Purpose
To compare self-administration of drops in both visually impaired glaucoma subjects and retina subjects.
Design
Prospective, observational study.
Methods
Setting: Distinct glaucoma and retina practices. Study Population: Subjects with glaucoma or retinal diseases with visual acuity of 20/60 or worse in 1 eye, significant field loss, or both. Observation Procedures: Subjects were video recorded self-instilling a drop onto the worse eye. Main Outcome Measure: Proper instillation of eye drop onto ocular surface.
Results
We included 409 subjects (205 glaucoma, 204 retina). Differences between the groups included the following: glaucoma subjects included fewer females ( P = .05), included fewer white persons ( P < .005), had worse visual acuity ( P < .005), had less self-reported arthritis ( P < .05), were younger ( P < .005), and had more previous exposure to drop use ( P < .005). Glaucoma subjects had more bilateral impairment (60% vs 42%; P < .0005). Retina subjects instilled more drops (1.7 vs 1.4; P = .02) and more frequently touched the bottle to the eye (47% vs 33%; P = .003). Of subjects claiming not to miss the eye, nearly one third from each group ( P = .32) actually missed. Approximately one third of each group could not get a drop onto the eye (30% retina vs 29% glaucoma; P = .91). Among subjects placing 1 drop onto the eye without touching the adnexae, there was a trend for glaucoma patients to perform better, although both groups did poorly (success, 39% glaucoma vs 31% retina; P = .09).
Conclusions
Among visually impaired subjects, regardless of cause, drop administration was a problem. Both groups wasted drops, contaminated bottles, and had inaccurate perception of their abilities. This has implications for future therapeutic delivery systems.
We recently published data regarding eye drop self-instillation in patients with impaired vision resulting from glaucoma, but are unaware of data concerning patients with retinal disease who also might have impaired vision and are required to instill eye drops as part of their therapy. These treatments may include topical antibiotics related to intravitreal injections or endophthalmitis, topical steroids or nonsteroidal drugs during a postoperative course or for uveitis, or artificial tears for dry eyes. There are also no previous reports in a computerized search of PubMed comparing the abilities of these populations (glaucoma and retinal disease) to administer eye drops. This information may be helpful in assessing the effects of central and peripheral vision loss on eye drop administration.
In the present study, we extend our observational video recorded study of eye drop instillation to a population of subjects with visual impairment (low vision or blindness) resulting from diagnoses associated with retinal pathologic features such as age-related macular degeneration, diabetic retinopathy, diabetic macular edema, and pseudophakic cystoid macular edema. The intent of this study was to observe the eye drop instillation behavior of subjects with retinal diseases and to compare it with a parallel group of glaucoma subjects. Depending on self-declared race, glaucoma and retinal diseases are either the number 1 or number 2 causes of irreversible blindness in the United States. Based on United States census data from 2000, the estimated prevalence of low vision is 1.98% and the prevalence of blindness is 0.78% among those older than 40 years. The specific causes of visual impairment vary greatly by race and ethnicity, with older white persons predominantly affected by age-related macular degeneration, and older African Americans affected by glaucoma and diabetes. Age-related macular degeneration, diabetic retinopathy, primary open-angle glaucoma, and cataracts account for approximately 75% of the cases of blindness and 85% of visual impairment in adults older than 40 years.
Poor patient adherence, even if unintentional, can have adverse effects on a person’s response to therapy, as well as their disease progression. The group that may most need and may be most persistent with therapy is the group with the greatest visual impairment. Although those who are visually impaired may be more motivated, they in fact may have more difficulty administering eye drops.
Methods
Study Population
We enrolled subjects with glaucoma, retinal disease, or both who had Early Treatment Diabetic Retinopathy Study (ETDRS) visual acuity of 20/60 or worse in at least 1 eye, demonstrated Hodapp-Parrish-Anderson criteria for moderate or severe visual field damage on their most recent visual field test in 1 eye, or both. Two groups were enrolled simultaneously between September 2008 and September 2009; one group consisted of glaucoma subjects from a glaucoma subspecialty practice and the second group consisted of subjects with retinal disease from a private subspecialty retina practice. The catchment areas from both retinal and glaucoma practices were similar and reflected an adequate sample of other patients with these diagnoses. All patients admitted to self-instilling their own eye drops.
Patients were classified as having glaucoma if glaucoma was the primary diagnosis to explain their impaired vision. Patients were categorized as having retinal disease if the retina diagnosis was determined to be their primary cause for impaired vision.
We stratified subjects as having low vision those with an ETDRS visual acuity (VA) worse than 6/18 (20/60) but better than 6/60 (20/200) or those as blind with an ETDRS VA of 6/60 (20/200) or worse in at least 1 eye. We excluded eyes with light perception or no light perception vision.
Design
This was a prospective, consecutive, observational study. Inclusion criteria if subjects met the above vision criteria were: 18 years of age or older; admission that the patient self-instilled eye drops; confirmed diagnosis of glaucoma according to examination of the optic disc, visual field, and optic nerve, or confirmed diagnosis of a retinal disease based on the specialist’s clinical examination results (after which subjects were classified in the retina or glaucoma group based on the above definitions); English language proficiency; and willingness to participate in an observational study. We allowed other visual comorbidities such as cataract. Exclusion criteria were unwillingness to participate, significant language barrier, prior adverse reaction to artificial tears, requiring assistance with eye drop administration, and dementia.
Observation Procedures
Subjects completed a questionnaire regarding their prior experience with eye drops. If vision was too poor to complete it, the questionnaire was read to them. We gave subjects an unopened sterile 5-mL bottle of artificial tears and instructed them to instill 1 drop onto their worse-seeing eye using their dominant hand in a manner similar to what they might normally use at home. We video recorded their administration of this drop. No practice sessions were given, and recordings were made on the subject’s first attempt of administration. This was the same procedure we have used previously.
We gave subjects access to a mirror and a reclining chair, and they were allowed to sit, stand up, or lie down, consistent with their usual technique. If both eyes had the same level of visual impairment, we selected the left eye. Trained personnel or an investigator performed all video recording and administered all questionnaires. After the video recording was complete, an investigator or a trained technician offered all subjects suggestions for improved techniques of drop instillation. We did not evaluate the results of this counseling.
The main outcome measure was successful instillation of 1 eye drop onto the eye. Secondary end points were bottle–eye contact, the inability to instill a drop, and the patients’ perceptions of their performance. Half (0.5) of a drop was defined as an amount between 0 instilled (i.e., landing on the cheek) and 1 drop that landed fully on the ocular surface—where at least a portion of the drop landed on the eye and also a portion on the lid or other location.
Statistical Analysis
A single masked investigator (A.L.H.) analyzed all surveys and video using Stata software version 10.0 (StataCorp LP, Cary, North Carolina, USA) and graded each video for the following: location of the instilled drop (eye, eyelid, canthus, cheek, or other location), number of drops instilled, distance of the bottle tip to ocular surface, any touch between the bottle tip and the ocular surface, and the number of attempts (approach of bottle toward eye) made to administer 1 drop. This investigator was not aware of subjects’ identities, diagnoses, or VAs at the time the videos were graded.
Because practitioners may have varying definitions of clinical success for eye drop instillation, we defined success on 2 levels: (1) putting any number of drops onto the surface of the eye, and (2) placing just 1 drop on the eye without touching the surface of the eye. Two different multivariate logistic regression models were created, the second with a more strict definition. The variable for visual impairment resulting from either a retinal or glaucoma cause was included in each of the models as a categorical value (0 or 1).
Results
We enrolled 409 subjects (205 in the glaucoma group and 204 in the retina group) who fulfilled VA or visual field criteria, or both, in either eye. Only a small number (< 1%; 3/412) of subjects who qualified declined to participate in the study. All videos were readable for outcomes assessment by a masked observer (A.L.H.). Table 1 shows major diagnoses for each group. The most common glaucoma diagnosis was open-angle glaucoma (182/205), and the most common retina diagnoses were age-related macular degeneration (exudative or nonexudative; 80/204) and complications related to diabetes (32/204). There were 32 subjects that had both retinal and glaucoma diagnoses; however, they remained in the retina or glaucoma group alone, depending on the primary cause for their impaired vision. Those who had both diagnoses did not perform significantly differently from either the glaucoma or retina group and were not presented separately in the regression models because of the small sample size. Glaucoma subjects tended to use eye drops for chronic intraocular pressure-lowering therapy, and retina patients had used drops after surgery or related to intravitreal injections.
| Glaucoma | No. (%) | Retina | No. (%) |
|---|---|---|---|
| Open-angle glaucoma | 182/205 (88.8) | Age-related macular degeneration | 80/204 (39.2) |
| Neovascular glaucoma | 7/205 (3.4) | Diabetic complications: NPDR, PDR, DME | 32/204 (15.7) |
| Traumatic glaucoma | 7/205 (3.4) | Cystoid macular edema | 18/204 (8.8) |
| Angle closure glaucoma | 3/205 (1.5) | Vein occlusion (central or branch) | 14/204 (6.9) |
Table 2 shows baseline characteristics of both groups of subjects and statistically significant differences between the groups. The glaucoma group had fewer females ( P = .05), had fewer white persons ( P < .005), had better visual acuity ( P < .005), were younger ( P < .005), were more educated ( P < .005), had more previous exposure to eye drop use ( P < .005), and had less self-reported arthritis ( P < .05) than retina subjects. Mean ETDRS VA was 0.78 ± 0.9 among glaucoma subjects and 1.12 ± 0.6 among retina subjects ( Table 2 ). A greater proportion of retina subjects had visual acuity of 20/200 or worse (54% vs 27%; P = .00). Among the glaucoma subjects, the automated visual field mean defect was −14.5 ± 8.0 (range, −33.8 to 2.6). We did not compare moderate visual field loss with severe visual field loss in this study.
| Variable | Glaucoma (n = 205), n (%) | Retina (n = 204), n (%) | P Value |
|---|---|---|---|
| Female | 112 (54.6) | 131 (64.2) | .05 |
| White | 150 (73.2) | 173 (85.2) | < .005 |
| Mean age, years | 68.9 | 73.5 | < .005 |
| Education beyond high school | 71 (34.6) | 37 (18.2) | < .005 |
| Previous eye surgery (yes) | 175 (85.4) | 138 (68.0) | < .005 |
| Using eyedrops > 6 months | 193 (94.1) | 136 (70.5) | < .005 |
| Mean no. of eye medications | 2.0 | 1.6 | < .05 |
| VA (logMAR) | 0.78 ± 0.87 | 1.12 ± 0.59 | < .005 |
| Bilateral visual impairment | 124 (60.5) | 86 (42.4) | < .005 |
| Tremor a | 7 (3.5) | 21 (10.3) | < .05 |
| Arthritis a | 53 (26.2) | 86 (42.4) | < .05 |
| History of CVA/TIA a | 20 (9.9) | 22 (10.8) | .76 |
Approximately one third of both groups could not get a drop onto the ocular surface (30% retina group vs 29% glaucoma group; P = .91). Retina subjects instilled more drops onto the eye (1.7 vs 1.4; P = .03) and more frequently touched the bottle tip to the eye (47% vs 33%; P = .003) than glaucoma subjects. Retina subjects also made more attempts (defined as bringing the bottle toward the eye) at instilling the eye drops (1.3 retina group vs 1.2 glaucoma group; P = .07). Among subjects placing only 1 drop onto the eye without touching the ocular surface, there was a trend for glaucoma patients to perform slightly better, although both groups did poorly (success, 39% glaucoma group vs 31% retina group; P = .09). Three percent of retina subjects used a mirror, whereas 16% of glaucoma subjects used one ( P < .005).
Subjects from both groups also were asked about perceptions of their own abilities to instill eye drops. Of the 143 subjects (68 in the retina group, 75 in the glaucoma group; P = .75) who said they never missed their eye, 29% of retina subjects and 35% of glaucoma subjects actually did miss the eye ( P = .32). Nearly 80% of subjects overall stated they had no trouble instilling eye drops, yet only 28% of retina subjects and 29% of glaucoma subjects ( P = .85) were able to instill a drop onto the eye. Of these 80% who reported no problems with self-instillation, 50% of retina subjects and 35% of glaucoma subjects touched the bottle tip to their eyes ( P = .006). Of the 290 subjects (148 in the retina group, 142 in the glaucoma group) who denied touching the bottle tip to the eye, 41% of retina subjects and 24% of glaucoma subjects did touch the bottle to the eye on review of the videos ( P = .002).
Tables 3 and 4 display factors that we thought might predict success of eye drop instillation. In logistic regression models controlling for age, race, gender, previous eye surgery, bilaterality of disease (VA 20/60 or worse in both eyes), type of disease (retinal vs glaucoma), self-reported history of stroke, and self-reported history of arthritis, only age was a significant factor in predicting a subject’s success at instilling an eye drop. We included some factors in our logistic regression models that may affect dexterity (arthritis, stroke). Self-reported tremor was not included in the model because of the small sample size. Although the presence of arthritis or previous stroke was not found to be a significant factor in determining success in the model, when tested by univariate analysis, arthritis seemed to have a more detrimental effect on the success of glaucoma subjects than on that of the retina subjects ( P = .01).
| Factor | Success, n (%) | Odds Ratio | 95% Confidence Interval | P Value |
|---|---|---|---|---|
| Gender | ||||
| Female | 168/241 (69.1) | 1.0 | — | — |
| Male | 119/165 (72.1) | 1.07 | 0.68 to 1.69 | .77 |
| Race | ||||
| Nonwhite | 62/84 (73.8) | 1.0 | — | — |
| White | 224/321 (69.8) | 1.03 | 0.58 to 1.83 | .92 |
| Type of impairment | ||||
| Unilateral | 140/197 (70.1) | 1.0 | — | — |
| Bilateral | 146/208 (70.2) | 1.12 | 0.72 to 1.78 | .60 |
| Type of disease | ||||
| Retina | 143/204 (70.4) | 1.0 | — | — |
| Glaucoma | 144/205 (70.9) | 0.92 | 0.58 to 1.48 | .75 |
| Previous eye surgery | ||||
| No | 73/95 (76.8) | 1.0 | — | — |
| Yes | 213/310 (68.7) | 0.77 | 0.44 to 1.36 | .36 |
| Age (years) | ||||
| < 70 | 130/157 (82.8) | 1.0 | — | — |
| ≥ 70 | 157/249 (63.1) | 0.34 | 0.20 to 0.58 | .00 |
| History of stroke/TIA | ||||
| No | 255/360 (70.8) | 1.0 | — | — |
| Yes | 28/42 (66.7) | 0.92 | 0.45 to 1.86 | .81 |
| History of arthritis | ||||
| No | 187/265 (70.6) | 1.0 | — | — |
| Yes | 96/137 (70.1) | 1.24 | 0.77 to 2.02 | .38 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
