Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Discontinuing in-hospital cardiac resuscitation - validating a clinical decision aid

Decision aid

The evidence rules for diagnostic tests or clinical decision rules are not the same as for treatments. Randomised trials have their place, but the hard work comes from:

When the rules have been shown to work, only then may a randomised trial make sense.

From Canada comes a superb example of a decision aid for perhaps one of the most difficult clinical decisions, discontinuing resuscitation attempts after a cardiac arrest [1].

Decision aid

The decision aid had previously been determined using data from two randomised trials [2]. It had been found that all resuscitated patients who were eventually discharged from hospital had:

♦ A witnessed arrest
♦ An initial cardiac rhythm of ventricular tachycardia or ventricular fibrillation
♦ A pulse within the first 10 minutes of chest compression

The aid proposed that physicians could safely withdraw resuscitative efforts on patients who did not satisfy the decision aid since none were discharged from the hospital.


The validation of the decision aid was carried out by analysis of a resuscitation registry at a medical centre in Georgia in which all in-hospital arrests between 1987 and 1996 were entered into a registry, and in which resuscitation was handled by multidisciplinary teams according to standard protocols. A detailed coding sheet accompanied each resuscitation attempt, and hospital records were reviewed to ensure that all resuscitation events were recorded.


After excluding resuscitation events for a variety of reasons (no chest compression, information missing) there were 2181 attempted resuscitations on 1884 patients. They had an average age of 65 years, and half were women.

Table 1 shows the results of applying the decision aid. There were 1912 patients in which the decision aid predicted some chance of a hospital discharge, and 17% of these were eventually discharged. There were 269 in which the decision aid predicted no chance of discharge. Fifty-three responded enough to be transferred to intensive care, and 26 remained alive for at least 24 hours (range 1 - 29 days). Three, (1%) patients were discharged.

Table 1: Results of applying the decision aid to 2181 attempted resuscitations

Decision aid      
The patient has a chance of discharge from hospital if any of the following is true: Applying the decision aid to 2181 attempted resuscitations
The arrest was witnessed 1 1721 arrest witnessed 287 (17%) discharged Some chance of discharge
    460 arrest NOT witnessed 40 (9%) discharged OF WHOM
The initial cardiac rhythm was ventricular tachycardia or ventricular fibrillation 2 49 initial rhythm VT or VF 10 (20%) discharged Some chance of discharge
    411 initial rhythm NOT VT or VF 30 (7%) discharged OF WHOM
Pulse was regained during the first 10 minutes of chest compression 3 142 pulse regained within 10 minutes 27 (19%) discharged Some chance of discharge
    269 pulse NOT regained within 10 minutes 3 (1%) discharged Decision aid predicts no chance of discharge

The three discharged patients were:

  1. 1 A 76 year old man with dementia, hypertension and COPD and oropharyngeal cancer who was eventually transferred to another hospital, required major medical aid, and who died two months later.
  2. A 43 year old man with COPD and alcoholic cardiomyopathy. He was discharged to a nursing home because of problems with caring for himself despite minimal ischaemic damage.
  3. A 65 year old woman who arrested following back surgery, with no ischaemic injury but who was discharged to a nursing home because of complications with surgery.


This decision aid was validated in a separate group of patients to that in which it was developed. Experts in resuscitation after cardiac arrest will have their own views about the applicability of decision aids in this most difficult situation. They may also question whether the situation in hospitals in North America, with high staffing levels, makes the decision aid valid in situations with lower staffing levels, where the actual witnessing of an event might be much less likely, and the decision aid less useful.

The strength of the study is that it demonstrates beautifully that decision aids can be derived for even difficult clinical situations, and that the methods required produce robust results when applied. For likelihood ratio aficionados, the positive likelihood ratio was 1.2 (not very useful), but the negative likelihood ratio was 0.06 (making a rule-out rule possible).


  1. C van Walgrave et al. Validation of a clinical decision aid to discontinue in-hospital cardiac arrest resuscitation. JAMA 2001 285: 1602-1606.
  2. C van Walgrave et al. Derivation of a clinical decision rule for the discontinuation in-hospital cardiac arrest resuscitation. Archives of Internal Medicine 1998 158: 129-134.
previous or next story in this issue