Improving Bias Correction Standards by Quantifying its Effects on Treatment Outcomes