[Skip to Navigation]
June 1990

Cox Regression Analysis in Clinical Research

Author Affiliations

From the National Eye Institute, Biometry and Epidemiology Program, National Institutes of Health, Bethesda, Md.

Arch Ophthalmol. 1990;108(6):888-889. doi:10.1001/archopht.1990.01070080132051

The Cox survivorship model1 has been used widely in medical research and recently in the ophthalmic literature to analyze prognosis following alternative treatments of uveal melanoma.2-5 The purpose of this short communication is to explain what Cox regression analysis is, so that investigators will understand its role in clinical research.

The Cox model is a particular type of survivorship model. More generally, survivorship models are used to describe and interpret the rate at which deaths occur over time in some population. The clinical investigator may use these models to interpret the rate at which morbid events occur as well, such as loss of vision, progression of retinopathy, and marked increase in intraocular pressure. When investigators are interested in determining how a variety of factors affect the rate at which events occur (eg, treatment, characteristics of the disease, or of the patient), they may wish to use the Cox

Add or change institution