of life among older adults, and observational data suggests that cataract surgery may improve visual and cognitive outcomes for dementia patients. However, we find that dementia patients are approximately half as likely to undergo cataract surgery compared to their non-dementia counterparts in a US Medicare population. This finding thus carries substantial public health implications, particularly as the prevalence of dementia increases in an aging US population. Dementia patients should perhaps be considered for cataract surgery more often, or undergo surgery earlier than they are currently.
Managing cataracts appropriately requires caution, given the challenges of surgery in dementia patients. There is a need for further research regarding reasons for the discrepancy in cataract surgery rates, including access to eye care as well as cataract surgery safety and outcomes among dementia patients, to determine whether and when benefits outweigh risks. However, it will be important to ensure that dementia patients are not inappropriately deprived of potential benefits from cataract surgery. These findings highlight the need for referral and coordination between primary care providers, geriatricians, neurologists, and ophthalmologists, to enable evaluation for diagnosis and appropriate management of visually-limiting cataracts.
Footnotes
ACKNOWLEDGMENTS
This work was supported by the National Institute on Aging (R03-AG056453). SP also received departmental support from Research to Prevent Blindness and is a consultant for Acumen, LLC (Burlingame, CA) and Verana Health (San Francisco, CA). VWH also received support from P50-AG047366. JDS also received support from R01-EY026641 and departmental support from Research to Prevent Blindness.
Findings presented at the American Geriatrics Society U13 GEMSSTAR Conference, March 25-27, 2018, Bethesda, MD, USA.
