Apnea testing in suspected brain dead children--physiological and mathematical modelling.

OBJECTIVE

To study the validity and safety of the traditional apnea test in children, and to evaluate a mathematical equation estimating the hemodynamic response to the apnea test.

DESIGN

A prospective clinical study.

SETTING

Pediatric ICU.

PATIENTS AND PARTICIPANTS

38 pediatric patients suffering severe brain injury aged 2 months to 17 years, undergoing apnea testing for brain death.

MEASUREMENTS AND RESULTS

Apnea tests were performed 61 times (once in 19 patients, twice in 15, and 3 times in 4 patients). Mean PaCO2 was 41.1 +/- 10.6 mmHg before apnea and increased to 68.0 +/- 17.6 at 5 min. PaCO2 increased to 81.8 +/- 20.1 and 86.0 +/- 25.6 at 10 and 15 min, respectively. There was a mean PaCO2 increase by 5.38 +/- 1.4 mmHg/min in the first 5 min, and 2.75 +/- 0.5 mmHg/min during the next 5 min. We found a statistically significant (p < 0.05) linear relationship between the natural logarithm of PaCO2, time, and the logarithm of the initial level of PaCO2. An inverse linear relationship (p < 0.05) was found between systemic mean arterial pressure (MAP) and initial level of PaCO2 presented as mathematical correlations and nomograms.

CONCLUSIONS

By using our model for predicting MAP and PCO2 prior to apnea testing, hemodynamic embarrassment can be anticipated and prevented, thus allowing a safer procedure in the detection of brain death. Despite the fact that continuous cardiorespiratory monitoring is important, hemodynamic disturbances can be estimated before the apnea test, thus allowing a safer approach to brain death detection.