Abstract
Purpose: :
Conventional Supra-Threshold (ST) testing is a rapid method for identifying abnormal points in the visual field (VF). Multisampling ST (Artes et al IOVS 2003 44:2582-7) gives better diagnostic precision but with an increase in test time. We propose a new ST testing strategy: Enhanced Supra-Threshold Testing Algorithm (ESTA) which resamples adaptively and we compare it to conventional testing paradigms in pilot study of patients and healthy subjects.
Methods: :
ESTA begins by testing all points once at a threshold calculated from the 95 centile of an age corrected normal response. These initial results are combined with spatial relationships between VF points (Strouthidis et al, IOVS 2006; 47:5356-5362), to form a cluster analysis giving a pointwise likelihood of abnormality. ‘Less definite’ locations were then selected for retesting. ESTA parameters were optimised for the Moorfields Motion Displacement Test (MDT) using frequency of seeing data from MDT constant stimuli experiments, in 26 patients. Reference limits for normality were set from the Moorfields MDT normative database, which comprises 120 healthy subjects. ESTA was then piloted on 13 glaucomatous patients (mean age: 66.5 [31,77] yrs; mean defect: -4.1 [range -0.5, -8.8 dB]) and 13 age-matched healthy subjects. Each subject underwent: ESTA, 2/3 MultiSampling (2/3MS) and Standard Automated Perimetry (SAP SITA standard - Humphrey Field Analyser).
Results: :
In this sample, ESTA (Mean Test Time (MT) = 92sec range:78sec to 135sec) was significantly faster (on average 3 times quicker) to perform than SAP SITA (p<0.001, MT = 352 sec; range: 226sec to 417sec). ESTA was also on average 75% quicker to perform than 2/3MS (p<0.001, MT=170sec; range 156sec to 208sec). ESTA showed >80% pointwise agreement for ‘defective’ points with locations flagged at the 99.5% level on Pattern Deviation SAP maps and those locations flagged by 2/3MS. Using the two-point fail criterion; 12 out of 13 patients and 0 healthy subjects failed with both ESTA and with SAP SITA (99.5% PD maps); while with 2/3MS, 13 glaucoma patients failed alongside 2 healthy subjects. This glaucoma patient 'missed' by ESTA had small dense defect, which was detected by ESTA but resulted with only a single point flagged.
Conclusions: :
ESTA offers a promising new strategy for rapid case-finding perimetric tests for visual field loss
Keywords: perimetry • visual fields