Heijl A, Lindgren G, Olsson J
Department of Ophthalmology, University of Lund, Sweden.
Arch Ophthalmol. 1987 Nov;105(11):1544-9. doi: 10.1001/archopht.1987.01060110090039.
We assessed the variability of results in normal subjects of computerized static threshold perimetry of the central 30 degrees field. Variability of measured threshold values was highly dependent on eccentricity. This included variability among individuals, test-to-test variability within individuals, and intratest variability. All values were significantly larger in the midperiphery than centrally. We found that the mean sensitivity decrement with age was eccentricity dependent, so that the age-corrected normal visual field became not only depressed but also steeper with age. Distributions of individual pointwise deviations from the age-corrected normal mean thresholds were significantly nongaussian. The dependency of variability on test point location, the nongaussian distributions of deviations from age-corrected means, and the variability of age-induced sensitivity reduction should all be considered in the interpretation of computerized visual fields, and particularly in the design of statistical programs for field analysis. Programs not considering these factors are likely to result in misleading analyses.
我们评估了正常受试者中央30度视野计算机静态阈值视野检查结果的变异性。测量阈值的变异性高度依赖于偏心度。这包括个体间的变异性、个体内测试间的变异性以及测试内变异性。所有数值在中周部均显著大于中央部。我们发现,平均敏感度随年龄的下降与偏心度有关,因此经年龄校正的正常视野不仅随年龄降低,而且随年龄变得更陡峭。个体与经年龄校正的正常平均阈值的逐点偏差分布显著非高斯。在解释计算机视野检查结果时,尤其是在设计视野分析统计程序时,应考虑变异性对测试点位置的依赖性、与经年龄校正均值的偏差的非高斯分布以及年龄引起的敏感度降低的变异性。不考虑这些因素的程序可能会导致误导性分析。