Some people prefer to think of SAT as OAT or TOAT (assuming that there is no instrument error involved), and TAT as indicated OAT allowing for compressibility problems at the thermometer.
TAT will exceed SAT by an amount depending on mach number and the characteristics of the transducer design, the latter determining how much of the temperature rise associated with the mach number is registered by the system.
Icing-wise it is all the same ... if the temperature is in the range determined during certification to present an icing risk, then you do the things appropriate to the aircraft to minimise the effects.