Systematic bias in real-world tonometry readings based on laterality?

Abstract

In research settings, the first eye examined tends to have a higher intraocular pressure (IOP) than the second. We sought to verify whether clinicians in Yorkshire, UK, measure IOP in right eyes before left and whether such behavioural factors affect IOP readings at the population level. We observed 128 IOP measurements taken by 28 ophthalmologists using Goldmann applanation tonometry (GAT) over a 4-month period in 2018, recording which eye was examined first. All IOP measurements on electronic patient records for Leeds Teaching Hospitals NHS Trust, UK, between January 2002 and June 2017 were extracted, yielding IOP readings for 562,360 eyes, analysed for evidence of systematic bias in IOP measurement. Right eye IOP was measured before left in 112/128 observations (87.5% (95% CI: 75.2%-94.2%)). For IOP measured by GAT, there was no statistically significant difference (p = 0.121) between right and left eye IOP (mean IOP 16.95 and 16.96 mmHg, respectively). Even values of IOP were reported more frequently than odd values (136,503/214,628 (63.6%) were even). Identical IOP readings for both eyes were recorded in 124,392/254,380 patients (48.9%) who had both eyes measured. Our study found no IOP difference based on laterality, but strong evidence of certain trends associated with IOP measurement by GAT, such as a preference for even values and the same IOP being recorded for both left and right eyes. Such effects may be explained by behavioural aspects of GAT and suggest that there are substantial opportunities for improvement in the way GAT is utilised in real world settings.