There are considerable challenges to conducting meta-analyses of costing data given the large variation in costs across healthcare settings. Further, costing data is often positively skewed. Recently, the ratio of means (RoM) and ratio of geometric means (RoGM) have been proposed as methods to pool continuous outcomes data from trials when outcomes are expressed in different units (or different settings) and/or when continuous outcomes are positively skewed. The objective of this study was to review and apply RoM and RoGM methods for synthesizing costing data from a set of randomized controlled trials comparing the costs for thyroidectomy using a harmonic device versus conventional techniques. In this illustrative example, we describe methods for assessing skewness of the data, converting costs to a common currency, adjusting for inflation and economic development differences, combining costing data using both RoM and RoGM, illustrating findings in an intuitive way for decision makers, and assessing heterogeneity. We also discuss advantages and disadvantages of applying the RoM and RoGM approaches for combining costing data. We conclude that there are situations where costing data can be combined despite the methodological challenges. However, analyses combining costing data should be accompanied by a detailed assessment of heterogeneity. As well, the conclusions drawn from such analyses are more robust when there is consistency between effect estimates across studies.