Repeatability and Concordance of Clinical Measurement of Amplitude of Accommodation

Abstract

Purpose:: Clinical measurement of the amplitude of accommodation (AA) provides an indication of the maximum accommodative ability. In order to demonstrate if a significant change in the AA has occurred, it is important to determine the degree of repeatability of this measurement. The aim of the present study was to compare the measurements of AA determined by using three different subjective clinical methods, including the push-up, push-down and minus lenses. These techniques differ in terms of either the apparent size of the target, the end point used or the accommodative components stimulated. The hypothesis was that the degree of repeatability of the three methods would be quite different and that the three techniques could not be interchangeable.

Methods:: The AA of the right eye was measured on two separate occasions in 61 visually normal subjects with a mean age of 19.7 years (range 18 to 32 years). All measurements were recorded by the same examiner. The repeatability of the tests and the concordance between tests was estimated by using the statistical method of Bland y Altman. We determined the mean difference (bias), the standard deviation of the differences (SD) and the 95% limits of agreement, i.e. the range over which 95% of values would be predicted to lie (bias ± (1,96 x SD)). When the Anderson-Darling test for normality was significant, we computed the 95^th percentile of the absolute differences instead of the 95% interval of agreement.

Results:: The 95% limits of agreement for the push-up, push-down and minus lens techniques were ±4.76, ±4.00 and ±2.52 D, respectively. The minus lenses method presented the best repeatability, with the smaller bias (-0.08 D) and the smaller 95% interval of agreement.It was observed higher values when measuring the AA with push-up technique, compared with the push-down and minus lens procedures. The biases were high in clinical terms, always above to 1.75 D, and the interval between the 95% limits of agreement correspond to substantial differences, always higher than ±4.50 D. The highest interval corresponds to the comparison of AA from minus lenses and from the push-up method (±5.65 D).

Conclusions:: The minus lenses method presented the best repeatability, with the smaller bias (-0.08 D) and the smaller 95% interval of agreement. The degree of agreement between the three techniques was poor.

Keywords: clinical research methodology