Assessment of depth of midazolam sedation using objective parameters

To determine the relationship between the clinically evaluated depth of midazolam-induced sedation and the cardiac beat-to-beat variability (RR variability) in ICU patients in the intensive care unit (ICU).

Prospective study.

ICU of a university hospital.

20 consecutive patients studied during weaning from mechanical ventilation and withdrawal of midazolam-induced sedation.

After clinical evaluation of depth of sedation according to the Ramsay sedation score, the RR variability over 512 RR intervals and predominant respiratory rate were measured. The power spectrum of RR variability was calculated by a fast Fourier transformation and the resulting total frequency band (0.016-0.35 Hz) was subdivided into a very low, a low, and a high frequency band. Stepwise multiple regression analysis in the first 10 patients (group 1) showed a significant relationship between depth of sedation and measures of RR variability combined with respiratory rate (r2 = 0.59; F = 12.1; p < 0.001). The more effective sedation was, the more depressed were both RR variability and predominant respiratory rate. Mean heart rate, mean respiratory rate, median deviation of RR intervals from the mean, and the ratio between spectral power density in the high and the low frequency bands proved to be the most important predictors of the Ramsay score (+/-1 level accuracy: 87%, p < 0.001). Using this regression equation, the Ramsay score was predicted in the remaining 10 patients (group 2) with a +/-1 level of accuracy of 81% (p < 0.001).

In ICU patients, a significant correlation is found between the depth of midazolam-induced sedation as assessed by the Ramsay sedation score and RR variability, with a clinically sufficient prediction accuracy. RR variability can serve as an objective, continuously available, and non-invasive measurement to monitor midazolam-induced sedation in intubated and mechanically ventilated patients.