The case of rapid response systems : Are randomized clinical trials the right methodology to evaluate systems of care ? *