Better time data from in-hospital resuscitations

Benefits of an undocumented defibrillator feature


Research and quality improvement (QI) related to in-hospital cardiopulmonary resuscitation attempts (“codes” from here forward) are hampered significantly by the poor quality of data on time intervals from arrest onset to clinical interventions.1

John A. Stewart, a hospital nurse in Seattle

John A. Stewart

In 2000, the American Heart Association’s (AHA) Emergency Cardiac Care Guidelines said that current data were inaccurate and that greater accuracy was “the key to future high-quality research”2 – but since then, the general situation has not improved: Time intervals reported by the national AHA-supported registry Get With the Guidelines–Resuscitation (GWTG-R, 200+ hospitals enrolled) include a figure from all hospitals for times to first defibrillation of 1 minute median and 0 minutes first interquartile.3 Such numbers are typical – when they are tracked at all – but they strain credulity, and prima facie evidence is available at most clinical simulation centers simply by timing simulated defibrillation attempts under realistic conditions, as in “mock codes.”4,5

Taking artificially short time-interval data from GWTG-R or other sources at face value can hide serious delays in response to in-hospital arrests. It can also lead to flawed studies and highly questionable conclusions.6

The key to accuracy of critical time intervals – the intervals from arrest to key interventions – is an accurate time of arrest.7 Codes are typically recorded in handwritten form, though they may later be transcribed or scanned into electronic records. The “start” of the code for unmonitored arrests and most monitored arrests is typically taken to be the time that a human bedside recorder, arriving at an unknown interval after the arrest, writes down the first intervention. Researchers acknowledged the problem of artificially short time intervals in 2005, but they did not propose a remedy.1 Since then, the problem of in-hospital resuscitation delays has received little to no attention in the professional literature.

Description of feature

To get better time data from unmonitored resuscitation attempts, it is necessary to use a “surrogate marker” – a stand-in or substitute event – for the time of arrest. This event should occur reliably for each code, and as near as possible to the actual time of arrest. The main early events in a code are starting basic CPR, paging the code, and moving the defibrillator (usually on a code cart) to the scene. Ideally these events occur almost simultaneously, but that is not consistently achieved.

There are significant problems with use of the first two events as surrogate markers: the time of starting CPR cannot be determined accurately, and paging the code is dependent on several intermediate steps that lead to inaccuracy. Furthermore, the times of both markers are recorded using clocks that are typically not synchronized with the clock used for recording the code (defibrillator clock or the human recorder’s timepiece). Reconciliation of these times with the code record, while not particularly difficult,8 is rarely if ever done.

Defibrillator Power On is recorded on the defibrillator timeline and thus does not need to be reconciled with the defibrillator clock, but it is not suitable as a surrogate marker because this time is highly variable: It often does not occur until the time that monitoring pads are placed. Moving the code cart to the scene, which must occur early in the code, is a much more valid surrogate marker, with the added benefit that it can be marked on the defibrillator timeline.

The undocumented feature described here provides that marker. This feature has been a part of the LIFEPAK 20/20e’s design since it was launched in 2002, but it has not been publicized until now and is not documented in the user manual.

Hospital defibrillators are connected to alternating-current (AC) power when not in use. When the defibrillator is moved to the scene of the code, it is obviously necessary to disconnect the defibrillator from the wall outlet, at which time “AC Power Loss” is recorded on the event record generated by the LIFEPAK 20/20e defibrillators. The defibrillator may be powered on up to 10 minutes later while retaining the AC Power Loss marker in the event record. This surrogate marker for the start time will be on the same timeline as other events recorded by the defibrillator, including times of first monitoring and shocks.

Once the event record is acquired, determining time intervals is accomplished by subtracting clock times (see example, Figure 1).

Figure 1. Sample event record

In the example, using AC Power Loss as the start time, time intervals from arrest to first monitoring (Initial Rhythm on the Event Record) and first shock were 3:12 (07:16:34 minus 07:13:22) and 8:42 (07:22:14 minus 07:13:22). Note that if Power On were used as the surrogate time of arrest in the example, the calculated intervals would be artificially shorter, by 2 min 12 sec.

Using this undocumented feature, any facility using LIFEPAK 20/20e defibrillators can easily measure critical time intervals during resuscitation attempts with much greater accuracy, including times to first monitoring and first defibrillation. Each defibrillator stores code summaries sufficient for dozens of events and accessing past data is simple. Analysis of the data can provide a much-improved measure of the facility’s speed of response as a baseline for QI.

If desired, the time-interval data thus obtained can also be integrated with the handwritten record. The usual handwritten code sheet records times only in whole minutes, but with one of the more accurate intervals from the defibrillator – to first monitoring or first defibrillation – an adjusted time of arrest can be added to any code record to get other intervals that better approximate real-world response times.9

Figure 2. Approximate time intervals

Next Article:

   Comments ()