Multiple Sclerosis, 1 October 2002, vol. 8, no. 5,
Uitdehaag B.; Adèr H.; Roosma T.; de Groot V.; Kalkers N.; Polman C.
 Departments of Neurology and Clinical Epidemiology and Biostatistics, VU Medical Centre, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands  Department of Clinical Epidemiology and Biostatistics, VU Medical Centre, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands  Department of Neurology, VU Medical Centre, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands  Department of Rehabilitation Medicine, VU Medical Centre, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands
The Multiple sclerosis functional composite (MSFC) has been recommended as a clinical outcome measure to be used in future MS trials.
A specific characteristic of the MSFC is that it is defined as a measure of impairment relative to a reference population.
Using different reference populations affects actual MSFC scores.
If the selection of a reference population also has an effect on sensitivity to change of the MSFC, comparison of data from clinical trials will be almost impossible when different reference populations are used.
We studied the effect of the selection of a reference population on the outcome of a trial by simulating 343 intervention trials and comparing results obtained by using three different reference populations: two previously published MS patient populations and a healthy population.
Scores of the healthy population were collected in the first part of the study.
The effects of sex, age and education level on test scores of healthy subjects were studied as well.
In the healthy controls, sex, age and education level had a different impact on individual test scores of MSFC components and overall MSFC score.
Our study shows that, with the use of the MSFC, the selection of different reference populations does not affect the trial statistics and significance, but it does affect comparability of results between different trials, and complicates the clinical interpretation of any observed change.