Consider adding survival within matched diagnostics

[edward-burn]

In addition to comparing characteristics for sample vs matched populations, it seems like it could be informative to also compare survival estimates for sample vs matched - what do you think @albertpratsu @daniprietoalhambra ?

[daniprietoalhambra]

For data with good mortality like cprd even having survival in the sample would be awesome!