Reconsideration of proximate Utstein-style end points.

Abstract

The Utstein-style template defines core and supplementary data for reporting out-of-hospital cardiac arrest information. The primary outcome statistic of the Utstein template is survival to hospital discharge (SHD). The SHD statistic is dependent on Utstein-defined out-of-hospital variables and multiple in-hospital variables that are undefined and uncontrolled. An example of one of these undefined in-hospital variables is the decision to place a patient on do-not-resuscitate status. At our municipal teaching hospital, 418 patients who had out-of-hospital cardiac arrest presented over a 4-yr period; 79 (19%; 95% confidence interval [CI], 15% to 23%) survived to hospital admission, with 54 (68%; 95% CI, 57% to 78%) subsequently being placed on do-not-resuscitate status. When patients on do-not-resuscitate status were included in the SHD calculation, the SHD rate was 5.3% (95% CI, 3.3% to 7.8%), and when patients on do-not-resuscitate status were excluded from the SHD calculation, the SHD rate was 6.1% (95% CI, 3.8% to 9.0%). These data show a relative 15% change in SHD resulting from a single in-hospital variable. Cardiac arrest survivors represent a small proportion of a total population; therefore, large numbers of study subjects are required for a statistically significant interpretation of the SHD statistic. This requirement for large study populations has resulted in recent studies that report results by using end points proximate to SHD when assessing the effect of individual interventions. It is logical that success of a specific intervention should be determined by the ability of the intervention to accomplish its purpose rather than the ability to improve SHD that is dependent on multiple variables. Furthermore, because in-hospital care is not standardized and uncontrolled variables exist, the primary Utstein end point of SHD should be reconsidered when evaluating cardiac arrest interventions.