Kaplan-Meier cumulative probability of undergoing filtration surgery in at least 1 eye.
Kaplan-Meier cumulative probability of undergoing filtration surgery in at least 1 eye for the treated ocular hypertension vs classic glaucoma groups.
Kaplan-Meier cumulative probability of undergoing filtration surgery in both eyes.
Kaplan-Meier cumulative probability of undergoing filtration surgery in both eyes for the treated ocular hypertension vs classic glaucoma groups.
Matthew G. Hattenhauer, Douglas H. Johnson, Helen H. Ing, David O. Hodge, Linda C. Butterfield, David C. Herman, Darryl T. Gray. Probability of Filtration Surgery in Patients With Open-angle Glaucoma. Arch Ophthalmol. 1999;117(9):1211–1215. doi:10.1001/archopht.117.9.1211
To investigate the probability of undergoing filtration surgery in either 1 or both eyes in patients in whom open-angle glaucoma was newly diagnosed.
Methods and Design
A retrospective community-based study of 295 residents of Olmsted County, Minnesota, in whom open-angle glaucoma was newly diagnosed between January 1, 1965, and December 31, 1980, was performed. Kaplan-Meier methods were used to estimate the cumulative probability of undergoing filtration surgery during a 20-year period.
At 20 years of follow-up, the Kaplan-Meier cumulative probability of undergoing filtration surgery in at least 1 eye was estimated to be 23% (95% confidence interval, 16%-30%), and in both eyes the estimate was 12% (95% confidence interval, 6%-17%). Patients with optic nerve damage at the time of diagnosis were more likely to undergo surgery than patients with elevated intraocular pressure but no damage (1 eye, 39% vs 15%; both eyes, 27% vs 5%).
This retrospective study of a white population newly diagnosed as having and treated for open-angle glaucoma indicates that while most patients did not undergo filtration surgery in the course of glaucoma therapy, at least one third of those with glaucomatous damage at the time of diagnosis underwent filtration surgery.
AS A WORLDWIDE cause of blindness, glaucoma ranks among the top permanent visually debilitating diseases.1 While there is no cure for open-angle glaucoma (OAG), there are various treatment options that lower intraocular pressure (IOP) and may, therefore, slow the progressive loss of visual field associated with the disease. Topical and systemic medications, laser trabeculoplasty, and filtration surgery are common methods used in an attempt to lower IOP. The role and timing of these therapeutic options often vary based on the clinician's training and experience, as well as the severity of the disease. Traditionally, ophthalmologists in the United States have followed a stepwise approach in the management of OAG, with extensive use of topical ocular hypotensive medications, followed by laser trabeculoplasty, and then filtration surgery. European clinicians have been more aggressive in their treatment of OAG, with filtration surgery recommended either as initial therapy or after a trial of a topical medication fails to control IOP or the progression of visual field loss.2,3 The reevaluation of treatment strategy in the United States has led to several large multicentered studies, including the Collaborative Initial Glaucoma Treatment Study4 and the Glaucoma Laser Trial,5 that compare surgical intervention with medical therapy in the treatment of OAG.
Consideration of a more aggressive approach to reduction of IOP with early surgery has evolved, in part, because of the fact that blindness from OAG can occur even while patients are receiving therapy.6 In a previous study,6 we reported the Kaplan-Meier 20-year cumulative probability of blindness in patients in whom OAG was newly diagnosed January 1, 1965, through December 31, 1980. This probability of blindness was estimated to be 27% (95% confidence interval, 20%-33%) in at least 1 eye and 9% (95% confidence interval, 5%-14%) in both eyes.6 At the time of initial diagnosis, those eyes with evidence of optic disc and/or visual field loss from glaucoma had a greater probability of blindness than did those eyes with only elevated IOP.6 This article provides information from the same study population regarding the probability that a patient in whom OAG was newly diagnosed will undergo filtration surgery during a 20-year period.
The present study is a retrospective, community-based, longitudinal study of residents of Olmsted County, Minnesota, in whom OAG was newly diagnosed from January 1, 1965, through December 31, 1980. This period was chosen to allow for adequate follow-up after diagnosis.
Study patients were identified through a computerized search of the database of the Rochester Epidemiology Project, a medical records linkage system established to study the occurrence and natural history of disease among the residents of Rochester, Minn, and the balance of Olmsted County.7,8 Using this system, population-based studies can be conducted because almost all medical care is provided to area residents by Rochester institutions affiliated with the Mayo Clinic or Olmsted Medical Center. A medical record system is used by the Mayo Clinic and its affiliated hospitals whereby all data on inpatient and outpatient care is assembled in 1 place. These medical histories can be retrieved for study because, for almost a century, the Mayo Clinic and its affiliated hospitals have maintained extensive indexes based on clinical and histological diagnoses made and procedures performed. The Rochester Epidemiology Project also maintains indexes for Olmsted Medical Center and its affiliated hospital, the Olmsted Community Hospital. Information on diagnoses and procedures performed for Olmsted County residents treated as inpatients at other regional facilities, such as the Minneapolis Department of Veteran's Affairs Hospital and the University of Minnesota, both 144 km (90 miles) away in Minneapolis, is also available. The result is the linkage of medical records from essentially all sources of medical care used by the local population in Olmsted County. The completeness of the Rochester Epidemiology Project's capture of data on ophthalmologic care provided to Olmsted County residents has been documented elsewhere.9,10
All medical records of patients with a new diagnosis of OAG, suspected glaucoma, or ocular hypertension made by ophthalmologists between 1965 and 1980 at 1 of the 2 institutions in Olmsted County were reviewed. A precoded data form was used to record demographic and clinical information that included the following: patient's age and sex, visual acuity and IOP, optic disc description and visual field test results at diagnosis, date of initiation of glaucoma therapy, type and subgroup of OAG, and intraocular surgical procedures performed after the diagnosis of OAG. If the patient developed blindness, the date of the diagnosis of blindness, the affected eye, and the cause of the blindness were recorded.
The patients were considered to have OAG or treated ocular hypertension on the date when therapy was initiated, whether medications, laser trabeculoplasty, and/or filtration surgery were used to treat the disease. Those patients in whom glaucoma was suspected or who had ocular hypertension but did not receive therapy were not included in this study. Our definition of OAG included primary OAG, exfoliative glaucoma, and pigmentary glaucoma.
This broadly defined group of OAG cases was then divided into 3 distinct subgroups. Classic glaucoma refers to those patients in whom at least 2 of the following 3 characteristics were noted in the patient record at the time of diagnosis: at least 1 documented IOP of 21 mm Hg or higher, optic nerve damage consistent with glaucoma, and visual field defects in a pattern consistent with that seen in glaucoma. This could include those patients with normal-tension glaucoma. The second subgroup, treated ocular hypertension, included those patients with an elevated IOP of 21 mm Hg or higher at the time of diagnosis but with a normal optic nerve and visual field test result (if performed) in whom therapy was initiated and maintained in an attempt to lower IOP. At the time of diagnosis, a visual field test with a Goldmann, tangent screen, or automated perimeter was performed in most patients. An IOP check and an optic disc evaluation were performed for all patients at the time of diagnosis. A third subgroup, referred to as unsupported glaucoma, included patients treated for OAG without documentation of the previously described clinical features.
For this study, filtration surgery included a trabeculectomy, a full-thickness filtering procedure, or a cyclodialysis procedure. Neither a combined cataract and trabeculectomy operation nor placement of a tube shunt were included as filtration procedures. In this retrospective review, the decision to perform a filtration procedure was not affected by ongoing clinical trials or specific guidelines, but rather was assumed to be based on the clinician's evaluation that the glaucoma was poorly controlled despite treatment with either medications or laser trabeculoplasty.
For patients meeting the study criteria, Olmsted County residence at the time of diagnosis was formally documented by the Department of Health Sciences Research personnel experienced in performing this task. This study was approved by the institutional review boards of the Mayo Clinic and Olmsted Medical Center.
Statistical analysis was performed by using the Kaplan-Meier methods to estimate the 20-year cumulative probability of undergoing filtration surgery in at least 1 eye, as well as in both eyes, adjusted for the period of individual follow-up.
Two hundred ninety-five patients in whom OAG was newly diagnosed were identified during the 20-year period. Primary OAG was diagnosed in 89.8% (265/295), exfoliative glaucoma in 8.5% (25/295), and pigmentary glaucoma in 1.7% (5/295). Based on the classification previously described, 177 patients (60%) were classified as having treated ocular hypertension, 114 (38.6%) were identified as having classic glaucoma, and 4 (1.4%) were classified as having unsupported glaucoma.
Sixty-three percent of patients were women; and most patients (95.9%) were white. The median age at diagnosis of OAG was 68 years; and the mean±SD follow-up time was 15±8 years (range, 9 days to 32 years) (Table 1).
Of the 295 patients, 29 (9.8%) were blind in at least 1 eye at the time of glaucoma diagnosis. All 29 of these patients were included in the analysis of the probability of undergoing a filtration procedure. In these 29 patients, there were 3 eyes that had a filtration procedure after the diagnosis of blindness was made. A total of 48 filtration procedures were performed, of which 41 were guarded trabeculectomies. Seven full-thickness procedures were noted. No cyclodialysis operations were performed during this period.
In Table 2, the Kaplan-Meier cumulative probability of undergoing filtration surgery in at least 1 eye is shown. A 23% probability was calculated when patients with treated ocular hypertension and classic glaucoma were considered together (Figure 1). Separately, the treated ocular hypertension group and the classic glaucoma group were calculated to have a 15% and 39% probability at 20 years, respectively (Figure 2). Because of the nature of Kaplan-Meier analysis, the sum of the probabilities for treated ocular hypertension and classic glaucoma exceeds 23%.
Table 3 shows the Kaplan-Meier cumulative probability of a patient newly diagnosed as having and treated for OAG undergoing a filtration procedure in both eyes. Considering treated ocular hypertension and classic glaucoma together, the probability was 12% at 20 years (Figure 3). When considered separately, the treated ocular hypertension and classic glaucoma groups were calculated to have a 5% and 27% probability at 20 years, respectively (Figure 4).
The results of this retrospective longitudinal study of a community-based population suggest that the probability of undergoing filtration surgery after the diagnosis of OAG is greater in those patients with glaucomatous damage at the time of diagnosis than in patients without evidence of such damage. At 20 years of follow-up, the cumulative probability of filtration surgery calculated by Kaplan-Meier methods was estimated to be 23% in at least 1 eye and 12% in both eyes. In those patients who had glaucomatous damage at the time of diagnosis, ie, classic glaucoma, the probability of undergoing filtration surgery in at least 1 eye was estimated to be 39%, while in both eyes, the estimate was 27%. This was much greater than the probability estimate for treated ocular hypertension, which was 15% in at least 1 eye and 5% in both eyes (Table 2 and Table 3).
The role of surgery in the management of glaucoma has changed since trabeculectomy surgery was introduced more than 30 years ago.11 The traditional American approach of medical therapy followed by laser trabeculoplasty and then filtration surgery in an effort to control IOP has been a subject of debate in the national and international ophthalmic communities. Advancements in surgical techniques, surgical technology, and supplemental agents such as mitomycin have led to increasing success rates for trabeculectomy and, consequently, use of trabeculectomy earlier in the course of glaucoma management.2,12- 15 Options for the reduction of IOP during the past 30 years have increased dramatically. It is possible that this data may overestimate the probability of filtration surgery in current ophthalmic practice because there are many new therapeutic options available, including an expanded arsenal of topical medications.
In this study, a combined cataract and trabeculectomy procedure was not included as a filtering procedure. The rationale for this revolves around the fact that a combined procedure is not always performed because the IOP is poorly controlled. At times, a cataract extraction in a patient with glaucoma is combined with a trabeculectomy in an effort to reduce the number of glaucoma medications. Because this study was retrospective in nature, the ability to determine the exact reason for a combined procedure was limited. Therefore, combined cases were not included in the analysis for the probability of filtration surgery. Eleven combined procedures were identified among the cohort of patients in this study.
The traditional stepped approach to therapy for glaucoma management was common among the patients in this study. Although trabeculectomy was not commonly the first line of therapy for OAG in this population, those patients with classic glaucoma had a 39% probability of having filtration surgery in at least 1 eye and a 27% probability in both eyes at 20 years of follow-up. In a previous study,4 the probability of blindness was calculated by Kaplan-Meier methods for this same cohort of patients. In the treated ocular hypertension group, this was estimated to be 4% in both eyes and 14% in at least 1 eye at 20 years. In the classic glaucoma group, the estimate for blindness at 20 years in both eyes was 22% and in at least 1 eye it was 54%.6 Data from that study, as well as from the present study, demonstrate that patients who have glaucomatous damage at the time of diagnosis are more likely to require filtration surgery during the course of glaucoma management. Those patients are also more likely to develop blindness from OAG than are those patients who have an elevated IOP but no evidence of glaucomatous optic disc or visual field changes. Presumably, patients with glaucomatous damage required a lower IOP and, therefore, surgical intervention occurred more frequently in an effort to achieve a lower pressure. Further analysis of the glaucoma management in those patients who developed blindness compared with those patients who did not develop blindness in this study population may provide further information regarding the development of blindness in OAG.
There are several limitations of this study, including the fact that the small number of patients resulted in fairly wide confidence intervals for the calculated probabilities in some cases. The retrospective nature of the study precluded the consistent availability of standardized documentation of ophthalmic data at the time of examination. The racially homogeneous nature of the population in Olmsted County (98% white in 1980 compared with 83% white for the United States) means that this information may not be applicable to other populations. On the other hand, the stability of this nonreferral population and the uniqueness of the Rochester Epidemiology Project made it possible to assemble long-term follow-up data on a population that is not expected to have overrepresentation of severe disease. Furthermore, in the population studied in this retrospective report, OAG was diagnosed between 1965 and 1980, and data were abstracted until the last follow-up in Olmsted County, with more than 30 years of follow-up available in some cases. In conclusion, the results of this retrospective community-based study of a predominantly white population suggest that most patients newly diagnosed as having and treated for OAG did not undergo filtration surgery as estimated by Kaplan-Meier methods during a 20-year period. However, those patients who had glaucomatous damage at the time of diagnosis were more likely to undergo filtration surgery compared with those who did not have glaucomatous optic disc or visual field damage. This suggests that early diagnosis of OAG is important in the context of filtration surgery.
Accepted for publication March 2, 1999.
This study was supported in part by Research to Prevent Blindness Inc, New York, NY; the Mayo Foundation, Rochester, Minn; and grant AR 30582 from the National Institutes of Health, Bethesda, Md.
Reprints: Douglas H. Johnson, MD, Mayo Clinic, 200 First St SW, Rochester, MN 55905.