There is no agreed method for determining rapidity of development of disability in patients with multiple sclerosis when each patient has had only a single assessment in the course of the disease. Previously a variety of methods have been tried including demarcating arbitrary severe and benign subgroups, the progression index (Expanded Disability Status Scale (EDSS) divided by disease duration) and variations on this method. The Multiple Sclerosis Severity Score (MSSS) provides a novel approach, relating scores on the Expanded Disability Status Scale (EDSS) to the distribution of disability in patients with comparable disease durations. Effectively the MSSS assigns to each EDSS its median decile score within this distribution. Thus a patient having an MSSS of x progresses faster than 10x% of the population and slower than (100-10x)%. So, for example, an MSSS of 5.0 means progressing at the median rate. A patient whose MSSS is 9.0 is a fast progressor, progressing faster than 90% of patients. A patient whose MSSS is 1.0 is a slow progressor, progressing faster than just 10% of patients.
Full details are to be found in Roxburgh, Seaman et al The Multiple Sclerosis Severity Score, accepted by Neurology 2004. Please quote this reference if you use MSSS in publications.
We applied the MSSS method to 9981 patients from 11 different countries to derive the global MSSS, a reference table for future disability comparisons. Thus anyone with known EDSS and disease duration can be assigned a global MSSS score, simply by finding the appropriate column and row in this table. Alternatively a local MSSS table can be generated for any particular set of data being analysed, provided it is large enough. However, unless there is some significant reason for doing so e.g. because the sample has a peculiarly fast or slow rate of disease progression, we would recommend simply applying the global MSSS.
Comparison of disease progression between two or more groups who differ for some variable can be made using the MSSS scores (whether using the global or a local table) and the Kruskal-Wallis test in the different groups. The Kruskal-Wallis test is similar to the ANOVA test, but is a non-parametric test, i.e. it does not assume normally distributed data. If there are only two genotype groups, the Kruskal-Wallis test is identical to the Wilcoxon test (also called the Mann Whitney U test). In simulations we have shown that the MSSS is the most powerful method devised to date to make these comparisons.
A typical variable would be genotype - though other variables could also be examined for their effect on disease progression provided that recruitment is not affected by this variable other than through its direct effect on disease progression.
MSSStest performs the following tasks
For more information about MSSStest please click here: MSSStest.
To download MSSStest click here: DOWNLOAD.
If you have would like any information about the MSSS or have any comments about the program please email Richard@Roxburgh.net.nz