Abstract
Purpose:
Phacoemulsification has been linked to lowered intraocular pressure (IOP) in patients with glaucoma, ocular hypertension, anatomic narrow angles, and in glaucoma suspects (henceforth, “glaucoma”), but the reported magnitude of change has varied.
Design:
Retrospective cohort study.
Participants and Controls:
Patients with glaucoma, June 2010 to May 2015, who underwent phacoemulsification (“surgical”) were matched one-to-one to patients who did not (“non-surgical”) on age, gender, type of glaucoma, baseline IOP, and number and type of glaucoma medications.
Methods:
Electronic medical record information was used to compare the matched surgical and non-surgical groups.
Main Outcome Measures:
Change in IOP, change in number of glaucoma medications, and likelihood of a glaucoma procedure within 36 months after phacoemulsification. IOP measures were obtained from Goldmann applanation tonometry when available (45%), and otherwise from Icare, Tono-pen, noncontact tonometry, and pneumotonometry.
Results:
Among 16,169 matched pairs, average IOP after the index date was lower in the surgical than non-surgical group throughout follow-up to 36 months. The difference was greatest during 1-18 months, during which IOP increased by 0.22 mm Hg from 16.49 mm Hg in the average non-surgical patient and decreased by 0.99 mm Hg from 16.50 mm Hg in the average surgical patient (difference-in-difference, 1.21 with 95% CI 1.12-1.30 mm Hg). The difference-in-difference was greatest for patients with ocular hypertension (2.00 mm Hg) and for patients with preoperative IOP ≥20 (2.46 mm Hg). A subgroup analysis with matched patients using only Goldmann applanation tonometry found similar results. By 30-36 months, 5% (95% CI 4-6%) fewer surgical patients used an ophthalmic medication. In the surgical group the odds of selective laser trabeculoplasty was reduced in patients with ocular hypertension (odds ratio [OR], 0.27; 95% CI 0.10-0.74) or glaucoma suspect (OR, 0.31; 95% CI 0.20-0.47), while the odds of glaucoma surgery were elevated in surgical patients with primary open angle glaucoma (OR, 1.48; 95% CI 1.08-2.01).
Conclusion:
In this community-based study, the association of phacoemulsification for cataract with IOP reduction was lower than in past referral-based studies. Surgeons should expect to reduce intraocular pressure approximately 1-2 mmHg with phacoemulsification in patients with preoperative IOP less than 20 mmHg.
Keywords: Glaucoma, Cataract Surgery, Electronic Health Record, Intraocular Pressure
Precis
In this community-based study of glaucoma patients and glaucoma suspects with mid-normal IOP, the association of phacoemulsification for cataract with IOP reduction was lower than reported in past referral-based studies. Surgeons should expect to reduce intraocular pressure approximately 1-2 mmHg with phacoemulsification.
Glaucoma is characterized by progressive retinal ganglion cell loss, vision loss, and potential blindness.1 Although glaucoma cannot be prevented, it can be treated using medication, lasers, and surgery to reduce the intraocular pressure (IOP) and related progressive loss of visual field. Numerous reports, including systematic reviews, meta-analyses, and randomized controlled trials, give evidence that cataract surgery can reduce IOP among glaucoma patients.2–10 This evidence has been widely accepted, with nearly half of American Glaucoma Society members surveyed responding that they would treat a patient with primary open angle glaucoma and visually significant cataract using phacoemulsification alone .11 However, the randomized controlled trials and observational studies that provided evidence for this association were quite small and based in referral centers, and used varying time frames to measure outcomes, while observational studies often lacked a control group. To strengthen the evidence, we conducted a controlled study to assess change in IOP after cataract surgery among patients with glaucoma, ocular hypertension, anatomic narrow angles, and in glaucoma suspects (henceforth, “glaucoma”). We hypothesized the following: (i) phacoemulsification for cataract is associated with reduced IOP among glaucoma patients, (ii) the association differs by type of glaucoma, (iii) the association is strongest among glaucoma patients with higher preoperative IOP as evidenced in past reports,3, 4, 7 and (iv) phacoemulsification reduces the frequency of subsequent glaucoma surgery. Secondary outcomes included use of glaucoma medications and use of glaucoma laser procedures.
METHODS
Setting and Population
Kaiser Permanente Northern California is a fully integrated healthcare delivery system. Most patients receive capitated, comprehensive services. With respect to cataract surgery, the health plan’s ophthalmologists perform clear cornea phacoemulsification using standardized phacoemulsification machines (Alcon, Irvine) and intraocular lenses (Alcon, Irvine).
The study included adult Kaiser Permanente members, ≤89 years, enrolled during June 2010 through May 2015 with at least 1 year of baseline enrollment and an inpatient or outpatient diagnosis of glaucoma (ICD-9 diagnosis code 365) after excluding glaucoma related to a congenital anomaly (365.14) or associated ocular disorders (365.5, 365.6). We identified glaucoma diagnoses starting in 2008 to ensure capture of prevalent disease, and we identified cataract surgeries starting in 1996 to ensure no history of cataract surgery before the start of the study. Eligible glaucoma patients were divided into two groups, those who underwent cataract surgery during the study period (“surgical”) and those who did not (“non-surgical”).
Among surgical patients, we included the first simple phacoemulsification for cataract using procedure codes (Supplemental Material).12 We excluded patients who, during the baseline period, did not have a preoperative measurement of IOP or had undergone selective laser trabeculoplasty or glaucoma surgery in an inpatient or outpatient setting (selective laser trabeculoplasty, CPT4 code 65855; Baerveldt shunt, 66180; Ahmed glaucoma shunt, Kaiser internal codes; trabeculectomy, CPT4 codes 65850, 66170, 66172, 66183, and 66185 and Kaiser internal codes). Minimally invasive glaucoma procedures were not commonly performed and were not included in the analysis. The date of cataract surgery was assigned as the index date.
Non-surgical patients were matched one-to-one to surgical patients on age, gender, glaucoma subtype, preoperative IOP, and number and classes of ophthalmic glaucoma medications dispensed during the baseline period. An index date was assigned to each non-surgical patient corresponding to the surgery date in the matched surgical patient. The non-surgical patient was required to be enrolled in the Kaiser health plan and without a selective laser trabeculoplasty during the one-year baseline period before the index date.
Patients were classified into glaucoma subtypes using the most specific diagnosis received by the patient during the baseline period using the hierarchy provided in the Supplemental Material. For the Goldmann subgroup analysis, we identified and matched patients whose IOP measures were obtained using Goldmann applanation tonometry.
Data Collection
The primary outcome was change in IOP before and after the index date. IOP was obtained from semi-structured fields in the electronic medical record using natural language processing.13 When multiple IOP measures were available, we prioritized Goldmann applanation tonometry, Icare (Raleigh, North Carolina), and Tono-pen (Reichert Technologies, Depew, New York), using other methods (noncontact tonometry, pneumotonometry) only when a more precise method had not been used. Preoperative IOP was classified into quartiles (≤13, 14-16, 17-19, ≥20 mm Hg). Before the index date, we used the single most recent IOP measurement, while after the index date, we used the single measurement recorded first during the period from 1 to 36 months after the index date.
Secondary outcomes included the number of drug classes and the use of glaucoma procedures used during the 36 months after the index date. The number of ophthalmic drug classes used by the patient during the baseline period was counted as 0, 1, 2, and ≥3. Drug classes included prostaglandin analogues, beta-adrenergic antagonists, alpha-adrenergic agonists, carbonic anhydrase inhibitor, and cholinergic agents (supplemental material). Use of two drugs from the same class represented switching within the class and therefore were counted as a single class. Ophthalmic medications that were dispensed as combinations were counted as two drug classes. We separately classified oral carbonic anhydrase inhibitors for patients who had at least two fills. Regarding glaucoma procedures, selective laser trabeculoplasty and glaucoma surgery performed during the 36 months after the index date were captured using the codes provided in the supplemental material.
Baseline demographic factors, systemic comorbidities, and ocular comorbidities, and procedures were obtained from membership, inpatient, and outpatient data recorded during the year before surgery. Anti-VEGF injections were obtained from pharmacy records. Charlson comorbidities were obtained from diagnostic and procedure codes recorded during the year before surgery.14
Statistical Analysis
Change in IOP was treated as a continuous outcome variable. We first assessed the change in IOP during 1 to 36 months after the index date by examining change in IOP during 6-month intervals (1-6, 7-12, … 30-36). Our matched pairs design ensured balance in the matching covariates between the surgical and non-surgical group, allowing for direct comparison of the outcomes between surgical and non-surgical patients. We report the average difference in the change in IOP between the surgical and non-surgical patients (average difference-in-difference), and the corresponding 95% confidence interval (CI). We also performed a multivariable analysis in which we used a linear regression models to fit the outcome of change in IOP in relation to surgery after adjusting for additional patient-level variables to determine the potential for additional confounding beyond the matching covariates. In addition, we performed subgroup analyses based on glaucoma subtype, preoperative IOP, and preoperative number of medication classes, and we conducted subgroup analysis using only Goldmann measures of IOP. For analysis of the secondary outcomes of change in number of glaucoma medication classes and likelihood of a glaucoma procedure, we assessed number of medication classes over time at 6 months, 12 months, 18 months, 24, and 36 months post-index.
For each glaucoma procedure, the odds ratio (OR) and corresponding 95% CI for the association between phacoemulsification and procedure was calculated using 2x2 contingency tables. All analyses were performed using SAS 9.4.
Institutional Review Board (IRB)/Ethics Committee approval was obtained. The research adhered to the tenets of the Declaration of Helsinki, and the study was HIPAA compliant. Consent was not required in this retrospective study.
RESULTS
We identified 19,932 patients aged 20-89 years with glaucoma who underwent phacoemulsification for cataract and had ≥1 year of enrollment in the Kaiser health plan before their surgery during June 2010 to May 2015. After exclusions, detailed in the supplemental material, 81% (N=16,169) of the surgical patients remained eligible for the study and could be matched with a non-surgical patient. The number of patients with Goldmann measurements before and within 36 months after surgery was 8,977. Because of the matching, patients and controls were identical or nearly identical on age, gender, type of glaucoma, baseline IOP, and number and type of glaucoma medications. Surgical patients were slightly more likely to be white, had more complete information on body mass index and smoking history, and had somewhat more codes recorded for Charlson and ocular comorbidities (Table 1). About one-third used a glaucoma medication, (average, 1.5 in each group). In addition, Table 1 shows that 46% of all patients were glaucoma suspects. Glaucoma suspects were less likely to have measurements by Goldmann applanation tonometry (Table 1). The time from the baseline IOP measurement to the index date was shorter in surgical patients (months: average 2.0, standard deviation [SD] 2.2) than non-surgical patients (average 5.0, SD 3.5). The time from the index date to the post-index measurement was also shorter in surgical patients (months: average 7.2, SD 8.6) than in non-surgical patients (9.9, SD 8.2).
Table 1.
Baseline characteristics of matched pairs followed for 36 months after index, of which surgical patients underwent phacoemulsification for cataract and non-surgical patients did not. Kaiser Permanente Northern California, 2010-15.*
| All IOP measurements (N=16,169), % | Goldmann only (N=7,283), % | ||||
|---|---|---|---|---|---|
| Characteristic | Coding | Non-surgical | Surgical | Non-surgical | Surgical |
| Glaucoma subtype | Ocular hypertension | 6.3 | 6.3 | 6.2 | 6.2 |
| Glaucoma suspect | 45.6 | 45.6 | 37.9 | 37.9 | |
| Primary open angle glaucoma | 21.0 | 21.0 | 26.8 | 26.8 | |
| Normal tension glaucoma | 5.4 | 5.4 | 6.9 | 6.9 | |
| Anatomic narrow angle | 7.3 | 7.3 | 6.4 | 6.4 | |
| Primary angle closure glaucoma | 3.1 | 3.1 | 3.3 | 3.3 | |
| Pseudoexfoliation glaucoma | 1.1 | 1.1 | 1.1 | 1.1 | |
| Unspecified | 10.2 | 10.2 | 11.4 | 11.4 | |
| Surgery year | 2010-2011 | 29.8 | 29.8 | 31.4 | 31.4 |
| 2012-2013 | 40.7 | 40.7 | 40.4 | 40.4 | |
| 2014-2015 | 29.5 | 29.5 | 28.2 | 28.2 | |
| Age, years | ≤69 | 32.5 | 32.0 | 31.0 | 30.3 |
| 70-79 | 44.6 | 43.6 | 45.9 | 45.2 | |
| ≥80 | 22.9 | 24.5 | 23.1 | 24.6 | |
| Gender | Female | 58.8 | 58.8 | 58.6 | 58.6 |
| Male | 41.2 | 41.2 | 41.4 | 41.4 | |
| Race/ethnicity | African American | 12.4 | 8.4 | 8.3 | 8.9 |
| Asian American | 16.2 | 17.4 | 14.9 | 14.5 | |
| Hispanic | 10.9 | 9.9 | 10.5 | 8.5 | |
| White | 58.9 | 62.9 | 63.9 | 66.8 | |
| Other | 1.6 | 1.4 | 2.4 | 1.3 | |
| Body mass index, kg/m2 | 12-18.4 | 1.6 | 1.6 | 1.5 | 1.7 |
| 18.5-24.9 | 30.2 | 31.9 | 30.1 | 32.2 | |
| 25-29.9 | 34.0 | 37.2 | 34.6 | 37.3 | |
| ≥30+ | 25.8 | 28.6 | 25.0 | 28.1 | |
| Missing | 8.4 | 0.8 | 8.8 | 0.8 | |
| Smoking | Ever | 34.7 | 42.7 | 35.1 | 43.8 |
| Never | 48.8 | 50.6 | 48.4 | 48.2 | |
| Missing | 16.5 | 6.7 | 16.5 | 8.0 | |
| Charlson comorbidity index** | 0 | 47.6 | 40.4 | 49.4 | 41.3 |
| 1 | 19.0 | 19.6 | 18.6 | 19.1 | |
| 2 | 13.8 | 16.1 | 13.9 | 16.1 | |
| ≥3 | 19.5 | 23.9 | 18.1 | 23.4 | |
| Ocular comorbidities | Age-related macular degeneration | 8.6 | 11.3 | 8.6 | 11.6 |
| Corneal disorders | 2.5 | 3.9 | 2.6 | 3.8 | |
| Epiretinal membrane | 3.2 | 5.1 | 3.4 | 5.1 | |
| Diabetic retinopathy | 5.5 | 8.4 | 4.6 | 8.0 | |
| Oral glaucoma medication | Carbonic anhydrous inhibitor | 0.1 | 0.2 | 0.3 | 2.7 |
| Ophthalmologic glaucoma medication** | Prostaglandin analogue | 22.8 | 22.8 | 29.9 | 29.9 |
| Beta-1-adrenergic antagonist | 16.0 | 16.0 | 20.8 | 20.8 | |
| Alpha adrenergic agonist | 5.2 | 5.2 | 6.7 | 6.7 | |
| Carbonic anhydrase inhibitor | 5.5 | 5.6 | 7.3 | 7.3 | |
| Cholinergic agent | 0.1 | 0.1 | 0.1 | 0.1 | |
| Number of ophthalmic glaucoma medications | 0 | 67.2 | 67.1 | 57.7 | 57.6 |
| 1 | 21.2 | 21.2 | 26.9 | 26.9 | |
| 2 | 7.6 | 7.6 | 9.9 | 9.9 | |
| ≥3 | 4.1 | 4.1 | 5.5 | 5.6 | |
Non-surgical patients were matched 1:1 to surgical patients on age, gender, type of glaucoma, preoperative IOP, and number and type of glaucoma medications using the closest possible match.
Total does not equal 100.0 because of rounding.
The reduction in IOP observed in surgical patients was greatest during the first 18 months after the index date (Figure 1), and we therefore focused our analysis of the difference-in-difference on the 12,240 matched pairs with IOP measurements recorded through 18 months. IOP increased by 0.22 mm Hg from the baseline measurement to the post-index measurement in the average non-surgical patient and decreased by 0.99 mm Hg in the average surgical patient for a difference-in-difference of 1.21 (95% CI 1.20-1.12) using all IOP measurements and 1.19 mm Hg in the Goldmann subgroup (95% CI 1.06-1.32) (Table 2). In multivariable analysis, the difference-in-difference was 1.12 mm Hg (1.01-1.23 mm Hg) in all patients and 1.06 mm Hg (0.91-1.21 mm Hg) in the Goldmann subgroup (supplemental Table 2). The reduction in IOP after phacoemulsification was evident for all glaucoma subtypes but was largest for ocular hypertension, anatomic narrow angle, and pseudoexfoliation glaucoma. It was smallest for normal tension glaucoma (Table 2 and Figure 2). The reduction in IOP observed in surgical patients showed a monotonic relationship with preoperative IOP and with the number of ophthalmic glaucoma classes dispensed in the baseline period (no medication, difference-in-difference of −1.45 with 95% CI −1.63, −1.28; 3+ medications, difference-in-difference of −0.60 (−1.16, −0.04). It was also associated with the patient’s age.
Figure 1.
IOP (mm Hg) in relation to months since index in 16,169 matched pairs with glaucoma, of which surgical patients underwent phacoemulsification for cataract and non-surgical patients did not. Kaiser Permanente Northern California, 2010-15.*
*Non-surgical patients were matched to surgical patients on age, gender, type of glaucoma, before index IOP, and number and type of glaucoma medications using the closest possible match. The circle represents IOP before the index date, which was matched in surgical and non-surgical patients. The bar above the circle is the higher IOP observed in non-surgical patients after the index date. The bar below the circle is the lower IOP observed in surgical patients after the index date.
**From the before index IOP measurement to the index date was 2.0 mm Hg (SD 2.2) months in surgical patients and 5.0 mm Hg (3.5) months in non-surgical patients. From the index date to the post-index measurement was 7.2 mm Hg (8.6) months in surgical patients and 9.9 mm Hg (8.2) months in non-surgical patients.
Table 2.
Change in IOP (mm Hg) (postoperative minus preoperative) in patients with glaucoma and post-operative IOP measurement recorded within 18 months of the surgery or index date, of which surgical patients underwent phacoemulsification for cataract and non-surgical patients did not. Kaiser Permanente Northern California, 2010-15.*
| Mean (standard deviation) |
|||||||
|---|---|---|---|---|---|---|---|
| Non-surgical (N=12,240) | Surgical (N=12,240) | Difference-in-difference (95% CI) | |||||
| Subgroup | N | IOP before index | IOP after index | IOP before index | IOP after index | All IOP measures (N=12,240) | Goldmann only (N=7,196) |
| Overall | 12,240 | 16.49 (3.68) | 16.71 (4.06) | 16.50 (3.72) | 15.51 (4.38) | −1.21 (−1.30, −1.12) | −1.19 (−1.32, −1.06) |
| Ocular hypertension | 752 | 18.71 (3.55) | 19.70 (3.79) | 18.66 (3.63) | 17.65 (4.75) | −2.00 (−2.45, −1.56) | −1.85 (−2.47, −1.22) |
| Glaucoma suspect | 5,175 | 16.55 (3.56) | 16.62 (3.75) | 16.55 (3.57) | 15.34 (4.10) | −1.29 (−1.43, −1.15) | −1.35 (−1.56, −1.13) |
| Primary open-angle glaucoma | 2,883 | 16.37 (3.76) | 16.45 (4.23) | 16.39 (3.84) | 15.46 (4.41) | −1.01 (−1.22, −0.79) | −0.88 (−1.14, −0.61) |
| Normal tension glaucoma | 748 | 13.82 (2.67) | 13.80 (2.93) | 13.86 (2.85) | 13.20 (3.15) | −0.65 (−0.94, −0.36) | −0.55 (−0.92, −0.18) |
| Anatomic narrow angle | 810 | 16.49 (3.49) | 16.60 (3.75) | 16.50 (3.57) | 14.83 (3.85) | −1.78 (−2.12, −1.44) | −1.87 (−2.34, −1.41) |
| Primary angle-closure glaucoma | 412 | 16.72 (3.60) | 16.85 (4.01) | 16.79 (3.77) | 15.42 (4.67) | −1.50 (−2.09, −0.91) | −1.49 (−2.24, −0.75) |
| Pseudoexfoliation glaucoma | 150 | 16.89 (2.79) | 17.38 (4.31) | 16.91 (3.05) | 15.85 (4.19) | −1.55 (−2.52, −0.58) | −1.89 (−3.10, −0.69) |
| Unspecified glaucoma | 1,310 | 16.46 (3.56) | 16.56 (4.02) | 16.47 (3.60) | 15.51 (4.51) | −1.06 (−1.38, −0.75) | −1.10 (−1.50, −0.69) |
| IOP before index ≤13 mm Hg | 2,550 | 11.67 (1.41) | 13.19 (3.00) | 11.61 (1.47) | 12.83 (3.51) | −0.30 (−0.47, −0.12) | −0.28 (−0.51, −0.05) |
| IOP before index IOP 14-16 mm Hg | 3,900 | 15.04 (0.83) | 15.67 (2.99) | 15.04 (0.83) | 14.68 (3.76) | −0.98 (−1.13, −0.83) | −1.03 (−1.21, −0.84) |
| IOP before index IOP 17-19 mm Hg | 3,304 | 17.91 (0.77) | 17.67 (3.11) | 17.92 (0.78) | 16.31 (3.96) | −1.37 (−1.54, −1.20) | −1.48 (−1.72, −1.23) |
| IOP before index IOP ≥20 mm Hg | 2,486 | 21.78 (2.09) | 20.22 (3.92) | 21.83 (2.12) | 17.81 (4.52) | −2.46 (−2.69, −2.22) | −1.99 (−2.34, −1.64) |
| No medications | 7,681 | 16.65 (3.63) | 16.82 (3.97) | 16.66 (3.66) | 15.46 (4.33) | −1.37 (−1.49, −1.25) | −1.45 (−1.63, −1.28) |
| 1 medication | 2,883 | 16.46 (3.72) | 16.45 (3.96) | 16.47 (3.79) | 15.33 (4.16) | −1.12 (−1.32, −0.92) | −1.09 (−1.35, −0.84) |
| 2 medications | 1,078 | 15.92 (3.45) | 16.11 (3.99) | 15.95 (3.56) | 15.16 (4.20) | −0.98 (−0.31, −0.65) | −0.57 (−0.99, −0.14) |
| 3+ medications | 598 | 15.23 (3.63) | 15.54 (4.39) | 15.24 (3.78) | 14.77 (4.55) | −0.78 (−1.25, −0.31) | −0.60 (−1.16, −0.04) |
| Age ≤69 | 3,788 | 16.73 (3.71) | 16.92 (4.06) | 16.71 (3.71) | 15.96 (4.63) | −0.93 (−1.11, −0.75) | −0.96 (−1.21, −0.71) |
| Age 70-84 | 7,463 | 16.40 (3.64) | 16.51 (3.96) | 16.41 (3.72) | 15.12 (4.09) | −1.40 (−1.52, −1.27) | −1.26 (−1.43, −1.10) |
| Age ≥85 | 989 | 15.92 (3.39) | 16.13 (4.11) | 16.12 (3.53) | 14.99 (4.23) | −1.34 (−1.68, −0.99) | −1.53 (−2.00, −1.05) |
Non-surgical patients were matched 1:1 to surgical patients on age, gender, type of glaucoma, preoperative IOP, and number and type of glaucoma medications using the closest possible match.
Figure 2.
IOP (mm Hg) in 12,240 matched pairs with glaucoma, of which surgical patients underwent phacoemulsification for cataract and non-surgical patients did not. Kaiser Permanente Northern California, 2010-15.
*Non-surgical patients were matched to surgical patients on age, gender, type of glaucoma, before index IOP, and number and type of glaucoma medications using the closest possible match. The circle represents IOP before index date, which was matched in surgical and non-surgical patients. The bar above the circle is the higher IOP observed in non-surgical patients after index. The bar below the circle is the lower IOP observed in surgical patients after index.
Abbreviations: OHTN, ocular hypertension; POAG, primary open angle glaucoma; NTG, normal tension glaucoma; ANA, anatomic narrow angle; PACG, primary angle closure glaucoma; PEX, pseudoexfoliation glaucoma.
By 30-36 months, the proportion of patients who received a dispensing of an ophthalmic medication was 5% (95% CI 3-6%) lower in surgical patients than non-surgical patients (Figure 3). By 36 months, 3.10% of non-surgical patients and 2.38% of surgical patients had undergone selective laser trabeculoplasty. Phacoemulsification was associated with a lower risk of selective laser trabeculoplasty in patients with pseudoexfoliation glaucoma (14 non-surgical, 6 surgical) and ocular hypertension (18 non-surgical, 5 surgical), and in glaucoma suspects (87 non-surgical, 27 surgical) (Table 3). By 36 months, 0.56% (n=91) matched controls and 0.84% (n=136) of phacoemulsification patients had undergone glaucoma surgery with either a trabeculectomy or tube shunt. Although the likelihood of glaucoma surgery was rare for most groups (Table 3), it was slightly elevated in surgical patients with primary open angle glaucoma (2.0%, N=68 non-surgical and 3.0% N=101 surgical patients; OR, 1.48; 95% CI 1.08-2.01).
Figure 3.
Counts of number of dispensed bottles of postoperative ophthalmic medication classes over time in 16,169 matched pairs with glaucoma. Each patient contributed multiple postoperative measures based on the timing of their dispensings.
*Non-surgical patients were matched 1:1 to surgical patients on age, gender, type of glaucoma, before index IOP, and number and type of glaucoma medications using the closest possible match. The number of preoperative dispensings was counted using 12 months of enrollment, while postoperative medications were counted using 6 months periods, except that the number at 6 months represents dispensings from 1 month to 6 months.
Table 3.
Association of phacoemulsification for cataract with selective laser trabeculoplasty and glaucoma surgery in the subsequent 36 months, surgical and non-surgical matched pairs with glaucoma. Kaiser Permanente Northern California glaucoma patients, 2010-15*
| Selective Laser Trabeculoplasty | Glaucoma Surgery | ||||||
|---|---|---|---|---|---|---|---|
| Subgroup | N pairs | Non-surgical, % | Surgical, % | OR (95% CI) | Non-surgical, % | Surgical, % | OR (95% CI) |
| All IOP measures | |||||||
| Overall | 16,169 | 3.10 | 2.38 | 0.76 (0.67-0.87) | 0.56 | 0.84 | 1.50 (1.15-1.96) |
| Ocular hypertension | 1,020 | 1.76 | 0.49 | 0.27 (0.10-0.74) | 0.00 | 0.10 | - |
| Glaucoma suspect | 7,373 | 1.18 | 0.37 | 0.31 (0.20-0.47) | 0.01 | 0.05 | 4.00 (0.44-35.8) |
| Primary open-angle glaucoma | 3,391 | 7.25 | 6.90 | 0.95 (0.79-1.14) | 2.03 | 2.98 | 1.48 (1.08-2.01) |
| Normal tension glaucoma | 879 | 7.39 | 5.57 | 0.74 (0.50-1.08) | 0.91 | 1.14 | 1.25 (0.49-3.19) |
| Anatomic narrow angle | 1,179 | 0.51 | 0.42 | 0.83 (0.25-2.74) | 0.17 | 0.25 | 1.50 (0.25-9.00) |
| Primary angle closure glaucoma | 503 | 1.39 | 1.39 | 1.00 (0.35-2.87) | 0.60 | 1.19 | 2.01 (0.50-8.09) |
| Pseudoexfoliation glaucoma | 181 | 7.73 | 3.31 | 0.41 (0.15-1.09) | 1.10 | 1.10 | 1.00 (0.14-7.18) |
| Unspecified glaucoma | 1,643 | 3.53 | 3.16 | 0.89 (0.61-1.31) | 0.37 | 0.55 | 1.50 (0.53-4.23) |
| Goldmann measures only | |||||||
| Overall | 8,977 | 4.29 | 3.38 | 0.78 (0.67-0.91) | 0.68 | 1.20 | 1.78 (1.30-2.44) |
| Ocular hypertension | 561 | 1.78 | 0.53 | 0.30 (0.08-1.08) | 0.00 | 0.18 | - |
| Glaucoma suspect | 3404 | 1.73 | 0.53 | 0.30 (0.18-0.51) | 0.06 | 0.06 | 1.00 (0.14-7.10) |
| Primary open-angle glaucoma | 2406 | 8.19 | 7.81 | 0.95 (0.77-1.17) | 1.83 | 3.37 | 1.87 (1.29-2.71) |
| Normal tension glaucoma | 616 | 8.44 | 6.66 | 0.77 (0.51-1.18) | 0.65 | 1.30 | 2.01 (0.60-6.72) |
| Anatomic narrow angle | 573 | 0.35 | 0.35 | 1.00 (0.14-7.12) | 0.00 | 0.52 | - |
| Primary angle closure glaucoma | 300 | 2.33 | 2.00 | 0.85 (0.28-2.57) | 1.67 | 2.00 | 1.20 (0.36-3.99) |
| Pseudoexfoliation glaucoma | 98 | 14.3 | 2.04 | 0.13 (0.03-0.57) | 1.02 | 2.04 | 2.02 (0.18-22.7) |
| Unspecified glaucoma | 1019 | 4.32 | 4.22 | 0.98 (0.64-1.50) | 0.49 | 0.49 | 1.00 (0.29-3.46) |
Non-surgical patients were matched 1:1 to surgical patients on age, gender, type of glaucoma, preoperative IOP, and number and type of glaucoma medications using the closest possible match.
DISCUSSION
We conducted a community-based retrospective cohort study to evaluate change in IOP after simple phacoemulsification in patients with glaucoma, ocular hypertension, anatomic narrow angles, and in glaucoma suspects. Non-surgical patients had an increase in IOP over the 3-year follow-up period, regardless of glaucoma subtype. In contrast, surgical patients experienced a decrease in IOP, regardless of glaucoma subtype. The decrease was greatest for patients with ocular hypertension, anatomic narrow angle, and primary angle closure glaucoma, and for those whose baseline IOP was most elevated. Strengths of the study included the large sample size, community-based setting, detailed information, use of controls with glaucoma who did not undergo phacoemulsification, careful matching of baseline characteristics, and use of contemporary statistical approaches.
In surgical patients, the average decrease in IOP ranged from 0.7 to 1.7 mm Hg across glaucoma types. Compared with the controls, whose IOP increased, this translated to a difference-in-difference of 0.7 to 2.0 mm Hg. The finding of a reduction is consistent with past reports. However, the magnitude of these reductions is not as large as reported previously. For ocular hypertension, Mansberger and colleagues5 studied 42 phacoemulsification patients whose average IOP declined from 23.9 to 19.8 mm Hg at 1 year, in comparison with the present study in which IOP declined from 18.7 to 17.7 mm Hg at an average of 7 months. It is important to note that in this community study the average IOP for the ocular hypertension group was below 21mmHg. For glaucoma suspects, Yoo and colleagues (8) studied 51 patients whose IOP declined from 16.0 to 12.7 mm Hg at 2 months, compared with the present patients, whose IOP declined from 16.6 to 15.3 mm Hg.
Primary open angle glaucoma has been examined in numerous studies and is summarized in a meta-analysis by Chen and colleagues,4 analyzing 461 patients in 9 studies. They report IOP declined from 17.7 to 15.4 mm Hg, compared with the present reduction from 16.4 to 15.5 mm Hg. Shoji and colleagues10 studied normal tension glaucoma in 35 patients in whom IOP declined from 16.7 to 14.7 mm Hg at 6 months, compared with the present reduction of 13.9 to 13.2 mm Hg in 879 patients.
As part of the EAGLE study, Azuara-Blanco and colleagues3 studied 208 patients, with primary angle closure or primary angle closure glaucoma, randomized to phacoemulsification vs. conventional treatment with laser peripheral iridotomy and topical medications. They noted that average IOP in the phacoemulsification group declined from 29.5 to 15.9 mm Hg and in the conventional treatment group from 30.3 to 18.4 mm Hg at 12 months. In the meta-analysis by Chen and colleagues4, higher IOP prior to phacoemulsification was found to be the single most common significant factor associated with a greater reduction in IOP after phacoemulsification. In the primary angle closure group, IOP declined from 20.2 to 14.2 mm Hg. Chen’s results can be compared with the present study’s reduction from 16.9 to 15.5 mm Hg for the primary angle closure group. Finally, pseudoexfoliation glaucoma was examined in the same meta-analysis by Chen and colleagues4, who found a reduction of IOP from 20.7 to 16.6 mm Hg in 132 patients from 6 studies, compared with the present study’s reduction from 16.8 to 15.4 mm Hg.
Multiple studies, including this one, found that the patients with lower pre-operative IOP had lower reductions after phacoemulsification.4 The IOP reductions we observed, however, were less than those reported in past studies. One striking difference between past studies and this one is the lower preoperative IOP in nearly every one of our glaucoma subpopulations, except for glaucoma suspects. In the OHTS study,5 for example, inclusion criteria included an IOP between 24 and 32mmHg in one eye and between 21 and 32mmHg in the fellow eye. In the EAGLE study3 the inclusion criteria were primary angle closure with IOP greater than 30 (38% of patients) or primary angle closure glaucoma with an IOP greater than 21mmHg on one occasion (61%). Consequently, pre-operative IOP was notably high, at 29.5 mm Hg, and was reduced substantially in both surgical cohorts. In addition, the ophthalmologists were given postoperative IOP targets of 15 to 20 mm Hg, with their care being observed by the study team, so that treatment may have been more closely monitored.
IOP reduction after cataract surgery in this study reached a nadir at about one year and increased slowly over the subsequent two years without returning to baseline by 36 months. This trajectory is consistent with past reports.5 Our finding that 3% of patients with primary open angle glaucoma underwent of subsequent glaucoma surgery compared to only 2% of nonsurgical patients may seem paradoxical. The differential rate of glaucoma surgery in our study may be due to a mismatch in glaucoma severity between surgical and nonsurgical groups, however we could not directly assess this. For comparison, Bojikian et al found that 6.7% of uncontrolled or marginally controlled open angle glaucoma patients required trabeculectomy after phacoemulsification.15 Kim et al16 reported that the sustained reduction of IOP associated with phacoemulsification did not slow the rate of glaucomatous visual field decay in patients, but there was no control group for comparison.
With respect to change in the number of dispensed glaucoma medications, our results showing a decrease after cataract surgery are consistent with the Chen meta-analysis4, the EAGLE study3, and an analysis of 8,604 glaucoma patients who underwent surgery in Taiwan during 1998-2011, although that study could not match the control group on baseline IOP.7 We also observed that ocular hypertensives and glaucoma suspects had a 3- to 4-fold decreased risk of selective laser trabeculoplasty within 36 months which parallels reduced medication prescribing.
In community-based settings, a variety of methods are used to measure intraocular pressure. The inclusion of IOP measurement techniques other than Goldmann applanation tonometry may be considered a weakness of this study, but it was worthwhile to assure representativeness. To address this potential limitation, we conducted subgroup analysis matching surgical patients with Goldmann applanation tonometry to non-surgical patients with Goldmann applanation tonometry and found results that were very similar to the main analysis. An additional weakness of this study was use of a single baseline IOP measurement, however, restricting to patients with multiple measurements could have reduced representativeness, and any imprecision similarly affected both surgery and non-surgery patients. In addition, we retained patients who had a history of vitrectomy or complications during surgery, but this comprised fewer than 2% of patients,13 and the effect on the results was negligible. A key weaknesses of the study was lack of information on the severity of glaucoma, although matching on baseline IOP and number of medications likely minimized any differences between the surgical and non-surgical groups.
In summary, phacoemulsification was associated with a reduction in IOP, a result that has been widely reported, although the baseline IOP in our patients was lower and the magnitude of the reduction was lower, averaging 0.7 to 1.9 mm Hg depending on the glaucoma subtype. The reduction was quite marked in patients with the highest IOP, averaging 2.0-2.5 mm Hg in those with IOP ≥20 mm Hg. In addition, the number of ophthalmic medication classes dispensed to surgical patients after their index date was lower than the number dispensed to nonsurgical patients after their index date. Phacoemulsification does not reduce the frequency of subsequent glaucoma surgery, however, and patients with primary open angle glaucoma were at slightly increased risk for subsequent surgical intervention. Further study will be needed to confirm the finding that the reduction in IOP observed in surgical patients showed a monotonic relationship with the number of ophthalmic glaucoma medication classes dispensed in the baseline period.
Implications:
Patients generally have a lower IOP after phacoemulsification. For patients with preoperative IOP below 20mmHg, the magnitude of IOP reduction with phacoemulsification is small and is dependent upon preoperative IOP. If larger magnitude reductions in IOP are desired desired at the time of cataract surgery, surgeons may consider combining the phacoemulsification with either minimally invasive surgery17 or with a conventional glaucoma procedure, as has been recommended by others.18
Supplementary Material
Acknowledgments
Financial Support: This project was funded by the National Eye Institute R01 EY027329. The project also used products developed under earlier research grants provided by NEI R21 EY022989, Kaiser Permanente’s Community Benefit program, and the Garfield Memorial Fund, Kaiser Permanente. These sponsors had no role in the design or conduct of this research.
Abbreviations:
- CI
confidence interval
- IOP
intraocular pressure
- OR
odds ratio
- SD
standard deviation
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Meeting Presentation: This material has not been presented previously.
Conflict of Interest: No conflicting relationship exists for any author.
Supplemental Material: This article contains additional online-only material.
REFERENCES
- 1.Weinreb RN, Aung T, Medeiros FA. The pathophysiology and treatment of glaucoma: a review. JAMA. 2014;311:1901–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Armstrong JJ, Wasiuta T, Kiatos E, et al. The Effects of Phacoemulsification on Intraocular Pressure and Topical Medication Use in Patients With Glaucoma: A Systematic Review and Meta-analysis of 3-Year Data. J Glaucoma. 2017;26:511–522. [DOI] [PubMed] [Google Scholar]
- 3.Azuara-Blanco A, Burr J, Ramsay C, et al. Effectiveness of early lens extraction for the treatment of primary angle closure glaucoma (EAGLE): a randomised controlled trial. Lancet. 2016;388:1389–1397. [DOI] [PubMed] [Google Scholar]
- 4.Chen PP, Lin SC, Junk AK, et al. The Effect of Phacoemulsification on Intraocular Pressure in Glaucoma Patients: A Report by the American Academy of Ophthalmology. Ophthalmology. 2015;122:1294–307. [DOI] [PubMed] [Google Scholar]
- 5.Mansberger SL, Gordon MO, Jampel H, et al. Reduction in intraocular pressure after cataract extraction: the Ocular Hypertension Treatment Study. Ophthalmology. 2012;119:1826–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Baek SU, Kwon S, Park IW, Suh W. Effect of Phacoemulsification on Intraocular Pressure in Healthy Subjects and Glaucoma Patients. J Korean Med Sci. 2019;34:e47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chen HY, Lin CL, Kao CH. Changes in glaucoma medication numbers after cataract and glaucoma surgery: A nationwide population: 10.1097/MD.0-based study. Medicine (Baltimore). 2019;98:e14128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yoo C, Amoozgar B, Yang KS, et al. Glaucoma severity and intraocular pressure reduction after cataract surgery in eyes with medically controlled glaucoma. Medicine (Baltimore). 2018;97:e12881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jimenez-Roman J, Lazcano-Gomez G, Martínez-Baez K, et al. Effect of phacoemulsification on intraocular pressure in patients with primary open angle glaucoma and pseudoexfoliation glaucoma. Int J Ophthalmol. 2017;10:1374–1378. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Shoji T, Tanito M, Takahashi H, et al. Phacoviscocanalostomy versus cataract surgery only in patients with coexisting normal-tension glaucoma: midterm outcomes. J Cataract Refract Surg. 2007;33:1209–16. [DOI] [PubMed] [Google Scholar]
- 11.Vinod K, Gedde SJ, Feuer WJ, et al. Practice Preferences for Glaucoma Surgery: A Survey of the American Glaucoma Society. J Glaucoma. 2017;26:687–693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Liu L, Herrinton LJ, Alexeeff S, et al. Visual outcomes after cataract surgery in patients with type 2 diabetes. J Cataract Refract Surg. 2019;45:404–413. doi: 10.1016/j.jcrs.2018.11.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Liu L, Shorstein NH, Amsden LB, Herrinton LJ. Natural language processing to ascertain two key variables from operative reports in ophthalmology. Pharmacoepidemiol Drug Saf. 2017. ;26:378–385. doi: 10.1002/pds.4149. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Deyo RA, Cherkin DC, Ciol MA, Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613–19. [DOI] [PubMed] [Google Scholar]
- 15.Bojikian KD, Chen PP. Intraocular Pressure After Phacoemulsification in Open-angle Glaucoma Patients With Uncontrolled or Marginally Controlled Glaucoma and/or With Severe Visual Field Loss. J Glaucoma. 2018;27(2):108–114. [DOI] [PubMed] [Google Scholar]
- 16.Kim JI, Rabiolo A, Morales E, et al. Cataract Surgery and Rate of Visual Field Progression in Primary Open Angle Glaucoma. Am J Ophthalmol 2019;201:19–30. [DOI] [PubMed] [Google Scholar]
- 17.Samuelson TW, Chang DF, Marquis R, et al. A Schlemm Canal Microstent for Intraocular Pressure Reduction in Primary Open-Angle Glaucoma and Cataract. Ophthalmology. 2019;126 (1):29–37. [DOI] [PubMed] [Google Scholar]
- 18.Freidman DS et al. Surgical strategies for coexisting glaucoma and cataract: an evidence-based update. Ophthalmology 2002, 109: 1902–13. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.