You have to consider the pressure gradient all the way down. You cannot just take the QNH and do corrections at MSL. Where the temperature is lower than ISA, for instance, the isobars lie closer together, giving a low pressure at the same true altitude - the opposite for warmer than ISA. ISA deviation varies as you climb, making it more difficult to calculate.
So if there is a QNH of around 1020, and you measure a true altitude of 38500', my guess is that the averge ISA deviation is below ISA. (had to correct this once, hope I got it right this time!)
Can some met expert please see if the above is right (easy to fog it up ...)
By the way, an ICAO eastern level in meteres is 11400m = FL374. You should be able to find a list of this in any Jeppesen IFR binder (1).
Previous discussion as to why not make all aircraft fly true altitude instead of pressure is not just because of conservatism. For performance reasons this would be not be practical, optimum and max flight level etc.