Determining occurrence of a sequence of numbers in the first 50,000 primes

Prime and IntegerDigits are Listable, ___ (i.e. BlankNullSequence) is a pattern object that can stand for any sequence of zero or more expressions.

This is a nice approach and quite fast:

Length @ Cases[ IntegerDigits @ Prime @ Range @ 50000,
                          {___, 5, ___, 4, ___, 3, ___}]

1588

Remarks

This metod and kglr's operator approach based on Count seem to be equally efficient. However a reliable estimation of efficiency in case of large sets of primes should be carefully compared (with refreshing the kernels) since the algorithms behind Prime use sparse caching and sieving, to gain an insight I recommend to think over this question What is so special about Prime?. Instead of Length one might exploit CountDistinct in case of duplicates presence.

out = Select[MatchQ[IntegerDigits[#], {___, 5, ___, 4, ___, 3, ___}] &]@
   Prime[Range[50000]];

Length@out

1588

Short@out

{1543, 2543, 5413, 5431, 5437, 5443, 5483, 5743, 5843, 8543, <<1568>>, 599243, 599413, 599843, 601543, 602543, 605413, 605443, 605543, 610543, 611543}

If you need just the count:

count = Count[{___, 5, ___, 4, ___, 3, ___}]@*IntegerDigits@*Prime@*Range;
count @ 50000

1588

Prime[Range[50000]] // Short
IntegerDigits /@ % // Short
Flatten[SequenceCases[#, {___, 5, ___, 4, ___, 3, ___}] & /@ %, 1] // Length

(* 1588 *)

If you want the actual primes, replace the last line by

FromDigits /@ Flatten[SequenceCases[#, {___, 5, ___, 4, ___, 3, ___}] & /@ %, 1]
(* {1543, 2543, 5413, 5431, 5437, 5443, 5483, 5743, 5843, ..., 602543, 605413, 605443, 605543, 610543, 611543} *)

This should be speedier...

Pick[p = Prime@Range@5*^4, StringMatchQ[IntegerString[p], "*5*4*3*"]]