Barriers to Vision Care for Nursing Home Residents Brenda K. Keller, MD, Thomas Hejkal, MD, PhD, and Jane F. Potter, MD Objective: To describe the prevalence of periodic eye examinations by eye professionals and to examine nursing facility resident characteristics associated with lack of periodic screening. Design: Retrospective chart review. Setting: Two Midwestern nursing facilities. Participants: Between 1995 and 1997, 134 subjects aged 60 and older were recruited from two metropolitan nursing facilities . Measurements: Nursing home charts were reviewed for: demographics, length of stay, date of eye examination, eye diagnosis, and visual acuity. Nursing assessments were used to obtain information about cognition, function, behavior, and the presence of Do Not Resuscitate or Do Not Hospitalize orders. The chart was reviewed for visual acuity, intraocular pressures, and the presence of eye pathology. Individuals who had not had eye examinations in the previous 2 years were screened by an ophthalmologist. This examination included external examination of the eye, fundoscopic examination, tonometry, and visual acuity with correction. Results: Only 62 (46%) of the subjects had been seen by an eye care professional in the previous 2 years. Visual acuity information was available for 37/64 pre-
Department of Internal Medicine, Section of Geriatrics, University of Nebraska Medical Center (B.K.K. and J.F.P.) and Department of Ophthalmology, University of Nebraska Medical Center, Omaha, Nebraska. Supported in part by University of Nebraska seed grant MGA/22-071-90201. Presented, in part, at the 50th Annual Scientific Meeting of the Gerontological Society of America, Washington, DC, 1997. Address correspondence to: Brenda K. Keller, MD, Department of Internal Medicine, Section of Geriatrics, 985620 Nebraska Medical Center, University of Nebraska Medical Center, Omaha, NE 68918-5620.
viously examined subjects. Of those with no eye examination in the previous 2 years (n ⫽ 72), visual acuity was obtained in 32 (44%) of subjects. New eye diagnoses were made in 64% (41/64). Logistic regression models with “eye examination within the past 2 years” as the dependent variable show that residents who do not desire hospital transfer are 80% less likely to have had an eye examination than those without this designation. Sex, age, length of stay, functional status, presence of severe dementia, behavior problems, or DNR orders do not change the likelihood that a resident would have been examined. Logistic regression models with “visual acuity measured” as the dependent variable show that residents with severe dementia are 12.6 times less likely to have acuity measured than those without dementia. Those with a length of stay in the facility less than 6 months are 10% less likely to have visual acuity measured. Conclusions: This study does not confirm that barriers still exist in the provision of eye care to all nursing home residents, but the prevalence of such assessments remains low. Additional screening results in a substantial increase in the identification of treatable eye diseases. Contrary to the original hypotheses that patient characteristics that make testing difficult would provide a barrier or disincentive to vision testing, this study did not show statistical differences in the rates of vision screening for those with dementia, behavior problems, or severe functional impairment. Severe dementia does seem to affect the ability of the eye care specialist to gather subjective data such as visual acuity. It also demonstrates that vision screening does take place on nursing home residents with a broad range of cognitive and functional abilities, and this screening results in the diagnosis of many treatable eye conditions. Future efforts should be made to increase vision screening and treatment in the nursing home. (J Am Med Dir Assoc 2001; 2: 15–21)
Copyright ©2001 American Medical Directors Association
Keywords: Vision; nursing home; screening; frail aged
BACKGROUND Visual impairment increases with age and nursing home residents are three times more likely to have visual impairment and almost five times more likely to be legally blind than those living in the community.1 The impact of visual
impairment goes beyond difficulty with reading. Visual impairment in nursing home residents has been associated with decreased ability to perform basic activities of daily living and increased problems with social isolation.2,3 Older individuals with visual impairment fall more often and have a higher rate
Keller et al. 15
of depression than those with normal vision.4,5 For all of these reasons, maximizing visual function is of great importance for frail nursing home residents. Identification of treatable eye pathology requires routine eye care. Traditionally, few nursing home residents have had adequate eye care although many physicians agree that a visual evaluation is a crucial component of a complete nursing home admission assessment.6 Surveys of nursing facilities indicate that only 10 to 15% of nursing home residents have had a recent evaluation by an optometrist or ophthalmologist.7-12 Although these studies, which took place in the 1970s, provide background regarding the availability of eye care, few follow-up studies have been published since the initiation of the Omnibus Budget Reconciliation Act of 1987. This act and the associated regulation were aimed at helping nursing home residents achieve their “highest practicable level of function.”13 Increased attention to the assessment and treatment of sensory impairment is one of the potential impacts of this act.
The reasons for lack of eye care are not clear. Early studies focused on the availability of eye care practitioners. In a survey of 513 facilities that took place in the 1980s, less than 4% of facilities had an optometrist or ophthalmologist serving in a full or part-time capacity. Sixty-seven percent had no vision examinations policy for their residents, and 65% did not provide for periodic vision testing.14 It is possible that factors other than provider availability act as barriers to adequate eye care for nursing home residents. Characteristics of the nursing home resident may also play a role in whether adequate eye care is delivered. Recent studies found that between 14 and 33% of subjects are “medically unable” to complete screening, although little is known about the conditions that lead to this designation.15-17 Cognitively impaired individuals and those who cannot communicate complaints of poor vision may not be referred for timely eye exams and are difficult to assess in screening exams that rely on subjective responses by the patient. Limited research has been conducted to examine resident characteristics associated with eye care practices in nursing home residents. Eye care may be limited by demographic characteristics such as advancing age or by markers of limitation of preventive care such as marked functional impairment and Do Not Resuscitate or Do Not Hospitalize status. Other barriers to routine eye care include characteristics that limit examination, such as severe dementia or behavioral difficulties. The purpose of this study is to describe the use of vision care among nursing home residents in two Midwestern nursing homes and to describe resident characteristics that limit the use of eye care and the assessment of visual acuity. We also describe the prevalence of known and of previously unrecognized eye disease in these nursing home residents. We can then estimate the potential benefits of routine eye care in nursing facilities. METHODOLOGY Design A retrospective chart review took place with examination of subjects not evaluated by an eye care professional in the previous 2 years. Setting Subjects were recruited from two Omaha metropolitan nursing facilities. Both of these facilities had contractual agreements with optometrists to provide on-site eye examinations for the residents. Participants
Fig. 1. This figure shows the reasons for nonparticipation in the study. “No DNR recorded” indicates the number of subjects missing information about resuscitation status. “No HT recorded” indicates the number of subjects missing information about their desire to transfer to the hospital in the case of serious illness. Only facility No. 2 formally recorded hospital transfer desires in the chart. 16 Keller et al.
Potential participants were identified by the nursing facility administrators, and all individuals residing in these facilities older than age 60 were invited to participate. Persons who were residents in the nursing home at the start of the study were considered eligible for enrollment (n ⫽ 311). For nonparticipants, 77 subjects died or left the facility during the recruitment process. Forty-four residents refused participation. For 56 potential subjects, the proxy decision maker could not be reached with two letters and a phone call (See Figure 1). JAMDA – January/February 2001
Table 1. Patient Characteristics Characteristic
Age (years) Sex Female Male Race White Black Hispanic/American Indian Marital status Never married Married Widowed Divorced Education (years) Severe dementia Yes No Behavior problems Yes No Functional status ADL Score ADL ⬍6 ADL ⱖ6 Length of stay (months) Months in residence Stay ⬍ 6 months Stay ⬎ 6 months Patient desires hospital transfer Yes No Patient desires DNR Yes No
134 (85) (15)
126 6 2
(94) (5) (⬍1)
10 14 94 16 132
(8) (10) (70) (12)
(23) (77) (20) (80)
26 107 133
28 105 134 13 121
Measurements The nursing home charts of all consenting subjects were reviewed for date of last eye examination, type of eye care provider (optometrist vs. ophthalmologist), ophthalmologic diagnoses, and visual acuity. For subjects who had been assessed by an eye care specialist in the past 2 years, information was gathered from the chart regarding visual acuity, intraocular pressures, and presence of eye pathology. Eye pathology was categorized into six groups: cataracts, glaucoma, agerelated macular degeneration, diabetic retinopathy, visual field loss, and other. Subjects could have more than one diagnosis. Nursing assessments and the Minimum Data Set were used to obtain information about demographics, length of stay at the nursing home, mental status, functional status, behavioral difficulties, and presence of Do Not Resuscitate or
Mean ⴞ SD 86 ⫾ 8.1
For the remaining 134 subjects, consent was obtained from both the resident and his/her durable power of attorney. Subjects were recruited and studied for 18 months from 1995 to 1997. The services provided through the study were not billed to the patient. This study was approved by the Institutional Review Board of the University of Nebraska Medical Center.
10.9 ⫾ 3
12.9 ⫾ 6
41.2 ⫾ 45
Do Not Hospitalize orders. Only one of the facilities routinely recorded Do Not Hospitalize orders in the medical record. The MDS section of Disease Diagnoses was also reviewed for presence of the following: cataracts, diabetic retinopathy, glaucoma, and macular degeneration. Functional status was further characterized by the Activities of Daily Living (ADL) scale18 completed by a nurse familiar with the residents abilities. Intervention Individuals who had not had eye examinations in the previous 2 years were screened by an ophthalmologist. Screening included external examination of the eye and surrounding tissues, fundoscopic examination, tonometry with the Tonopen, and standard assessments of visual acuity with correction. Eye pathology was categorized into six groups as described above. Cataracts were included in the analysis only if they were associated with visual acuity impairment of 20/40 or worse. Eye diagnoses were considered new if the patient had not seen an eye care professional in 2 years and if eye pathology discovered during this study was not described in the medical chart or on the MDS disease diagnoses section. Keller et al. 17
Fig. 3. The distribution of visual acuity scores for study subjects (n ⫽ 69). Some degree of visual impairment is present in 61% of subjects.
Fig. 2. This figure provides information about the number of subjects with visual acuity (VA) information measured and recorded in the medical record. Some subjects may have had an eye evaluation in the past 2 years but not had visual acuity assessed. “Optometry” indicates that the subjects primary eye care provider performed the evaluation. “Research Team” indicates subjects who had not seen an eye care professional in the past 2 years but had an eye evaluation by our research team ophthalmologist as part of this study. Of the total number of subjects from both facilities, 69/134 had visual acuity measured.
Data Collection and Analysis Descriptive statistics provide information about patient characteristics, the spectrum of eye diagnoses present in the nursing home population, and the characteristics of new eye diagnosis discovered in this study. Two separate logistic regression models were created with the dependent variables (1) presence or absence of eye examination in the past two years and (2) able or unable to complete visual acuity measurement. The following independent variables were entered simultaneously into both of the logistic regression models: sex; age; length of stay; presence or absence of severe dementia, behavior problems, and Do Not Resuscitate orders; and lowest quartile of ADL function. Subjects were divided into two groups of ADL function. Those with the lowest quartile ADL score were considered impaired. This cutoff was used because no one score cutoff could be determined with examination of scatter plots and no conventional cutoff point has been used consistently in previous literature. Logistic regression models were also constructed with the above dependent variables and the independent variables plus hospital transfer status for residents in one nursing facility (n ⫽ 83). This information was not available in the chart review of residents of the other facility and, therefore, could not be analyzed. 18 Keller et al.
All residents of two metropolitan nursing facilities were invited to participate. Proxy consent and adult assent were given for 134 residents. Subjects mean age was 86 years, 85% were women, and 94% were white. Do Not Resuscitate orders were noted on the charts of 82% of subjects, but only 20% of subjects, all in one facility, did not want to be hospitalized for major illness. Dementia was prevalent, and 23% of subjects were classified by nursing staff as severely demented (ie, nursing staff indicated resident never/rarely made decisions.) Subject characteristics are presented in Table 1. No demographic information was gathered on nonparticipants. Only 62 (46%) of the 134 subjects had been seen by an eye care professional in the previous 2 years. Of these previous examined subjects, 53 had been evaluated by an optometrist and nine had been evaluated by an ophthalmologist. Visual acuity information was available for 37 (60%) previously examined subjects, and tonometry values were present for 41 (66%). Seventy-two subjects had had no eye examination in the previous 2 years. These subjects were offered free on-site evaluations, and 64 were evaluated. Of these 64, visual acuity was obtained in 32 (44%) of subjects and tonometry in 41 (57%). Prevalence of vision acuity testing is presented in Figure 2. Information about eye diagnosis was available for 134 subjects, and visual acuity information was present for 69 of these 134 subjects. For individuals with measured visual acuity (n ⫽
Table 2. Logistic Regression Model Predicting Examination by an Eye Care Specialist in the Past 2 Years (N ⫽ 121) Parameters
Sex Age Length of stay Functional status Severe dementia Behavior Do Not Resuscitate
.584 .025 .005 .041 .592 .207 .511
3.0 1.0 0.9 0.9 0.8 0.6 1.4
0.9–9.4 0.9–1.0 0.9–1.0 0.9–1.0 0.2–2.6 0.4–1.0 0.5–3.9
.059 .177 .085 .827 .776 .070 .452
JAMDA – January/February 2001
Fig. 4. Known and newly diagnosed eye disease in nursing home residents.(n ⫽ 134) Diagnoses were called “Old” if they were recorded by the subjects primary eye care specialist or recorded in the minimum data set. New diagnoses were made by the research team ophthalmologist. 152 eye diseases were reported for 120/134 subjects.
69), 39% had normal vision, 31% had mild visual impairment (vision 20/40 –20/70), 15% had moderate vision loss (vision 20/70 –20/200), and 15% were legally blind (vision worse than 20/200) (See Figure 3). In this very old, frail population, the most common cause of visual impairment was cataract, present in 64 (51%) of the subjects. Other diagnoses included 30 subjects (26%) with “other” eye disease, 22 (17%) with age-related macular degeneration, 14 (11%) subjects with glaucoma, 4 (3%) with visual field deficits, and 4 (3%) with diabetic retinopathy. For subjects who had no ophthalmologic eye exam in the previous 2 years, new diagnoses were made in 64% (41/64) of the subjects. These included previously unrecognized cataracts (n ⫽ 17), glaucoma (n ⫽ 3), macular degeneration (n ⫽ 7), visual field loss (n ⫽ 3), and diabetic retinopathy (n ⫽ 1). Other eye pathologies noted in 30 subjects included blepharitis, ectropion, dry eyes, ocular hypertension, and optic neuritis. Figure 4. The director of nursing for each nursing facility was given information about subjects found to have eye pathology requiring treatment.
Logistic regression models with “eye examination within the past 2 years” as the dependent variable showed that the variables sex, age, length of stay in the facility, functional status, and presence of severe dementia, behavior problems, or DNR orders do not change the likelihood that a resident would have been examined. Table 2. However, residents who do not desire hospital transfer are 80% less likely to have had an eye examination. Table 3. To determine if a patients characteristics affected whether visual acuity was measured as part of the examination, logistic regression models with “visual acuity measured” as the dependent variable were constructed. These models compared those subjects with acuity tested (n ⫽ 69) to those without acuity tested (n ⫽ 65). These models show that residents with severe dementia are nine times less likely to have acuity measured than those with mild or no dementia. Table 4. When hospital transfer status is added to the model, severe dementia patients are 12.6 times less likely to have visual acuity measured, and those with a length of stay in the facility less than 6 months are 10% less likely to have visual acuity measured. As ex-
Table 3. Logistic Regression Model Predicting Examination by an Eye Care Specialist in the Past 2 Years for Subjects with Hospital Transfer (HT) Preference Documented (N ⫽ 83) Parameters
Sex Age Length of stay Functional status Severe dementia Behavior Patient desires hospital transfer Do Not Resuscitate
.795 .031 .008 .053 .770 .272 .720 .664
3.7 1.0 0.9 1.0 0.7 0.6 0.2 0.7
0.7–17.5 0.9–1.1 0.9–1.0 0.9–1.1 0.1–3.2 0.4–1.1 0.0–0.8 0.2–2.7
.099 .107 .292 .755 .653 .163 .035 .682
Keller et al. 19
Table 4. Logistic Regression Model Predicting Which Subjects Had Visual Acuity Assessed as Part of Their Eye Examination (N ⫽ 121) Parameters
Sex Age Length of stay Functional status Severe dementia Behavior Do Not Resuscitate
.587 .026 .005 .044 .730 .221 .518
0.8 1.0 0.9 0.9 9.0 1.1 0.4
0.2–2.5 0.9–1.0 0.9–1.0 0.8–1.0 2.1–37.8 0.7–1.7 0.1–1.2
.710 .654 .568 .132 .003 .572 .131
pected, hospital transfer status and DNR orders do not change the likelihood that a resident would have had visual acuity measured. Table 5. CONCLUSIONS Visual impairment was very common in this sample of nursing home residents, 69% of whom had visual acuity worse than 20/40. This is slightly higher than the prevalence of visual impairment noted in other studies of nursing home residents, where 20 to 50% were visually impaired (acuity worse than 20/40) and 11 to 30% met criteria for legal blindness (acuity worse than 20/200.)15,16,19 Cataract is the most common eye pathology in these nursing home residents, similar to that seen in the general population of older persons and a finding consistent with large population-based studies.15 This suggests that routine evaluation could lead to interventions that improve vision. Prospective studies have shown that nursing home residents benefit from cataract extraction, showing both improvement in visual acuity and reduction in depression.20 In our study, eye pathology was present in all of the residents seen by an optometrist/ophthalmologist in the last 2 years, whereas 86% of all residents examined by our team had eye pathology. This suggests a selection bias toward evaluation of residents with known or suspected eye disease which may falsely increase the prevalence of visual impairment in this study. This study confirms that barriers still exist in the provision of eye care to all nursing home residents. Although our study shows that only 46% of residents had been screened for visual impairment, nevertheless, this is much higher than the 10 to
Table 5. Logistic Regression Model Predicting Which Subjects Had Visual Acuity Assessed as Part of Their Eye Examination for Subjects with Hospital Transfer (HT) Preference Documented (N ⫽ 83) Parameters
Odds 95% CI Ratio
Sex Age Length of stay Functional status Severe dementia Behavior Do Not Resuscitate Patient desires hospital transfer
0.9 1.0 0.9 0.8 12.6 0.8 0.5 4.4
.9977 .4099 .0270 .0740 .0093 .5937 .3737 .0570
20 Keller et al.
0.1–5.0 0.9–1.0 0.9–0.9 0.7–1.0 1.8–85.5 0.5–1.1 0.1–2.1 0.9–20.3
15% screening rate described in reports generated in the 1970s and early 1980s. Increased awareness of the importance of vision care for nursing facility residents by the optometry and ophthalmology communities, increased emphasis on screening for sensory loss as part of the functional evaluation by the nursing community, and increased regulation of care of nursing home residents via the Omnibus Budget Reconciliation Act of 1987 all serve as reasons for this improvement in vision screening. The majority of residents with eye exams in the previous 2 years were seen by an optometrist, which is consistent with the contractual agreements with an optometrist to perform eye examinations maintained by both of these facilities. Contrary to the original hypotheses that patient characteristics that make testing difficult would provide a barrier or disincentive to vision testing, this study did not show statistical differences in the rates of vision screening for those with dementia, behavior problems, or severe functional impairment. Some of these patient characteristics, however, such as severe dementia, do seem to affect the ability of the eye care specialist to gather subjective data such as visual acuity. We were unable to complete acuity testing in 51% (33/65) of previously unscreened persons Figure 2. This study confirms the hypothesis that residents with severe and terminal chronic illness who have chosen to limit healthcare services are less likely to be seen by an eye care specialist. This is consistent with the conservative approach to treatment and intervention in this frail subset of nursing home residents. However, the presence of a Do Not Resuscitate order is not associated with lack of eye care in the last 2 years. Generalization of this study to other populations is limited by its relatively small size and use of nursing facilities with on-site eye care. This study may overestimate screening rates because the facilities studied had optometrists available on site. Fifty-seven percent of persons declined participation or were discharged or died before study procedures were completed. Another limitation of this study is that it relied on nursing home records for determination of eye care. This may underestimate visits performed outside of the nursing home. Although these opposing forces may cancel each other, larger population-based research is needed to further clarify the barriers to vision care in the nursing home. Information on payment source was not considered in this study, but this may also influence the pattern of eye care utilization among nursing home residents. This study is unique in its approach to examining patient characteristics rather than practitioner availability as a barrier JAMDA – January/February 2001
to vision screening. It shows that residents with terminal illness, and those who are less likely to undergo healthcare screening of any type, are less likely to undergo vision screening and should not, therefore, be targeted by nursing home visual screening programs. It also demonstrates that vision screening can and does take place in nursing home residents with a broad range of cognitive and functional abilities and that this screening results in the diagnosis of many treatable eye conditions. Future efforts should be undertaken to increase vision screening and treatment in the nursing home. ACKNOWLEDGMENTS The authors thank the patients who participated in this study and their families. The authors acknowledge the work of Joy Morton, BA, and Mark Eggleston, MD, who collected the data, Dale Mundy, BA, who assisted in data analysis, and Jackie Whittington for her technical assistance. REFERENCES 1. Klein R, Klein B, Linton K, DeMets DL. The Beaver Dam Eye Study: Visual Acuity. Ophthalmology 1991;98:1310 –1315. 2. Marx MS, Werner P, Cohen-Mansfield J, Feldman R. The relationship between low vision and performance of activities of daily living in nursing home residents. J Am Geriatr Soc 1992;40:1018 –1020. 3. Resnick HE, Fries BE, Verbrugge LM. Windows to their world: The effect of sensory impairments on social engagement and activity time in nursing home residents. J Gerontol 1997;52B:S135–S144. 4. Tinetti ME, Williams TF, Mayewski R. Fall risk index for elderly patients based on number of chronic disabilities. Am J Med 1986;80: 429 – 434. 5. Rovner BW, Zisselman PM, Shmuely-Dulitzki Y. Depression, and disability in older people with impaired vision. A follow-up study. J Am Geriatr Soc 1996;44:181– 84.
6. Mezey M, Lavizzo-Mourey RJ, Brunswick J, Taylor LA. Consensus among geriatric experts on the components of a complete nursing-home admission assessment. Nurse Pract 1992;17:50 – 61. 7. Corwin BC. Vision screening in nursing and retirement homes. Am J Optom Arch Am Acad Optom 1972;December:1008 –1011. 8. Leslie WJ, Greenberg DA. A survey and proposal concerning visual care at the nursing home level. J Am Optom Assoc 1974;45:461– 466. 9. Newell SW, Walser JJ. Nursing home glaucoma and visual acuity screening results in western Oklahoma. Ann Ophthalmol 1985;17: 186 –189. 10. Search D. Demonstration discussion of blindness: In-service training programs for alternate care facilities. New Outlook for the Blind 1976; 70:398 – 400. 11. Snyder L, Pyrek J, Smith KC. Vision, and mental function of the elderly. Gerontologist 1976;16:491– 495. 12. Verma S. Serving the Visually Impaired Residents in Nursing Homes. Harrisburg, PA: Pennsylvania Dept. of Aging, 1989. 13. Institute of Medicine, Improving the Quality of Care in Nursing Homes. Washington, DC: National Academy of Science Press, 1986. 14. Beliveau M, Yeadon A, Aston S. Innovative Curriculum Development Research to Develop In-Service Training Curriculum for Providers of Long-Term Care to Elderly Blind/Visually Impaired: Final Report. Philadelphia: Pennsylvania College of Optometry, 1986. 15. Tielsch JM, Javitt JC, Coleman A, et al. The prevalence of blindness and visual impairment among nursing home residents in Baltimore. N Engl J Med 1995;332:1205–1209. 16. Mitchell P, Hayes P, Wang JJ. Visual impairment in nursing home residents: The Blue Mountains Eye Study. Med J Aust 1997;166:73–76. 17. Horowitz A. The relationship between vision impairment and the assessment of disruptive behaviors among nursing home residents. Gerontologist 1997;37:620 – 628. 18. Lawton MP, Brody EM. Assessment of older people: self maintaining and Instrumental Activities of Daily Living. Gerontologist 1969;9: 179 –186. 19. Whitmore WG. Eye disease in a geriatric nursing home population. Ophthalmology 1989;96:393–398. 20. Marx MS, Werner P, Billig N, et al. Outcomes of cataract surgery in nursing home residents. Psychosomatics 1995;36:254 –261.
Keller et al. 21