Investigators assessed the time required to detect statistically significant visual field (VF) progression on standard automated perimetry (SAP) after comparing various frequencies of testing.
Study design
This observational study included 1,072 eyes of participants with 5 or more VF tests over a 2- to 5-year period. The data were used to derive the longitudinal measurement variability of SAP mean deviation using linear regressions.
The variability estimates were then used to simulate real-world VF data to evaluate the time to detect progression based on various rates of VF loss and testing frequencies.
Outcomes
The authors found that, as expected, the time required to detect VF progression decreases as testing frequency increases, though not proportionally. The results show that 2 reliable baseline tests followed by twice-yearly testing would detect mean deviation loss ranging from -0.5 dB/year to -2 dB/year with sensitivity comparable to detection using thrice-yearly testing.
Additionally, over a period of 10 years, performing 1 confirmatory test decreased the percentage of patients with VF progression to 11% compared with 19% when no confirmatory test was performed. However, 2 confirmatory tests resulted in much smaller specificity gains (9%).
Limitations
This study assumes a linear rate of VF loss, whereas true progression is more likely to be nonlinear or episodic. Additionally, the study relied on trend-based analysis. An event-based analysis would be helpful to evaluate the power and time to detect progression.
Clinical significance
While thrice-yearly VF testing in the first 2 years has been advised to rule out rapid progression, this study suggests that twice-yearly testing is sufficient. As such, the authors recommend obtaining 2 reliable tests at baseline, followed by semiannual testing and confirmation of progression through repeat testing in the initial years of follow-up.
These recommendations are important when considering the impact on the public health system, as well as the rates of false-positive detections with increased testing frequency. However, the findings do not replace the clinician's judgement, especially when caring for patients with advanced VF loss or for younger patients with mild loss who may require more frequent testing.