advertisement
Editors Selection IGR 12-4
Progression: A linear model
Comment by Ananth Viswanathan on:
24975 Prediction of glaucomatous visual field loss by extrapolation of linear trends, Bengtsson B; Patella VM; Heijl A, Archives of Ophthalmology, 2009; 127: 1610-1615
See also comment(s) by Felipe Medeiros •It is reassuring to see that, after fifteen years of research into the subject, visual field deterioration in glaucoma is still well described by a linear model of the rate of change. In 1995, McNaught and colleagues1 fitted a variety of models, linear amongst them, to the first five visual fields in a series and measured the ability of the models to predict the final (15th) visual field in the series. Although they tested 221 different models including quadratic, cubic, complex polynomial and exponential, they found that the linear model gave the best prediction, with a mean error of -7dB.
After fifteen years of research into the subject, visual field deterioration in glaucoma is still well described by a linear model of the rate of change
Bengtsson et al. (1721) used a very similar approach, using a linear model fitted to the first five visual fields in a series to predict the full dataset. The model performed well in the present study, with the model which was based on the first five fields being within 10% of a model based on the entire dataset.
Patients in whom linear regression analysis suggests dangerously rapid rates of visual field progression may be candidates for significant alterations in therapy
An important difference between the present study and the previous work by McNaught et al.1 is that the present study used a summary measure of the visual field (the Visual Field Index) rather than examining individual points. Although the use of summary measures may be relatively insensitive to change as compared to pointwise measures (Chauhan, et al.2; Birch, et al.3), it may be of practical use if other more sophisticated analyses are not available. (Chauhan, et al.4) The authors correctly point out that patients in whom linear regression analysis suggests dangerously rapid rates of visual field progression may be candidates for significant alterations in therapy.
Conflict of interest: I wrote the PROGRESSOR software which implements pointwise linear regression of visual field data.
References
- McNaught AI, Crabb DP, et al. Modelling series of visual fields to detect progression in normal tension glaucoma. Graefes Arch Clin Exp Ophthalmol 1995; 233: 750-755.
- Chauhan BC, Drance SM, et al. The use of visual field indices in detecting changes in the visual field in glaucoma. Invest Ophthalmol Vis Sci 1990; 31: 512-520.
- Birch MK, Wishart PK, et al. Determining progressive visual field loss in serial Humphrey visual fields. Ophthalmology 1995; 102: 1227-1234; discussion 1234-1235.
- Chauhan BC, Garway-Heath DF, et al. Practical Recommendations for Measuring Rates of Visual Field Change in Glaucoma. Br J Ophthalmol 2008.
Comments
The comment section on the IGR website is restricted to WGA#One members only. Please log-in through your WGA#One account to continue.Log-in through WGA#One
