Survey of "do not resuscitate" orders in a district general hospital.

OBJECTIVE: To evaluate the local use of written "Do not resuscitate" orders to designate inpatients unsuitable for cardiopulmonary resuscitation in the event of cardiac arrest.

DESIGN: Point prevalence questionnaire survey of inpatients' medical and nursing records.

SETTING: 10 acute medical and six acute surgical wards of a district general hospital.

PARTICIPANTS: Questionnaires were filled in anonymously by nurses and doctors working on the wards surveyed.

MAIN OUTCOME MEASURES: Responses to questionnaire items concerning details about each patient, written orders not to resuscitate in the medical case notes and nursing records, whether prognosis had been discussed with patients' relatives, whether a "crash call" was perceived as appropriate for each patient, and whether the "crash team" would be called in the event of arrest.

RESULTS: Information was obtained on 297 (93.7%) of 317 eligible patients. Prognosis had been discussed with the relatives of 32 of 88 patients perceived by doctors as unsuitable for resuscitation. Of these 88 patients, 24 had orders not to resuscitate in their medical notes, and only eight of these had similar orders in their nursing notes.

CONCLUSIONS: In the absence of guidelines on decisions about resuscitation, orders not to resuscitate are rarely included in the notes of patients for whom cardiopulmonary resuscitation is thought to be inappropriate. Elective decisions not to resuscitate are not effectively communicated to nurses. There should be more discussion of patients' suitability for resuscitation between doctors, nurses, patients, and patients' relatives. Suitability for resuscitation should be reviewed on every consultant ward round.