If none of the above methods allow calculation of the standard deviations from the trial report (and the information is not available from the trialists) then, in order to perform a meta-analysis, an author may be forced to impute (‘fill in’) the missing data or to exclude the study from the meta-analysis: see Chapter 16 (Section 16.1.3). A narrative approach to synthesis may also be used. It is valuable to tabulate available results for all studies included in the systematic review, even if they cannot be included in a formal meta-analysis.