To compare visual acuity outcomes, vision-related quality of life, and complications related to cataract surgery in eyes with and without glaucoma.
Retrospective cohort study.
Cataract surgery outcomes in cases with and without glaucoma from the Veterans Affairs Ophthalmic Surgical Outcomes Data Project were compared.
We identified 608 glaucoma cases and 4306 controls undergoing planned cataract surgery alone. After adjusting for age, pseudoexfoliation, small pupil, prior ocular surgery, and anterior chamber depth, we found that glaucoma cases were more likely to have posterior capsular tear with vitrectomy (odds ratio [OR] 1.8, P = .03) and sulcus intraocular lens placement (OR 1.65, P = .03) during cataract surgery. Glaucoma cases were more likely to have postoperative inflammation (OR 1.73, P < .0001), prolonged elevated intraocular pressure (OR 2.96, P = .0003), and additional surgery within 30 days (OR 1.92, P = .03). Mean best-corrected visual acuity (BCVA) and Visual Function Questionnaire (VFQ) scores significantly improved after cataract surgery in both groups ( P < .0001), but there were larger improvements in BCVA ( P = .01) and VFQ composite scores ( P < .0001) in the nonglaucoma vs the glaucoma group. A total of 3621 nonglaucoma cases (94.1%) had postoperative BCVA 20/40 or better, compared to 466 glaucoma cases (89.6%) ( P = .0003).
Eyes with glaucoma are at increased risk for complications and have more modest visual outcomes after cataract surgery compared to eyes without glaucoma. Despite this, glaucoma patients still experience significant improvement in vision-related outcomes after cataract extraction. Further study is needed to explore potential factors that influence cataract surgery outcomes in glaucomatous eyes.
Cataracts and glaucoma are leading causes of visual impairment worldwide and these age-related conditions often coexist. The medical treatment for glaucoma has been associated with cataract development and increased rates of cataract surgery. Cataracts and glaucoma are particularly relevant in the aging veteran population in the United States. Eighty-five percent of all ophthalmic surgical procedures in the Veterans Health Administration system are cataract surgeries. Other than refractive error, glaucoma and cataracts are leading eye diagnoses in veterans undergoing routine eye examination.
It can be difficult to tease out the relative contribution of cataracts and glaucoma on visual function. Visual impairment and loss of quality of life associated with glaucoma and cataract have been well documented. Cataracts can have a significant impact on quality of life among patients with glaucoma of all severity levels. Cataract extraction can improve visual function and quality of life among glaucoma patients. In the Collaborative Initial Glaucoma Treatment Study, vision-related quality of life in glaucoma patients acutely improved after cataract surgery. A study from Sweden showed that 73% of glaucoma patients (8 of 11) with low vision undergoing cataract surgery had unchanged or better vision at 5 years after surgery.
The role of cataract surgery in the management of glaucoma is evolving, particularly in the context of intraocular pressure (IOP) control. IOP lowering after cataract surgery has been observed in patients with ocular hypertension, pseudoexfoliation (PXF), and angle closure glaucoma (ACG). However, there are some patients who have worse IOP control after cataract surgery and the need for glaucoma medications can increase after cataract extraction. Success of glaucoma filtering surgery is also impacted by cataract surgery to varying degrees, depending on the timing of the procedures relative to each other. There is less information on complications unrelated to IOP in glaucoma patients undergoing cataract surgery. Studies that specifically address complications of cataract surgery in glaucoma patients often focus on patients with PXF and ACG because these conditions are associated with zonular instability and shallow anterior chamber (AC) depth. Evaluating the overall risks of cataract surgery in glaucoma patients is very relevant in weighing the potential benefits related to visual and IOP outcomes.
The Veterans Affairs Ophthalmic Surgery Outcomes Database Pilot Program provides a unique opportunity to study and compare clinical outcomes of cataract surgery performed in a large population of veterans with and without glaucoma. We compare perioperative complications and pre- and postoperative visual acuity, as well as vision-related quality of life as measured by the National Eye Institute–Visual Function Questionnaire (NEI-VFQ), in patients with and without glaucoma undergoing cataract surgery in the Veterans Affairs Healthcare System.
The original Ophthalmic Surgery Outcomes Database Pilot Project was a quality improvement program designed for comparative assessment and enhancement of the quality of cataract surgery across the Veterans Affairs System. Specific data elements, predetermined by the Ophthalmic Surgery Outcomes Database Committee, were abstracted from preoperative, intraoperative, and postoperative records of patients undergoing cataract surgery at 5 different Veterans Affairs hospitals in the United States. After the quality improvement program was completed, the National Surgery Office within the Veterans Affairs Administration compiled the data generated by the Ophthalmic Surgery Outcomes Program into a de-identified database. It is the de-identified database that we have retrospectively reviewed for this paper with approval from the Veterans Affairs Boston Healthcare System Institutional Review Board and a Data Use Agreement.
Case records of patients undergoing cataract surgery that had glaucoma status noted, along with both pre- and postoperative NEI-VFQs filled out, were included for analysis. Glaucoma status was decided based on previous history of glaucoma of any type. Details on optic nerve status and glaucoma therapy were not collected for the database, but cases on more than 3 glaucoma medications or with a history of prior glaucoma surgery were identified in the dataset and defined as high-risk glaucoma cases for this study. All cases, including the high-risk glaucoma cases, underwent cataract surgery alone as a planned procedure. The 25-item NEI-VFQ was administered to patients pre- and postoperatively within 30–90 days of surgery. The questionnaire used for this study has been tested for validity and reliability in different populations.
Outcomes used for this analysis included postoperative best-corrected visual acuity (BCVA), VFQ scores, and perioperative events. BCVA values were converted to the logarithm of the minimal angle of resolution (logMAR) scale prior to computation for quantitative analysis. We analyzed pre- and postoperative VFQ subscales of general vision, ocular pain, near activities, distance activities, and vision-specific subscales of social functioning, mental health, role difficulties and dependency, driving, color vision, peripheral vision, and composite score. Intraoperative events that were analyzed were intraocular lens (IOL) placement, use of pupillary expansion, placement of capsular tension ring, corneal wound burn, corneal trauma, intraoperative floppy iris syndrome (IFIS), iris prolapse, iris trauma, zonular dehiscence with vitrectomy, zonular dehiscence without vitrectomy, anterior capsular tear, posterior capsular (PC) tear with vitrectomy, PC tear without vitrectomy, choroidal effusion, choroidal hemorrhage, and conversion to large incisional surgery.
Postoperative events recorded in the database were based on a retrospective review of all postoperative clinic visits within the 30-day period. Postoperative events that were analyzed include: corneal stromal edema at 1 month, intraocular inflammation at 1 month, refraction outside of the ±0.75 diopter target range at 1 month, cystoid macular edema (CME), clinically significant macular edema (CSME), retained lenticular material with repeat surgery within 30 days, retained lenticular material without repeat surgery within 30 days, intraocular pressure of less than 5 mm Hg for more than 1 week, intraocular pressure of more than 25 mm Hg for more than 1 week, retinal detachment, infectious and sterile endophthalmitis, further surgery in operative eye within 30 days, hospital admission postoperatively, and death within 30 days. Complications including CME, CSME, endophthalmitis, retained lens material, and retinal detachment were noted if such events occurred at any time within 30 days after surgery.
Data were analyzed using JMP Pro Version 11.2.1. (SAS Institute Inc., Cary, NC, 1989-2013) and Stata 11 (StataCorp. 2009. Stata Statistical Software: Release 11. College Station, TX: StataCorp LP). Fisher exact test was used to test statistically significant differences (α = 0.05) for categorical variables, Wilcoxon signed rank and rank-sums test for continuous variables, nominal logistic regression for adjusted odds ratios (OR) with likelihood test, and McNemar test for binomial pre- and postoperative BCVA. Missing values were not included for statistical analysis.
Of the 4923 eyes included in the Ophthalmic Surgery Outcomes database, 4914 eyes met our inclusion criteria: 608 (12.4%) had a documented history of glaucoma and the remaining 4306 cases were used as controls. Of those with glaucoma, 164 eyes were on more than 3 IOP-lowering medications or had prior incisional glaucoma surgery and were classified as high-risk glaucoma cases. Table 1 summarizes the baseline clinical and ophthalmic characteristics of this cohort. Overall, the glaucoma group had a higher mean age (74.5 ± 9.5 vs 70.0 ± 9.6, P < .0001) and lower body mass index (28.2 ± 5.5 vs 29.3 ± 6.3, P = .0003) compared to the nonglaucoma group. More cases with glaucoma were on selective alpha-blockers (98/608, 16.1% glaucoma vs 528/4299, 12.3% nonglaucoma, P = .009). A larger proportion of cases with glaucoma compared to those without glaucoma had small pupils (102/604, 16.9% vs 389/4300, 9.1%), PXF (59/608, 9.7% vs 64/4289, 1.5%), and prior ocular surgery (88/608, 14.5% vs 247/4306, 5.7%) ( P < .0001 for all). Glaucoma cases also had a shallower mean AC depth compared to nonglaucoma cases (3.08 ± 0.48 vs 3.17 ± 0.45, P < .0001). There were 1705 nonglaucoma cases (40.91%) with preoperative BCVA better than 20/40, compared to 215 glaucoma cases (36.5%, P = .04).
|Glaucoma (N = 608)||No Glaucoma (N = 4306)||P Value a|
|Age (mean ± SD)||74.5 ± 9.5||70.0 ± 9.6||<.0001|
|Preoperative BCVA, logMAR (median)||0.39||0.39|
|Preoperative BCVA (mean ± SD)||0.55 ± 0.47||0.52 ± 0.43||.10|
|Preoperative BCVA 20/40 or better||215 (36.5%)||1705 (40.9%)||.04|
|Small pupil||102 (16.9%)||389 (9.1%)||<.0001|
|Diabetes||259 (42.7%)||1791 (41.7%)||.66|
|Hypertension||489 (80.6%)||3402 (79.3%)||.48|
|1||0 (0 %)||14 (0.33%)|
|2||88 (14.5%)||772 (18%)|
|3||477 (78.6%)||3223 (75.1%)|
|4||42 (6.9%)||282 (6.6%)|
|BMI||28.2 ± 5.5||29.3 ± 6.3||.0003|
|Selective alpha blocker||98 (16.1%)||528 (12.3%)||.009|
|Nonselective alpha blocker||95 (15.6%)||571 (13.3%)||.11|
|Pseudoexfoliation syndrome||59 (9.7%)||64 (1.5%)||<.0001|
|Age-related macular degeneration||59 (9.8%)||337 (7.9%)||.11|
|Prior ocular surgery||88 (14.5%)||247 (5.7%)||<.0001|
|Anterior chamber depth||3.08 ± 0.48||3.17 ± 0.45||<.0001|
Table 2 summarizes intraoperative events during cataract surgery in eyes with and without glaucoma. Univariate analysis showed that more eyes with glaucoma required sulcus IOL placement (40/607, 6.6% glaucoma vs 157/4297, 3.6% nonglaucoma, P = .001) and use of a pupil expansion device (88/608, 14.5% glaucoma vs 341/4299, 7.9% nonglaucoma, P < .0001). The rate of PC tear with vitrectomy was also higher in cases with glaucoma (30/607, 4.9%) vs cases without glaucoma (116/4302, 2.7%; P = .005). After adjusting for age, small pupil, PXF, AC depth, and prior ocular surgery, the likelihood of sulcus IOL placement (OR 1.65, P = .03) and PC tear with vitrectomy (OR 1.80, P = .03) was still higher in the glaucoma group compared to the nonglaucoma group. Interestingly, the multivariate analysis adjusting for age, small pupil, PXF, AC depth, and prior ocular surgery showed that glaucoma cases were less likely to sustain iris prolapse or trauma (OR 0.47, P = .03 and OR 0.26, P = .03, respectively), even though there was no difference between the groups in the univariate analysis. Mean operative times were similar in the glaucoma group (43.2 ± 21.1 min) and nonglaucoma group (45.2 ± 25.4 min, P = .61), even though attending surgeons performed a larger percentage of glaucoma cases (76/608, 12.5%) compared to those without glaucoma (323/4304, 7.5%, P < .0001).
|Glaucoma||No Glaucoma||P Value a||Adjusted Odds Ratio||95% CI||P Value b|
|N||Total N||%||N||Total N||%|
|Pupil expansion device||88||608||14.5||341||4299||7.9||<.0001||1.09||0.77||1.54||.59|
|Corneal wound burn||2||608||0.3||4||4303||0.1||.16||3.73||0.46||24.24||.19|
|Intraoperative floppy iris syndrome||39||608||6.4||214||4303||5||.14||0.97||0.63||1.43||.87|
|Zonular dehiscence with vitrectomy||4||607||0.7||22||4296||0.5||.55||1.74||0.48||4.95||.36|
|Zonular dehiscence without vitrectomy||7||607||1.1||38||4296||0.9||.49||0.88||0.28||2.33||.81|
|Anterior capsular tear||25||608||4.1||132||4302||3.1||.17||1.17||0.66||1.96||.57|
|Posterior capsular tear with vitrectomy||30||607||4.9||116||4302||2.7||.005||1.8||1.06||2.95||.03|
|Posterior capsular tear without vitrectomy||5||607||0.8||49||4302||1.1||.68||0.6||0.18||1.55||.32|
|Conversion to large incisional surgery||5||607||0.8||19||4288||0.4||.20||1.93||0.58||5.33||.26|
Postoperative events in glaucoma and nonglaucoma cases are summarized in Table 3 . Univariate analysis showed that the following complications occurred significantly more frequently in glaucoma cases compared to nonglaucoma cases: corneal stromal edema at 1 month, inflammation at 1 month, 30-day refraction outside 0.75 diopters from target, IOP <5 mm Hg or >25 mm Hg for more than 1 week, and further surgery in the operative eye within 30 days. After adjusting for age, small pupil, PXF, AC depth, and prior ocular surgery, the following events were still more likely to occur in cases with glaucoma compared to those without: inflammation at 1 month (OR 1.73, P < .0001), IOP >25 mm Hg for >1 week (OR 2.96, P = .0003), and further surgery in the operative eye within 30 days (OR 1.92, P = .03). The difference between the groups for the 30-day refraction outside of 0.75 diopter from target just met statistical significance in the multivariate analysis (OR 1.31, P < .0499).
|Glaucoma||No Glaucoma||P Value a||Adjusted OR b||95% CI||P Value c|
|N||Total N||%||N||Total N||%|
|Corneal stromal edema||44||592||7.4||222||4145||5.4||.0449||1.15||0.76||1.70||.49|
|Inflammation at 1 month||130||591||22||570||4130||13.8||<.0001||1.73||1.34||2.22||<.0001|
|30-day refraction < or >0.75 diopter of target refraction||101||559||18.1||574||3892||14.7||.04||1.31||1.00||1.70||.0499|
|Cystoid macular edema||21||597||3.5||90||4172||2.2||.06||1.36||0.76||2.31||.28|
|Retained lenticular material with repeat surgery within 30 days||6||595||1||25||4193||0.6||.27||1.86||0.65||4.59||.23|
|Retained lenticular material without repeat surgery within 30 days||15||595||2.5||66||4193||1.6||.12||1.78||0.87||3.35||.11|
|IOP <5 mm Hg for >1 week||4||595||0.7||4||4154||0.1||.01||4.38||0.54||28.54||.15|
|IOP >25 mm Hg for >1 week||25||595||4.2||48||4147||1.7||<.0001||2.96||1.66||5.10||.0003|
|Further surgery in operative eye within 30 days||21||602||3.5||73||4239||1.7||.006||1.92||1.05||3.35||.03|
|Readmission within 30 days||6||599||1.00||52||4242||1.2||.84||0.72||0.27||1.62||.45|
|Death within 30 days||1||600||0.2||7||4245||0.7||1.00||0.48||0.02||3.89||.53|
When subgroup analysis was performed on the high risk glaucoma cases, we found no significant differences between the rates of complications in the high-risk and low-risk glaucoma cases, except that 5 high-risk glaucoma cases out of 164 (3%) had intraoperative floppy iris syndrome compared to 34 low-risk glaucoma cases out of 444 (7.7%) ( P = .04). In the univariate and multivariate analysis, there was no significant difference in risk for IFIS in either the high-risk or overall glaucoma group compared to nonglaucoma cases.
Visual outcomes at 1 month after cataract surgery are summarized in Table 4 . Mean logMAR BCVA significantly improved after cataract surgery in the glaucoma group (0.55 ± 0.47 preoperative vs 0.16 ± 0.35 postoperative) and the nonglaucoma group (0.52 ± 0.43 preoperative vs 0.07 ± 0.22 postoperative, P < .0001 for both). When analyzed separately, even cases with high-risk glaucoma showed improvement in mean BCVA after surgery (0.62 ± 0.59 vs 0.20 ± 0.32, P < .0001). However, analysis of variance showed that the improvement in mean BCVA was more pronounced in the nonglaucoma group compared to the glaucoma group ( P = .012). There were also more patients in the nonglaucoma group (3621/3848, 94.1%) who achieved 20/40 or better BCVA compared to the glaucoma group (466/520, 89.6%; P = .0003).
|Glaucoma||No Glaucoma||P Value b|
|Preoperative||Postoperative||P Value a||Preoperative||Postoperative||P Value a|
|BCVA 20/40 or better, n (%)||215/589 (36.5%)||466/520 (89.6%)||1705/4168 (40.9%)||3621/3848 (94.1%)||.0003 c|
|BCVA (median)||0.40||0.10||0.40||0||<.0001 d|
|BCVA (mean ± SD)||0.55 ± 0.47||0.16 ± 0.35||<.0001||0.52 ± 0.43||0.07 ± 0.22||<.0001||<.0001|
|VFQ composite score (mean ± SD)||64.29 ± 20.49||80.18 ± 19.14||<.0001||63.15 ± 19.74||84.72 ± 16.33||<.0001||<.0001|