This specificity issue is bothering me. This 'medical' meaning 'very important' - so if there's a test with 90% accuracy, why on heaven's earth do we not simpl

This specificity issue is bothering me. This 'medical' meaning 'very important' - so if there's a test with 90% accuracy, why on heaven's earth do we not simply run the test 3 times to get 'considerably greater accuracy'? 10% error is so large it's tough to make heads or tails of the data? Does someone know if this makes sense i.e. if we can simply run the test 3 times and get better data. Secondly - why are these health authorities not doing proper, state-wide tests? Here we are with an economy in meltdown, trying to 'model model model' with a 10 Trillion dollar economy wouldn't it make sense for at least ONE (or a few) freaking comprehensive tests that give us some good data?