Procedure-specific do-not-resuscitate orders. Effect on communication of treatment limitations

Do-not-resuscitate (DNR) orders are often inaccurately communicated between physicians and nurses or residents. Structured, procedure-specific DNR order forms have been suggested to improve communication, but no data exist to support this impression.

The level of agreement between attending physicians and nurses or residents in their understanding of the DNR orders of critically ill patients was measured before and after instituting a structured DNR order form. Caregivers were asked (1) about the clinical events to which the DNR order applied, (2) whether the DNR order withheld all elements of cardiopulmonary resuscitation, and (3) whether other treatments were to be withheld. Results were reported as kappa +/- SE.

Nurses (n=41) and residents (n=34) showed only fair to moderate agreement with attending physicians (n=53) for the 76 evaluable patients before initiation of the DNR order form. After initiation of the structured DNR order form, nurses showed higher levels of agreement for the second (0.67 +/- 0.14) and third (0.69 +/- 0.13) components but not the first (0.39 +/- 0.15) component of the DNR order. Residents showed higher levels of agreement for the second (0.90 +/- 0.10) and third components (0.81 +/- 0.13) but not the first (0.57 +/- 0.17) component. Nurses compared with residents had lower levels of agreement with attending physicians for most aspects of the DNR order.

A structured DNR order form improves agreement in understanding of some but not all components of the DNR order.