Dear Tung Nguyen Thanh

The Cochrane handbook stated that if the distribution is symmetrical then the median is a fair approximation for the mean and could be used in a meta-analysis. Furthermore, if the distribution is approximately normally distributed the width of the interquartile range will be approximately 1.35 standard deviations. However it should be noted that one of the reasons for reporting results as median and IQR is that the distribution is skewed and then the abovementioned approach may not be appropriate.

7.7.3.5  Medians and interquartile ranges

The median is very similar to the mean when the distribution of the data is symmetrical, and so occasionally can be used directly in meta-analyses.  However, means and medians can be very different from each other if the data are skewed, and medians are often reported because the data are skewed (see Chapter 9, Section 9.4.5.3).

Interquartile ranges describe where the central 50% of participants’ outcomes lie. When sample sizes are large and the distribution of the outcome is similar to the normal distribution, the width of the interquartile range will be approximately 1.35 standard deviations. In other situations, and especially when the outcomes distribution is skewed, it is not possible to estimate a standard deviation from an interquartile range. Note that the use of interquartile ranges rather than standard deviations can often be taken as an indicator that the outcomes distribution is skewed

Best Carsten

Hi Tung,

Any method may be making an unsuitable assumption. Since medians are used for skewed/kurtotic distributions, the nack-conversion may be misleading, particular for MA purposes. Consider converting your Means to Medians?

Hope this helps,

Matt

Dear Tung Nguyen Thanh

The Cochrane handbook stated that if the distribution is symmetrical then the median is a fair approximation for the mean and could be used in a meta-analysis. Furthermore, if the distribution is approximately normally distributed the width of the interquartile range will be approximately 1.35 standard deviations. However it should be noted that one of the reasons for reporting results as median and IQR is that the distribution is skewed and then the abovementioned approach may not be appropriate.

7.7.3.5  Medians and interquartile ranges

The median is very similar to the mean when the distribution of the data is symmetrical, and so occasionally can be used directly in meta-analyses.  However, means and medians can be very different from each other if the data are skewed, and medians are often reported because the data are skewed (see Chapter 9, Section 9.4.5.3).

Interquartile ranges describe where the central 50% of participants’ outcomes lie. When sample sizes are large and the distribution of the outcome is similar to the normal distribution, the width of the interquartile range will be approximately 1.35 standard deviations. In other situations, and especially when the outcomes distribution is skewed, it is not possible to estimate a standard deviation from an interquartile range. Note that the use of interquartile ranges rather than standard deviations can often be taken as an indicator that the outcomes distribution is skewed

Best Carsten

Hi,

The Cochrane Handbook for Systematic Reviews of Interventions mentions that "The median is very similar to the mean when the distribution of the data is symmetrical, and so occasionally can be used directly in meta-analyses.  However, means and medians can be very different from each other if the data are skewed, and medians are often reported because the data are skewed." (Link: https://handbook-5-1.cochrane.org/chapter_7/7_7_3_5_mediansand_interquartile_ranges.htm)

The median data are better to be avoided for meta-analysis and can be reported as additional tables:

• results that do not fit into ‘Data and analysis’ tables, for example skewed data reporting a median and range.

Use Wan's method (2014) ~820 cited. Superior to Hozo's method. Hereby the PDF and Excel.

thank you very helpful information

Hello Tung,

There seems to be a brand new technique (McGrath et al, 2020).

https://journals.sagepub.com/doi/full/10.1177/0962280219889080

But I do not have any Excel sheet.

Best wishes!