Problems of Statistical Evaluations

Why are there so many discrepancies between the figures of the different studies?
Why is it so difficult to produce reliable statistics in LCH?
How can the statistics be improved?

Statistics are always based on appropriate count, with both a clear definition and appropriate numerator and denominator.
Statistics in LCH are in general quite difficult for at least three major reasons:

  • The events are extremely rare and any study needs both a large geographic area and a large observation time.
  • The definition slightly changes with time.
  • The observator – who is counting events – can only access a biased denominator. (For example a national reference center can despite the quality of its observation loose many cases not referred to it).

Due to these limitations, results reported in studies can show large discrepancies for some indicators like the percentages of organs involved and of sequelae, or the incidence rate.