Novel method to quantify loss of heart rate variability in pediatric multiple organ failure.

OBJECTIVE

To develop a power-law model for measurement of heart rate variability (HRV) and to compare this model with established methods for measuring HRV in a group of children with organ failure (OF).

DESIGN

Prospective, observational study.

SETTING

Pediatric intensive care unit of a tertiary children's hospital.

PATIENTS

A total of 104 measurements were made on 50 patients (median age, 8 months; range, 2 days to 16 yrs) and categorized into three groups according to the number of simultaneous organs failing: 0-1 OF, 2 OF, and >/=3 OF.

INTERVENTIONS

Heart rate was recorded over a 5-min period when patients were hemodynamically stable. The power-law model represents a power function relating frequency distribution to magnitude of effect (in this case, squared deviation from the mean heart rate). Plotting the data on a bi-logarithmic scale produces a regression line for each measurement, described in terms of r2, slope, and x-intercept. Comparison with other HRV measures included two time-domain measures (sd of the normal R-R intervals and the square root of the mean squared differences of successive normal R-R intervals), one frequency-domain method (power spectral analysis), and one nonlinear method (detrended fluctuation analysis).

MEASUREMENTS AND RESULTS

For the power-law model, patients exhibited a similar r2 of.87 (.09) (mean [sd]) and slope of -1.80 (0.29), regardless of the degree of OF. HRV could thus be described purely in terms of x-intercept, which demonstrated a left shift with increasing OF (p <.001). This was independent of age and heart rate. Loss of HRV with increasing OF was demonstrated by all methods; however, only the power-law model was able to discriminate between each OF group. Using the model, change in HRV in individual patients over successive days often concurred qualitatively with the change in OF status.

CONCLUSION

The power-law model is an appropriate measure of HRV in pediatric patients, being neither age nor heart rate sensitive. Loss of HRV occurs with increasing OF; this effect was better demonstrated by the model compared with other measures of HRV.