Accuracy of continuous jugular bulb oximetry in the intensive care unit

To address the accuracy of a bedside jugular bulb oxygen saturation (SjO2) catheter monitor (Baxter-Edwards, Santa Ana, CA) versus in vitro co-oximetry measurements in the intensive care unit (ICU).

By prospective protocol, we compared blood gas measurements with simultaneously recorded continuous bedside oximetric monitor values for 31 ICU patients with traumatic brain injury undergoing jugular bulb catheter monitoring. For suboptimal fiberoptic light signal quality indices, the catheter was repositioned, flushed, or both before drawing the sample for in vitro measurement. Laboratory and bedside monitor data were examined for association using the chi2 and paired t tests and a linear regression model.

We assessed 195 samples (median, 5 per patient; range, 1-14) who were monitored an average of 3.4 (range, 1-6) days. The in vivo monitor (range, 32-94%) and in vitro co-oximetry (range, 38-93%) values had acceptable correlation (y = 0.94x + 4.4, r2 = 0.80). For bedside monitor detection of jugular bulb desaturation (SjO2 < 50% for 10 min), the kappa statistic was 0.35, the sensitivity was 45 to 50%, and the specificity was 98 to 100%.

Continuous ICU SjO2 monitoring correlates significantly with in vitro values, but less so than previously described during intracranial surgery. Although sensitivity of the bedside monitor to detect confirmed desaturations remains an issue, the high specificity indicates that it is less of a concern that patients may be misdiagnosed as having desaturations resulting in unnecessary interventions. Nonetheless, suspected jugular bulb desaturation should be verified before taking therapeutic actions.