SHORT-TERM REPEATABILITY OF DIURNAL INTRAOCULAR PRESSURE PATTERNS IN GLAUCOMATOUS INDIVIDUALS1
Realini T, Weinreb N, Wisniewski S.*
Ophthalmology, January 2011
Do patients treated for primary open-angle glaucoma
show a repeatable diurnal IOP pattern?
It is known that IOP varies spontaneously over time,
and a common presumption is that IOP follows a similar
circadian pattern from day to day.2-6 This assumption
supports the clinical practice of scheduling patients' visits
at similar times of the day so as to minimize IOP variation
between visits. However, only limited and conflicting
data, gathered without Goldmann tonometry, characterize
the reproducibility of circadian IOP patterns
over multiple 24-hour periods.7,8 Given this limitation
and the recent report that nonglaucomatous eyes did
not have a repeatable intervisit diurnal pattern,9 further
investigation of the repeatability of diurnal patterns in
glaucomatous eyes is also warranted. Realini et al conducted
a prospective, observational cohort study of the
short-term repeatability of diurnal IOP measurements in
patients with primary open-angle glaucoma (POAG).
Forty-seven individuals were enrolled, and their 12-hour diurnal IOP curves were recorded using Goldmann tonometry at two visits 1 week apart. Corresponding IOP measurements at each time point were compared from the first visit to the second visit, and the changes in corresponding IOP measurements over periods between those time points were compared as well. The authors reported that IOP was generally lower at the second visit than at the first visit. Average agreement of corresponding individual IOP values between the first and second visits was fair to good, as characterized by an intraclass correlation coefficient. However, average agreement of corresponding IOP changes over time was consistently poor, over both shorter and longer time periods during the diurnal curve.
The authors concluded that diurnal IOP patterns are not repeatable in treated POAG patients. They cited their earlier study of diurnal IOP measurements in nonglaucomatous eyes to conclude that POAG does not disrupt or mask an existing inherent diurnal pattern.9 Furthermore, they stated that their study does not provide evidence for the theories that treatment with IOP-lowering drops masks an existing diurnal pattern or creates a new one in POAG patients. The authors then considered possible contributing factors to the lower average IOP noted at the second visit; they added that lower IOP was not likely to be caused by equipment or personnel bias, environmental factors, or an improvement in therapeutic compliance. Given that the same phenomenon was observed in several groups of nonglaucomatous eyes,9,10 the authors propose the possibility of “white-coat” ocular hypertension at the first study visit, supported by the relationship between cortisol levels and IOP.
Although the study's data were collected during a limited number of visits, included only daytime IOP measurements, and involved mainly a white population, the study was strengthened by several factors. Its design was prospective, and IOP was measured by Goldmann tonometry using a strict measurement protocol based on the Ocular Hypertension Treatment Study (OHTS). Even though the ethnic diversity of the study population was limited, the results were applicable to clinical practice in that participants were on a stable medication regimen with environmental factors (eg, diet and physical activity) not standardized.
In summary, the authors conclude that single-day diurnal IOP measurements do not accurately capture IOP variation over time, and they stated that standardizing the time of IOP measurements might not be the best method of minimizing IOP variation. They suggested that multiple pre- and posttreatment IOP measurements might be a better alternative, but details of these measurements have yet to be established.
*Financial disclosures: The authors stated that they hold no proprietary interest in the materials discussed herein.
INTEROBSERVER AGREEMENT AND INTRAOBSERVER REPRODUCIBILITY OF THE SUBJECTIVE DETERMINATION OF GLAUCOMATOUS VISUAL FIELD PROGRESSION11
Tanna AP, Bandi JR, Budenz DL, et al.*
Ophthalmology, January 2011
How much interobserver and intraobserver variability
exists in the determination of visual field progression?
Is glaucoma progression analysis software useful in
determining progression?
Although visual field testing is a crucial part of identifying
glaucomatous progression, the determination of
visual field progression itself is not straightforward.
Long-term fluctuation is common,12-16 and gold standard
criteria for visual field progression do not exist.
There is often little agreement among experts on the
classification of series of visual fields as progressive or
not.13,17-19 Although several automated visual field
analysis algorithms are available, ophthalmologists
often resort to a subjective review of serial visual field
tests to determine progression, a method the reproducibility
and interobserver variability of which are
largely unknown.16,18,20,21
The authors of this study sought to identify the extent of interobserver agreement and intraobserver reproducibility of visual field progression classification. They investigated whether automated progression analysis software lessens interobserver variability.
One hundred eyes of 83 patients with baseline visual field loss were retrospectively enrolled, and five serial Humphrey visual field tests (Carl Zeiss Meditec, Inc., Dublin, CA) were obtained for each eye. The principal investigator selected series of fields that demonstrated either stability or progression. The de-identified visual field printouts were then shared with five subspecialtytrained “expert readers,” who were blinded to the purpose of the study and the identities of the other experts. The expert readers evaluated the series three times: first to establish interobserver agreement, second to evaluate intraobserver variability, and third with the Glaucoma Progression Analysis (Carl Zeiss Meditec, Inc.) results available to determine their effect on interobserver agreement. The series were divided into four categories of progression—none, questionable, probable, or definite.
Majority consensus for dichotomized categories (progression or not) agreed 89% of the time in rounds one and two of analysis. Intraobserver reproducibility over the three rounds of analysis was “good to excellent” for the aforementioned raw classification categories, but reproducibility for the analysis of dichotomized categories was “fair to good.” There were no significant differences in interobserver agreement from rounds one to three. These results contrast with a 2003 study of another progression analysis program, which found that the software significantly improved intraobserver and interobserver variability.21
The authors speculated that the level of interobserver agreement might be due to the fact that the different experts used different criteria for progression. The lack of improvement with progression software might reflect diffuse progression not detected by the software or progression detected by the software that was classified as artifact by experts. The study was limited by the lack of “official” criteria for visual field progression, undefined progression-likelihood categories, and the exclusion of visual fields with ambiguous findings, which might have artificially lowered interobserver variability. The study conditions did not resemble clinical practice in that visual fields could not be repeated when possible progression was noted. The experts were also masked to all information except visual fields and could not determine whether other clinical data would confirm or refute possible progression. Finally, although this study suggested that progression analysis software did not improve interobserver agreement, the visual field graders were all experienced, glaucoma-trained subspecialists. The study did not answer the question of whether automated glaucoma progression analysis benefits comprehensive ophthalmologists and others who may be less experienced in visual field evaluation.
*Financial disclosures: Angelo P. Tanna, MD, is a consultant to and receives honoraria for lectures from Carl Zeiss Meditec, Inc. This study was supported in part by research grant P30 EY-014801 awarded by the National Eye Institute, Bethesda, Maryland; by an unrestricted grant from Carl Zeiss Meditec, Inc.; and by an unrestricted grant from Research to Prevent Blindness, New York, New York.
META-ANALYSIS OF ONE- VERSUS TWOSITE PHACOTRABECULECTOMY22
Gdih GA, Yuen D, Yan P, et al.* Ophthalmology, January 2011
Does one- or two-site phacotrabeculectomy lower
IOP or decrease complications more?
Age is a risk factor for both cataracts and glaucoma,
two of the three most common causes of blindness
worldwide,23 and the two conditions commonly coexist.
Combined phacotrabeculectomy is one of several
options for the surgical treatment of these two conditions.
However, there is mixed evidence on the advantages
of one procedure over another. A 2002 review of
the surgical treatment of coexisting cataract and glaucoma
determined that there was weak support for
1 to 2 mm Hg of extra IOP lowering with two- versus
one-site phacotrabeculectomy.24 The current study is a
meta-analysis that reviewed prospective randomized
controlled trials of one- versus two-site phacotrabeculectomy.
Trials with at least 12 months of followup
were included, and the main outcome measure was
a reduction in IOP from baseline. Ten articles were included in a qualitative synthesis,25-34 and seven of
those26,29-34 were included in the quantitative
meta-analysis.
In studies that reported glaucoma patients' diagnoses, most included open- and closed-angle glaucoma. Patients with uveitic, neovascular, phacomorphic, steroid-induced, and traumatic glaucomas were commonly excluded. Surgical and anesthetic techniques as well as postoperative regimens did not differ significantly between one- and two-site surgeries. In the studies comparing surgical time, the duration of onesite surgery averaged 13 minutes less than two-site surgery. Although IOP reduction was numerically greater for one- versus two-site surgery at all follow-up points except 12 months, the differences were not statistically significant. There were also no statistically significant differences in reductions in glaucoma medication or changes in visual acuity between the two groups. In those studies that reported complications, the rates of intraoperative and postoperative complications were similar for one- and two-site surgeries.
In summary, the authors reported that their metaanalysis showed no statistically significant differences in primary or secondary outcome measures between oneand two-site phacotrabeculectomy. In their absence, factors that incorporate surgeons' preferences and individual patients' characteristics might take precedence when choosing between one- and two-site phacotrabeculectomy.
*Financial disclosures: The authors cited that they hold no proprietary interest in the materials discussed herein.
Section Editor James C. Tsai, MD, is the chairman and Robert R. Young professor of ophthalmology and visual science at Yale University School of Medicine in New Haven, Connecticut. He acknowledged no financial interest in the products or companies mentioned herein. Dr. Tsai may be reached at (203) 785-7233; james.tsai@yale.edu.
Kristy Ahrlich, MD, is an instructor and glaucoma fellow at Yale University Department of Ophthalmology and Visual Science in New Haven, Connecticut. She acknowledged no financial interest in the products or companies mentioned herein. Dr. Ahrlich may be reached at (203) 785-2020; kristy.ahrlich@yale.edu.
