Likelihood of successful extubation in critically ill patients is estimated by respiratory parameters, the predicitive power of which is only modest. Although the sensitivity of a rapid shallow breathing index (RR/TV) is 100%, its specificity is only 63%. We hypothesized that brain natriuretic peptide (BNP), a marker of pressure or volume overload of the heart, could improve this prediction in the cardiac population.
Baseline BNP (Triage BNP, Biosite) was measured in 23 consecutive CCU patients who had reached the point of attempted extubation according to usual clinical criteria.
Extubation failed in two patients (F). The 21 patients with successful extubation (S) had age 69 ± 4 years, mean arterial pressure 74 ± 3 mmHg, hemoglobin 10.4 ± 0.3 g/l, creatinine 1.3 ± .14 mg/dl, period of intubation 5.5 ± 1 days, ejection fraction 32 ± 3% and RR/TV 35 ± 6.1 breaths*min-1. Baseline BNP was 1212 ± 254 pg/ml and its increase in response to a 60 minute spontaneous breathing trial (SBT) was 308±184 pg/ml. In the failure group (n=2) none of the individual parameters were outside the 95% confidence intervals.
In summary, the preliminary data of this ongoing trial suggest that BNP is not predictive of which patients will fail extubation.
Our data suggests that BNP measurement is not useful in determining a patient’s likelihood for successful extubation.
D. Hersh, None.