Purpose
To evaluate use of medical, laser, or incisional surgical interventions for glaucoma after laser peripheral iridotomy (LPI).
Design
Retrospective longitudinal cohort study.
Methods
All enrollees aged ≥21 years in a US managed-care network who underwent bilateral LPIs in 2001–2011 were identified. The mean numbers of pre- and post-LPI glaucoma medication classes prescribed and the proportion of enrollees requiring cataract or glaucoma surgery within 2 years after the LPIs were determined. Multivariable logistic regression assessed factors associated with enrollees’ prescription of ≥1 glaucoma medication class after bilateral LPIs.
Results
Of the 1660 patients undergoing bilateral LPIs, 1280 (77.1%) had no pre- or post-LPI prescriptions for any glaucoma medication class. Of the remaining patients, 251 (66.1%) required more glaucoma medication classes after than before the procedures, whereas 44 (11.6%) used fewer after the procedures; 85 (22.4%) were prescribed the same number before and after the LPIs. A total of 167 patients (10.1%) underwent cataract surgery and 79 (4.8%) received glaucoma surgery over the 2-year follow-up. Black patients had a 130% increased odds for glaucoma medication–class prescriptions after bilateral LPIs, compared with white patients ( P = .02). The odds of post-LPI glaucoma medication use increased by 21% for every additional 5 years of age ( P < .0001).
Conclusion
Most patients undergoing bilateral LPIs received no pre- or post-LPI glaucoma medication–class prescriptions and had no cataract or additional glaucoma surgery within 2 years after LPIs. Clinicians should alert black or older patients and those already taking glaucoma medications before the procedure of their higher odds of requiring medications afterward.
Anatomically narrow- or closed-angle entities are defined by a spectrum of findings including iridotrabecular meshwork contact, trabecular meshwork dysfunction, and occasionally elevated intraocular pressure (IOP). Most eyes with angle closure possess a component of relative pupillary block with increased iridolenticular resistance to aqueous movement from its site of production in the posterior chamber to its site of egress at the anterior chamber angle. A laser peripheral iridotomy (LPI) often successfully eliminates the relative pupillary block component of the angle-closure process. Creation of the LPI allows aqueous to flow more freely from the posterior chamber to the anterior chamber with a change in the iris configuration such that it often opens up the angle. LPI is also recommended as prophylactic treatment for primary angle-closure suspects (PACS) or patients with occludable angles.
Most studies of patients undergoing LPIs have focused on the impact of the procedure on IOP and anatomic outcomes (ie, ultrasound biomicroscopy, anterior chamber optical coherence tomography, gonioscopic findings), such as changes in anterior chamber depth, changes in angle appearance, and progression to angle closure. A few small studies have described the incidence of progression of persons considered PACS to primary angle closure (PAC) or primary angle-closure glaucoma (PACG) after LPI; recurrence of angle closure due to plateau iris syndrome, lens relocation, or ciliary block ; or complications of LPI such as corneal damage. However, little is known about the need for additional interventions for glaucoma following LPI, such as IOP-lowering medications, cataract extraction, or laser or incisional glaucoma surgery; or if there are factors that increase or decrease the probability of requiring additional interventions following LPIs. Such information is important for clinicians to adequately advise patients what to expect after the procedure. Using data from a large managed-care network, we identified 1660 patients who underwent bilateral LPIs and followed them continuously for 2 years to determine the need for medical, laser, or incisional glaucoma surgery following the iridotomies.
Methods
Data Source
The Clinformatics Data Mart database (OptumInsight, Eden Prairie, MN) contains detailed claims data of all beneficiaries in a nationwide US managed-care network. The dataset contains healthcare claims data for all individuals with ≥1 International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes for eye-related diagnoses (360–379.9); ≥1 Current Procedural Terminology codes for any eye-related visits, diagnostics, or therapeutic procedures (65091–68899 or 92002–92499); or any other claim submitted by an ophthalmologist or optometrist from 2001 to 2011. For each enrollee, we had access to all medical claims for ocular and nonocular conditions and sociodemographic information, including age, sex, race, education level, and income. The database also captures information on all outpatient medications filled. All persons in the medical plan were also fully enrolled in the pharmacy plan. The database has been used in the past to study patients with glaucoma. Since all the data for these analyses were de-identified to the investigators, the University of Michigan Institutional Review Board approved this retrospective longitudinal cohort study as a nonregulated study.
Participants and Sample Selection
Individuals were included in the analysis if they met the following criteria: age ≥21 years, continuous enrollment in the medical plan, and the receipt of bilateral LPIs (CPT code 66761) within a 3-month period. Eligible enrollees were also required to have ≥2 years in the plan prior to the first LPI and ≥2 years in the plan after the second LPI ( Supplemental Figure , available at AJO.com ).
Individuals were excluded if they had any record of cataract surgery or laser or incisional glaucoma surgery during their time in the plan prior to the LPIs. Also excluded were those with noncontinuous enrollment and those who did not undergo bilateral LPIs within a 3-month time frame. Because this data source does not reliably capture the laterality of the eye undergoing the LPI, we limited our study to those undergoing this procedure on 2 separate occasions within a 3-month period so that if either or both eyes required subsequent medical or surgical interventions for glaucoma, we would know with greater certainty that the interventions were performed on at least 1 eye that had undergone LPI surgery. Since this study’s focus was on the effectiveness of LPI, we did not consider enrollees who underwent surgical iridectomy.
Reason for the Laser Peripheral Iridotomy
To attempt to understand the reason why each enrollee underwent LPI, we queried the database to determine the primary ICD-9-CM diagnosis code listed on the encounter when the initial LPI was performed for each patient.
Outcomes
In the 2 years following the second LPI, we assessed the proportion of enrollees who were prescribed glaucoma medications or underwent cataract or glaucoma surgery. Eight medication classes were identified: topical beta-blockers, alpha-agonists, prostaglandin analogues, topical carbonic anhydrase inhibitors, oral carbonic anhydrase inhibitors, miotics, adrenergics, and combination medications. We quantified the number of medication classes ever prescribed for a given patient during the time period of interest. For example, during the 2-year period prior to first LPI, if a patient was taking a prostaglandin analogue for 6 months followed by a topical beta-blocker for 3 months, we considered him to have taken 2 medication classes during that period. Likewise, if she took both medications simultaneously during the 2-year interval, she would also be considered to have taken 2 classes. Patients taking a fixed combination were treated as taking each individual component. Similar quantification of medication use was done for the 2-year interval after the LPIs. The following surgeries were also considered: cataract surgery, laser glaucoma surgery (iridoplasty, laser trabeculoplasty), or incisional glaucoma surgery (trabeculectomy, glaucoma drainage device) ( Supplemental Table 1 , available at AJO.com ).
Analyses
Statistical analyses were performed using SAS software, version 9.3 (SAS Institute, Cary, North Carolina, USA). Participant characteristics were summarized for the entire sample using mean values and standard deviations for continuous variables and frequencies and percentages for categorical variables. We determined the proportion of enrollees who were prescribed IOP-lowering medications and the number of agents prescribed in the 6, 12, and 24 months prior to bilateral LPI and during the 6, 12, and 24 months after the LPIs. Since glaucoma medication use can fluctuate owing to addition, substitution, or discontinuation of medications, to help simplify the quantification of medication use prior to and after the LPIs, we looked at the number of glaucoma medication classes each enrollee was prescribed on the specific date corresponding to 6, 12, and 24 months prior to the initial LPI and 6, 12, and 24 months afterwards and determined means for each time point. The proportion of enrollees who required cataract extraction or laser or incisional glaucoma surgery following bilateral LPI was also captured, along with the timing of these procedures relative to the LPIs. In order to explore the rate of complications such as aqueous misdirection or corneal decompensation after LPI, we also assessed the proportion of patients who underwent vitrectomy or keratoplasty after the bilateral LPIs. A sub-analysis was performed to determine whether receipt of cataract and glaucoma surgery following LPI varied by race. If a patient underwent more than 1 of these surgeries, only the first was considered. Only 15 patients (0.9%) underwent more than 1 of these surgeries within 2 years following bilateral LPIs. If a patient underwent a combined cataract and glaucoma surgery on the same day, both were captured.
Multivariable logistic regression was performed to assess the impact of sociodemographic characteristics of the patients and other factors on the odds of receiving IOP-lowering medication after bilateral LPI. Covariates included in the model were maximum number of glaucoma medication classes prescribed in the 2 years prior to the initial LPI, primary diagnosis on the date of the initial LPI, age, sex, race, education, income, and Charlson Comorbidity Index, a measure of overall health. Model covariates were determined to be noncollinear by checking correlation coefficients.
Results
A total of 1660 beneficiaries met the inclusion criteria ( Figure ). The mean ± standard deviation age of the beneficiaries at enrollment was 56.9 ± 11.2 years. There were 1142 female subjects (68.8%), and the racial distribution (of 1511 subjects for whom data were available) included 1244 whites (82.3%), 50 blacks (3.3%), 122 Latinos (8.1%), 75 Asian Americans (5.0%), and 20 individuals of other races (1.3%). Of 1619 subjects with available education data, there were 1065 patients (65.8%) who had higher than high school–level education; 390 of 1567 patients (24.9%) had incomes >$100 000 ( Table 1 ). A total of 1373 persons (82.7%) underwent gonioscopy (CPT code 92020) at least once while in the plan. The primary diagnosis listed on the date of the initial LPI was most commonly anatomical narrow angle (n = 955, 47.9%), followed by PACG (n = 300, 15.1%) and chronic angle-closure glaucoma (CACG) (n = 220, 11.0%). Some patients saw more than 1 eye care provider on the date of the initial LPI and may have received different diagnoses by the different providers, so the total number of diagnoses adds up to more than 1660 ( Supplemental Table 2 , available at AJO.com ).
| Variable | Value | Total, N (%) |
|---|---|---|
| Total | 1660 (100.0) | |
| Sex (N miss = 0) | Male | 518 (31.2) |
| Female | 1142 (68.8) | |
| Race (N miss = 149) | White | 1244 (82.3) |
| Black | 50 (3.3) | |
| Latino | 122 (8.1) | |
| Asian | 75 (5.0) | |
| Other | 20 (1.3) | |
| Education (N miss = 41) | Less than high school | 25 (1.5) |
| High school diploma | 529 (32.7) | |
| Some college | 657 (40.6) | |
| College diploma | 403 (24.9) | |
| Advanced degree | 5 (0.3) | |
| Income (N miss = 93) | <$30 000 | 151 (9.6) |
| $30 000-$59 999 | 505 (32.2) | |
| $60 000-$99 999 | 521 (33.2) | |
| $100 000-$124 999 | 195 (12.4) | |
| ≥$125 000 | 195 (12.4) | |
| Urban/rural status (N miss = 4) | Urban | 1530 (92.4) |
| Larger rural | 66 (4.0) | |
| Small rural | 60 (3.6) | |
| Mean (SD) age (in years) at enrollment (N miss = 0) | 56.9 (11.2) |
Classes of Glaucoma Medications Prescribed Before and After Bilateral Laser Peripheral Iridotomy
Of the 1660 enrollees who underwent bilateral LPIs, 1280 (77.1%) had no record of any classes of glaucoma medications prescribed before or after the LPIs and 380 (22.9%) required glaucoma medications either before or after the procedure. Thirty (1.8%) were prescribed ≥1 medication class before the first LPI but were not prescribed glaucoma medications after the second LPI; 205 (12.3%) did not use any classes of glaucoma medications prior to the first LPI but were prescribed ≥1 glaucoma medication class after the second LPI; and 145 (8.7%) were prescribed medications both before the first and after the second LPI. Among the 380 patients taking ≥1 class of glaucoma medications either before or after the LPIs, 251 (66.1%) required more glaucoma medication classes after the LPIs, 85 (22.4%) required the same number of medication classes before and after, and 44 (11.6%) required fewer glaucoma medication classes following the LPIs. Among 175 patients on IOP-lowering medication prior to LPI, 30 patients (17.1%) were not prescribed any classes of glaucoma medication post-LPI and 145 patients (82.9%) were prescribed ≥1 glaucoma medication class post-LPI. Table 2 shows the proportion of patients taking medications at 6,12, and 24 months before and after the LPIs and among those taking medications, and the mean and standard deviation number of classes taken at each time point.
| Patients Taking Medication, N (%) a | Mean (SD) Number of Medication Classes for Those Receiving Glaucoma Medications b | ||||||
|---|---|---|---|---|---|---|---|
| Before First LPI | After Second LPI | P Value c | N | Before First LPI | After Second LPI | P Value d | |
| Total | 175 (10.5) | 350 (21.1) | <.0001 | 145 | 1.48 (0.76) | 1.80 (0.97) | <.0001 |
| 6 months | 57 (3.4) | 127 (7.7) | <.0001 | 24 | 1.21 (0.41) | 1.33 (0.56) | .42 |
| 12 months | 63 (3.8) | 140 (8.4) | <.0001 | 26 | 1.23 (0.43) | 1.42 (0.58) | .17 |
| 24 months | 53 (3.2) | 165 (9.9) | <.0001 | 36 | 1.11 (0.32) | 1.25 (0.50) | .10 |
a The interpretation is that 175 patients were taking glaucoma medications at any point during the 2 years prior to the first LPI and 350 patients were taking glaucoma medications at any point in the 2 years after the second LPI. Likewise, there were 57 patients who were taking medications at 6 months prior to the first LPI and 127 taking medications at 6 months after the second LPI.
b Data indicating that among the 145 patients who were taking glaucoma medications at any point during the 2 years prior to the first LPI and at any point after the second LPI, the mean and standard deviation number of medications they were taking was 1.48 (0.76) before and 1.80 (0.97) after. Likewise, among the 24 patients who were taking glaucoma medications at 6 months prior to the first LPI and at 6 months after the second LPI, the mean number of medications taken before the first LPI was 1.21 (0.41) and afterward 1.33 (0.56).
Factors Associated With Receipt of ≥1 Class of Glaucoma Medication After Bilateral Laser Peripheral Iridotomy
Blacks had 130% increased odds of being prescribed ≥1 classes of IOP-lowering medication after bilateral LPIs compared with whites (adjusted odds ratio [OR] = 2.30, confidence interval [CI] = 1.13–4.66). Asians and Latinos did not differ from whites in their receipt of IOP-lowering medication after the LPIs ( P = .63 and P = .81, respectively). Older age (adjusted OR = 1.21, CI = 1.11–1.32) was associated with increased odds of requiring ≥1 class of IOP-lowering medication after bilateral LPIs. Each additional 5 years of age conferred a 21% increased odds of being prescribed medications. Each additional glaucoma medication class prescribed before the initial LPI increased the odds of receiving glaucoma medications after the LPI 17-fold (adjusted OR = 16.73, CI = 10.59–26.44). Compared to persons carrying a diagnosis of narrow angles, those diagnosed with intermittent angle-closure glaucoma had a 124% increased odds of requiring ≥1 class of medication following LPI (adjusted OR = 2.24, CI = 1.18–4.27) and those with chronic angle-closure glaucoma had an 80% increased odds of requiring ≥1 class of medication following LPI (adjusted OR = 1.80, CI = 1.13–2.84). There was no association between sex, education, or income and odds of being prescribed IOP-lowering medications after bilateral LPIs ( Table 3 ).
| Covariate | Level | OR (95% CI) | P Value |
|---|---|---|---|
| Age at enrollment | 5-year increments | 1.21 (1.11, 1.32) a | <.0001 |
| Sex (ref = male) | Female | 0.98 (0.70, 1.37) | .91 |
| Race (ref = white) | Black | 2.30 (1.13, 4.66) a | .02 |
| Latino | 0.93 (0.51, 1.69) | .81 | |
| Asian | 1.20 (0.57, 2.51) | .63 | |
| Other | 0.62 (0.14, 2.84) | .54 | |
| Personal income (ref = <$30K) | $30K-$60K | 1.23 (0.69, 2.18) | .49 |
| $60K-$100K | 1.12 (0.60, 2.10) | .72 | |
| $100K-$125K | 0.95 (0.44, 2.05) | .90 | |
| >$125K | 0.52 (0.23, 1.120) | .12 | |
| Education (ref = less than high school) | High school diploma | 0.54 (0.18, 1.59) | .26 |
| Some college | 0.65 (0.22, 1.93) | .44 | |
| College diploma or higher | 0.71 (0.22, 2.24) | .55 | |
| Charlson Comorbidity Index | 0.95 (0.90, 1.01) | .08 | |
| Number of medication classes before LPI | 16.73 (10.59, 26.44) a | <.0001 | |
| Primary diagnosis (ref = anatomical narrow angle) | Primary angle-closure glaucoma, unspecified | 1.23 (0.73, 2.07) | .44 |
| Chronic angle-closure glaucoma | 1.80 (1.13, 2.84) a | .01 | |
| Acute angle-closure glaucoma | 0.62 (0.26, 1.50) | .29 | |
| Intermittent angle-closure glaucoma | 2.24 (1.18, 4.27) a | .01 | |
| Multiple diagnoses | 1.33 (0.78, 2.25) | .30 | |
| Other diagnoses | 2.18 (1.29, 3.68) a | .004 |
Cataract Surgery After Bilateral Laser Peripheral Iridotomy
A total of 167 patients (167/1660; 10.1%) underwent ≥1 cataract surgery after the second LPI. When distributed by age (<65 or ≥65), 70 of 1253 patients (5.6%) younger than age 65 underwent subsequent cataract surgery while 97 of 407 patients (23.8%) age ≥65 underwent cataract surgery after the LPIs. As expected, the mean ± standard deviation age at enrollment for enrollees who did not undergo cataract surgery was younger than those who underwent surgery (55.8 ± 10.9 years vs 66.1 ± 10.0 years; P < .0001) ( Table 4 ). Of those who underwent cataract surgery after bilateral LPIs (n = 167), 103 patients (61.7%) had the procedure within 1 year after LPIs and 64 patients (38.3%) had the procedure during the second year following the LPIs. Supplemental Table 3 (available at AJO.com ) shows the rates of cataract surgery stratified by age, sex, and race. While the numbers are small, these rates are 2–3 times higher than others reported in the literature for persons of comparable demographic profiles.
| Cataract Surgery After Second LPI | ||||
|---|---|---|---|---|
| No | Yes | Total | P Value | |
| Age at enrollment, n (row %) | ||||
| Under 65 | 1,183 (79.2) | 70 (41.9) | 1,253 (75.5) | <.0001 a |
| 65 and over | 310 (20.8) | 97 (58.1) | 407 (24.5) | |
| Total | 1,493 (89.9) | 167 (10.1) | 1,660 (100.0) | |
| Age at enrollment, mean ± SD (range) | 55.8 ± 10.9 (23, 89) | 66.1 ± 10.0 (29, 84) | 56.9 ± 11.2 (23, 89) | <.0001 b |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
