THE GUIDELINES FOR CARDIOPULMONARY RESUSCITAtion (CPR) and Emergency Cardiovascular Care (ECC) are probably the most widely implemented and best-known guidelines in medicine. In the setting of cardiac arrest, health care professionals want and need simple, practical, and effective guidelines. As the American Heart Association (AHA) and International Liaison Committee on Resuscitation (ILCOR) revise their Consensus on Science and Treatment Guidelines in 2005, it is imperative to assess how these guidelines are developed. Despite the major reassessment and publication of new CPR and ECC guidelines every 5 to 8 years for the past 3 decades, survival from cardiac arrest remains dismal. Have the guidelines and guideline development process improved or compromised the treatment of patients in cardiac arrest? Do they reflect the reality of cardiac arrest treatment? Are they responsive, or impenetrable, to new ideas and concepts in ECC? Are there ways to improve the guidelines process and, therefore, the guidelines themselves? The studies by Wik et al and Abella et al in this issue of JAMA document a major problem in the treatment of patients in cardiac arrest. Using a sternal pad to monitor chest compressions and ventilations, Wik et al obtained data from paramedics and nurse anesthetists performing CPR on 176 adult patients with out-of-hospital cardiac arrest in 3 cities in Europe. They found that chest compressions were not provided 48% of the time with patients in cardiac arrest. Despite using a compression rate of 121/min, these rescue personnel, with the documented interruptions, delivered only 64 chest compressions per minute. Abella et al, using the same monitoring device to observe 67 in-hospital cardiac arrests, found that patients did not receive chest compressions 24% of the time during the resuscitation. Other problems identified in the CPR segments analyzed included ventilations of 20/min or more (61%), compression rates less than 90/min (28%), and inadequate compression depth (37%). Although neither of these studies was powered to assess patient survival, Abella et al found a trend showing that patients who had longer periods without chest compression had worse resuscitation outcome. These reports are consistent with previous studies documenting low chest compression rates and high ventilation rates when CPR is performed by health care professionals. They also complement studies looking at how laypersons and health professionals deliver CPR in cardiac arrest simulations. Assar et al demonstrated that laypersons taught single-rescuer CPR take an average of 16 seconds for each ventilatory pause. A recent study from our CPR Research Group showed that medical students needed 14 seconds to deliver 2 breaths during CPR and delivered only 43 chest compressions per minute after AHA standard CPR training because of pauses for ventilations. Students taught a simplified chest-compression-only CPR delivered 113 compressions per minute immediately after training and 91 compressions per minute when tested 6 months later. Thus, laypersons, medics, physicians, nurses, medical students, and other health care professionals do not perform CPR according to published guidelines. However, this conclusion is not surprising. Indeed, studies demonstrating poor retention of CPR skills have documented the poor performance of CPR for more than 3 decades. Does the quality of CPR make a difference in patient outcomes? Although there are no randomized controlled trials (RCTs) to answer this question, observational studies in both experimental models and humans indicate that the quality of CPR is likely to affect patient outcome. Kern et al demonstrated that when animals received realistic 16-second pauses for ventilations, 24-hour neurologically intact survival was 13% compared with 80% in the group receiving continuous chest compressions. Yu et al showed that 100% of animals receiving more than 80 compressions per minute were resuscitated whereas only 10% of those receiving fewer than 80 compressions per minute survived.