The purpose of this study is to examine the commonly held assumption that time is measured and documented accurately during resuscitation from cardiac arrest in the hospital.
Methods: A two-pronged approach was used to evaluate the accuracy of time documentation and measurement. First, two existing databases-the National Registry of Cardiopulmonary Resuscitation (NRCPR) and a 240-bed hospital's repository of cardiac arrest records-were evaluated for completeness and accuracy of documentation on resuscitation records of times required for calculating the Utstein gold-standard process intervals-recognition of pulselessness to starting cardiopulmonary resuscitation (CPR), delivery of first defibrillation shock, successful intubation, and epinephrine (adrenaline) administration. Second, nurses from a 900-bed hospital were interviewed to determine timepieces used during resuscitations, and timepieces were assessed for coherence and precision.
Results: : From the NRCPR database that included 10,689 pulseless cardiac arrests submitted by 176 hospitals, time data for calculating the Utstein intervals were missing for 10.9% of the interventions; negative intervals were calculated for 4%. From 232 consecutive resuscitation records from the 240-bed hospital, 85 records were identified from non-monitored units with staff who provided only CPR. Defibrillation, intubation and epinephrine administration were delayed until after arrival of advanced life support (ALS) responders; unlikely intervals of 0 min from event recognition to these ALS interventions were calculated for 11.5%. Sixty-seven nurses from the 900-bed hospital were interviewed; when documenting information during resuscitations, 21 (31.3%) reported using only patient room clocks, 30 (44.8%) only their watches, and 16 (23.9%) several timepieces. In all in-patient units in the same hospital, 241 timepieces (nurses' and physicians' watches, clocks in patient rooms, defibrillator clocks, central station monitors, and nursing station clocks) were compared to atomic time. The mean absolute time difference from atomic clock was 2.83 min (S.D. +/-5.9 min), median 1.88 min, and range 52.1 min slow to 72.85 min fast. There was no difference among timepieces (P = 0.35).
Conclusions: Missing time data, negative calculated Utstein gold-standard process intervals, unlikely intervals of 0 min from arrest recognition to ALS interventions in units with CPR providers only, use of multiple timepieces for recording time data during the same event, and wide variation in coherence and precision of timepieces bring into question the ability to use time intervals to evaluate resuscitation practice in the hospital. Practitioners, researchers and manufacturers of resuscitation equipment must come together to create a method to collect and document accurately essential resuscitation time elements. Our ability to enhance the resuscitation process and improve patient outcomes requires that this be done.