Cardiopulmonary arrest and mortality trends, and their association with rapid response system expansion

In reply: Levinson and Mills focus on one of the most important challenges in health care — the appropriate management of patients at the end of life. It is correct to suggest that the rapid response system-associated reduction in inhospital cardiopulmonary arrest may be due to more than just prevention through early intervention, but may also be due to increased attention to end-of-life care and avoidance of inappropriate resuscitation. Up to 30% of all rapid response calls are for patients with end-of-life issues. Perhaps if we develop ways of identifying these patients earlier, a more appropriate management plan could be developed in cooperation with patients and their carers.

O'Callaghan highlights the fact that the MERIT trial provided no evidence of significant improvement of cardiac arrest, unplanned intensive care unit admissions or unexpected death. However, as discussed in our MERIT publication1 in detail and in our Journal article2 to some degree, the MERIT trial was underpowered and the control hospitals acted in a manner similar to that of hospitals with medical emergency team (MET) systems in place (ie, over 35% of cardiac arrest team calls in control hospitals were made for patients without cardiac arrest). Also, the implementation and uptake of MET systems were not optimal (two-thirds of patients did not have a MET call despite meeting the criteria). Thus, the MERIT trial results are inconclusive but not in conflict with the results presented in our Journal article, as “absence of evidence is not evidence of absence”. Indeed, our follow-up article showed that MET hospitals had significant reduction in mortality outside the intensive care unit compared with non-MET hospitals.3

The increased hospital admissions over the study period were unlikely to explain the reduction in hospital mortality, as the increased admissions mostly occurred among frail older people, who have an increased risk of cardiac arrest and mortality. Our analyses adjusted for age and other possible confounders. The increase in admissions showed the increased severity and complexity of the conditions for which the patients were admitted, not a change in hospital health care delivery.