The interpretation of changes in serial intraocular pressure (IOP) measurements are critical in the diagnosis and management of glaucoma. Differentiating true change from test-retest variability (TRV) in IOP measurements using Goldmann applanation tonometry (IOPG) is essential when reviewing these measurements. We, therefore, determine clinically applicable guidelines for differentiating TRV from the true change in IOPG by quantitatively summarizing the published evidence from 18 studies.
Studies reporting TRV in clinically meaningful ways, such as the coefficient of repeatability or the limits of agreement, were collated and analyzed. A single linear model quantitatively summarized the TRV for the commonly encountered combinations of patient types and intraobserver versus interobserver testing. The best model determined n-fold effects providing asymmetric SEs.
For the common intraobserver/interobserver visit combination the model results for healthy subjects and glaucoma/ocular hypertensive patients were TRV (±SE) values of 3.16 (2.67 to 3.74) and 4.26 (3.12 to 5.84) mm Hg, respectively. For the interobserver/intervisit case for glaucoma/ocular hypertension patients, the estimated TRV was 5.50 (3.43 to 8.76) mm Hg, for which no published coefficient of repeatability or limits of agreement data exists at present. The overall model was significant at P<0.018.
Potential improvements in the diagnosis and management of glaucoma are made by providing clinicians with well-estimated values for TRV in IOPG derived from the evidence base. These values not only enable individual clinicians to better interpret changes in IOPG, but they also provide a basis for standardizing the interpretation of changes in IOPG between clinics, clinicians, and patient types.
John Curtin School of Medical Research, Australian National University, Acton, Australian Capital Territory, Australia
The John Curtin School of Medical Research at the Australian National University paid the publication fees for this review.
Disclosure: The authors declare no conflict of interest.
Reprints: John G. Pearce, PhD, P.O. Box 133, Parkes, NSW 2870, Australia (e-mail: email@example.com).
Received September 18, 2018
Accepted November 14, 2018