Table 6.

Limitations and caveats of multidimensional measures

Interpretation may not be straightforward, particularly if clinical relevance of (some) components are not immediately obvious

An overall score lacks a clear dimension, which complicates the interpretability of the score

Components should be normalized or weighted without obscuring the clinical meaning

Components may shift in opposite directions (improvement vs harm) which might obscure interpretation of treatment efficacy

Components should capture the expected (biological) effects of the intervention under investigation

Increasing the number of components not necessarily increases sensitivity

Redundant components might cause a large change in the composite score in patients that have symptoms in that domain, while the change may be smaller or absent in patients with symptoms in other domains

Increasing sensitivity to change does not necessarily lead to higher sensitivity for treatment effects

Dichotomization of the results (e.g. ‘no evidence of disease activity’) will inherently cause loss of information