If it's not very large file, I think the simpliest way is to put the whole file into an string as explained here.
Once done, split and regex will do the rest of the work:

use strict;
use warnings;

my $text = <<EOF;
=================================================
Record 0

Name                  : John [John]
I/C                   : >123< [123]
Sex                   : 'Male' [Male]
Ref                   : >12< [12]
==================================================
Record 1

Name                  : Pretty [Pretty]
I/C                   : >125< [125]
Location              : 'Female' [Female]
Ref                   : >12< [12]
==================================================
Record 2

Name                  : Kim [Kim]
I/C                   : >124< [124]
Location              : 'Male' [Male]
Ref                   : >11< [11]
==================================================
EOF

my @goods = grep /Ref\s+:\s+>12</, split /=+/, $text;

# For this example 

print join "\n==================================================\n",  
+@goods;
[download]

Hope it helps,
deibyz