Alternative method for evaluating interventions where randomisation may not be possible.
The authors discuss propensity score matching, a technique for evaluating causal inference (which can be applied to intervention evaluation) when randomisation is not possible, perhaps for practical or ethical reasons.