Assessing Precision of Hodapp-Parrish-Anderson Criteria forStaging Early Glaucomatous Damage in an Ocular

Hypertension Cohort: A Retrospective StudyTutul Chakravarti, MBBS, DO, DNB, PG Dip (in Epidemiology)

original clinical study

Purpose: the aims of the study were (1) to stage early glaucomatous field defects following Hodapp-Parrish-anderson (HPa) criteria in an ocular hypertension (oH) cohort with 7-year follow-up and (2) to recog-nize the limitation(s) of this staging system.Design: a retrospective review of subjects with oH.Methods: a retrospective review of the visual fields of 183 oH sub-jects was conducted. a Humphrey Field analyzer was used for standard automated perimetry in all these participants, tested once every half year for 7 years or until the onset of conversion (study end point). only nerve fiber bundle glaucomatous abnormalities were identified as per the ocu-lar Hypertension treatment study guideline and considered as converted glaucoma cases from oH. subsequently, those fields were classified as possible, probable, and definite glaucoma cases following HPa criteria. this study also recognized the limitations of glaucoma Hemifield test (gHt) interpretation and the depth and distribution of real scotoma. the criterion for reproducible abnormalities was 3 consecutive abnormal fields with defects in a similar location.Results: Early glaucomatous defects were identified in 22 patients (27 fields) of 183 oH subjects. seven of 27 eyes were identified as possible glaucoma, 12 as probable glaucoma, and 8 as definite glaucoma. the gHt was interpreted as within normal limits for 5 cases, borderline for 11 cases, and outside normal limits for 11 cases.Conclusions: Hodapp-Parrish-anderson criteria can recognize subtle nerve damage to diagnose early glaucoma. Visual field abnormalities classified as definite glaucoma cases are less prone to false labeling than normal fields, whereas field defects identified as probable/possible glau-coma may remain underdiagnosed for imprecise gHt interpretations and poor specification of cluster analysis in pattern standard deviation.

Key Words: ocular hypertension, oHts criterion, visual field, standard automated perimetry (saP), gHt, Psd, Hodapp-Parrish-anderson criteria

(Asia-Pac J Ophthalmol 2017;1:2127)

Glaucoma, a chronic progressive optic neuropathy, is one of the leading causes of blindness worldwide.1 For diagnosing

and managing glaucoma, perimetry is elementary. reproducible field defects demonstrated by test results are the most conclusive and definite means of confirming a diagnosis of chronic open-angle glaucoma.2 the primary endpoints for the ocular Hyperten-sion treatment study (oHts) were the development of either a confirmed visual field (VF) abnormality or a confirmed clinically significant optic disc abnormality attributed to primary open-an-gle glaucoma (Poag).3 in the oHts endpoints for Poag con-version were reached by both disc and VF findings in up to 10% of the cases, by disc only in around 50%, and by VF only in ap-proximately 40%.4 to identify the earliest glaucomatous damage and borderline cases, Hodapp-Parrish-anderson (HPa) criteria for minimal abnormality are the most accepted method, widely used in offices and clinical studies for the past 40 years.5 the 3 minimum criteria for diagnosing acquired glaucomatous damage are as fol-lows: (1) a glaucoma Hemifield test (gHt, Fig. 1) outside normal limits (onl) on at least 2 fields; (2) a cluster of 3 or more nonedge points in a location typical for glaucoma, 2 of which are depressed on the pattern deviation plot at a P value of less than 5% and 1 of which is depressed at a P value of less than 1% on 2 consecutive fields; or (3) a pattern standard deviation (Psd) that occurs in less than 5% of normal fields on 2 consecutive fields.

according to oHts protocol, an abnormal VF is described as having a gHt onl and/or a corrected pattern standard devia-tion with a P value of less than 0.05.6 they also included ander-sons second criteria as an abnormality in their secondary defini-tion.6 the structure and Function Evaluation study findings also indicated that gHt, onl, and Psd locations worse than the 5% probability level had high specificity for detecting the develop-ment of early glaucomatous VF loss.7

although the oHts group attempted to classify the types of initial VF defects and recorded incidence of such defects during fol-low-up, they did not question the accuracy of this staging method, which could cause underdiagnosis of the borderline and interme-diate stages of glaucoma disease.6 However, we aimed to analyze and classify early field change of 183 ocular hypertension (oH) subjects at the moment of their glaucomatous conversion in this ret-rospective study on the basis of anderson criteria and by verifying gHt interpretations and the depth and location of real scotomas.

the gHt onl is considered to be the single most efficient method of VF interpretation and is claimed to be the most ac-curate among these 3 criteria with worldwide approval.2,5 the ac-curacy of the gHt was challenged by Zalta,8 who demonstrated that the 5 mirror-image zones of the gHt were not as sensitive in detecting early loss as had been previously thought. Katz et al9 also found disagreement in gHt interpretations.

although the precision of gHt analysis was questioned, none verified all 3 criteria as a whole to check the overall precision of this classification method at the moment of glaucomatous conver-sion in an oH cohort. ocular hypertension participants included

From the Eye and glaucoma care and Vivekananda institute of Medical sciences, West Bengal, india.

received for publication august 30, 2015; accepted February 23, 2016.Verifying the accuracy of glaucoma Hemifield test for detection of early

glaucomatous visual field loss: a study was presented at the international congress on glaucoma surgery 2010, new delhi, and at aioc 2015, new delhi.

the author has no funding or conflicts of interest to declare.reprints: tutul chakravarti, MBBs, do, dnB, Pg dip (in Epidemiology), Eye

and glaucoma care, 18/32, dover lane, Kolkata-29, Pin-700 029, West Bengal, india. E-mail: drtutul@hotmail.com; titirchakravarti@gmail.com.

copyright 2016 by asia Pacific academy of ophthalmologyissn: 2162-0989doi: 10.1097/aPo.0000000000000201

Copyright 2017 Asia-Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.

Asia-Pacific Journal of Ophthalmology Volume 6, Number 1, January/February 2017 www.apjo.org | 21

in this study were without clinically evident optic disc or VF dam-age at the study entry point.

the purpose of this study was to classify the early glauco-matous VF abnormalities on the basis of the 3 HPa criteria for minimal abnormality in an oH cohort at 7-year follow-up and to understand the limitation(s) of each criterion.

materials and methodsthis was a retrospective cohort study from october 2010 to

december 2012. the participants included 183 oH subjects, with a mean sd age of 62.64 11.33 years, intraocular pressure (ioP) between 22 and 32 mm Hg, and normal VFs with standard automated perimetry (saP) on at least 2 separate occasions (table 1). all participants included in this study were of white ethnicity without clinically evident optic disc damage and best-corrected snellen visual acuity of at least 20/40 in both eyes.

slit-lamp findings for all participants were within normal limits (Wnl) and showed open angles on gonioscopy at the study

entry point. subjects with existing ocular or systemic comorbidi-ties known to possibly affect the VF were excluded from this study. Participants with diabetes mellitus were also excluded from this study because it could affect the VF. according to protocol, we ex-cluded oH subjects with a history of intraocular surgery (except for

table 1. Demographics and Ophthalmic Status of 183 Ocular Hypertensive Subjects (366 Eyes/VFs)

figure 1. The GHT compares pattern deviation probability scores in 5 zones in the upper field with corresponding scores in the mirror-image lower field. Reprinted with permission from The Field Analyzer Primer, Essential Perimetry, 3rd ed.

demographics Mean sd (range) age, y race BcVa Mean sd follow-up, yophthalmologic characteristics ioP, mm Hg VFs

global indices Mean sd saP Md, dB Mean sd saP Psd, dB

table 3. Classifications and Clinical Characteristics of the 27 Eyes/Visual Fields With Early Glaucomatous Damage

Classifications Based on HPA Criteria

Possible glaucomaProbable glaucomadefinite glaucoma

64.62 11.33 (34106)White

20/40 in both eyes7.1 3.4

>22 and