The Relationship between Conjunctival PO 2 and Arterial PO 2 in 16 Normal Persons

Abstract

Semi-invasive transconjunctival oxygen monitoring uses a miniaturized Clark electrode to measure the oxygen tension of the palpebral conjunctiva (PcjO₂). We designed this study to determine the normal value for the conjunctival index (PcjO₂/PaO₂), to evaluate the correlation between PcjO₂ and PaO₂ over a wide PaO₂ range, to evaluate the short-term stability of the conjunctival index with changes in PaO₂, and to evaluate some of the practical aspects of conjunctival oxygen monitoring. Materials and Methods: The Insight Eyelid Oxygen Monitor was used. We measured PcjO₂, pH, PaO₂, and PaCO₂ and calculated the conjunctival index in 16 healthy volunteers under conditions of normoxemia, hypoxemia, and hyperoxemia. In addition, subjects used a questionnaire to report their experience with the eyelid sensor. Results: The mean PcjO₂ (66 ± SD 32 torr, range 14-157 torr) was significantly different from the mean PaO₂ (99 ± 47 torr, range 34-221 torr) (P < 0.001). The mean conjunctival index was 0.68 ± 0.15 (range 0.41-0.98). The correlation coefficient for the overall relationship between PcjO₂ and PaO₂ was 0.91. When the conjunctival index was determined on the same individual at two PaO₂ levels, there was a statistically significant mean difference of 0.16 in the two conjunctival index values (P < 0.002). Our staff learned to use the conjunctival oxygen monitoring system with a minimum of practice. All subjects found the placement of the sensor at least mildly uncomfortable, six subjects stated that they would prefer not to have conjunctival oxygen monitoring used on them if they were critically ill, five subjects found the eyelid sensor more uncomfortable than arterial puncture, one subject experienced severe irritation and pus formation, and two experienced mild itching and irritation. Conclusions: Conjunctival oxygen monitoring is a relatively simple but not entirely innocuous means of monitoring tissue oxygenation. Although the overall correlation between PcjO₂ and PaO₂ is good in hemodynamically normal persons, the inter- and intra-individual variability in the conjunctival index limits its usefulness as an accurate reflection of PaO₂. When PcjO₂ monitoring is used, we recommend that it be used in conjunction with periodic measurement of PaO₂ to assess overall tissue oxygen delivery.