Repeatability and reproducibility of applanation resonance tonometry: a cross-sectional study

BACKGROUND: To assess repeatability (intra-observer variability) and reproducibility (inter-operator variability) of intraocular pressure (IOP) measurements with servo-controlled Bioresonator Applanation Resonance Tonometry (ART) and to evaluate possible influential factors. METHODS: The study included 178 patients (115 glaucoma and 63 controls; one eye per subject). IOP was measured once with a Goldmann applanation tonometer (GAT) and twice by ART (ART1, ART2), in randomized sequence, by a single operator to assess intra-operator variability. Each ART measurement consisted on 3 readings. To assess inter-operator variability 2 evaluators performed 2 measurements each (in random order) on the same patient. Repeatability and reproducibility were assessed by the coefficient of variation (CoV) and intraclass correlation coefficient (ICC). RESULTS: In the entire cohort, ART1 was 0.4 ± 2.2 mmHg (-7.0 to 5.7 mmHg) higher than ART2 (p = 0.03) regardless of test order. Intra-operator CoV was 7.0% ± 6.3%, and ICC was 0.80-0.92. Inter-operator CoV ranged between 5.7% ± 6.1% and 8.2% ± 7.2%, and ICC between 0.86 and 0.97. ART1 and 2 were respectively 1.7 ± 3.1 and 1.3 ± 3.1 mmHg higher than GAT (p < 0.01). Test-retest difference with ART fell within ±1 mmHg in 41% of cases, within ±2 mmHg in 70%, within ±3 mmHg in 85%. 15% had a test-retest difference higher than ± 3 mmHg; Bland-Altman 95% intervals of confidence were -3.9 and +4.6 mmHg. Results were unaffected by age, diagnosis, central corneal thickness, keratometry, operator, randomization sequence. CONCLUSIONS: In most cases ART repeatability and reproducibility were high, with no differences due to patients' characteristics. ART measurements overestimated GAT by a mean of 1.3-1.7 mmHg.