One of two approaches to reporting defibrillation results involves what has been termed the defibrillation threshold. While the term defibrillation threshold has often been used in a vague fashion to denote the minimum level of the relevant independent variable at which defibrillation begins to be observed, it has sometimes been more carefully specified in terms of one of a number of definitive algorithms: an example, in one of the more intuitively attractive algorithms, one induces fibrillation and then attempts to defibrillate with a specific waveform of given amplitude. If the initial episode is successful, it is successively repeated with 20 percent less current each time, until failure is observed. If the initial episode is unsuccessful, it is successively repeated, with 20 percent more current each time, until success is achieved. In either case, the threshold is defined as the average of the last two shocks. A nice feature of this and other simple ways of characterizing the level of a variable needed for defibrillation is that very few fibrillation-defibrillation episodes are ordinarily required. Thus, the approach is quite inexpensive in terms of investigator time and animal usage. Recent studies have shown, however, that, except when there is a rapid transition from ineffective to effective shocks with an increase in the variable, the derived threshold may assume a wide range of values and yield little insight into the percent success associated with the shock.
The aim of the work proposed here is to further evaluate, using exact mathematical methods, the limitations of various methods currently used in specifying defibrillation threshold and then to employ the same methods (and/or more elaborate versions thereof) in checking somewhat more involved algorithms of a category which might be simple and straightforward enough for investigators to employ in defibrillation studies while at the same time capable of yielding credible results. The long-term benefits of this work are (a) the possibility of responding to societal concerns by achieving an appreciable decrease in animal usage while still obtaining information which will be helpful in the design of defibrillators for clinical use (or alternatively, showing that threshold methods are really not attractive in comparison with the more detailed methods involving curves of percent success versus a relevant variable) and, (b) eventually minimizing the negative effects of the amount of misinformation which is currently in the literature as a consequence of the improper use of the defibrillation threshold concept.
| McDaniel, W C; Schuder, J C (1987) The cardiac ventricular defibrillation threshold: inherent limitations in its application and interpretation. Med Instrum 21:170-6 |
