Accuracy of oxygen analyzers at subatmospheric concentrations used in treatment of hypoplastic left heart syndrome.

INTRODUCTION

The immediate survival of infants with hypoplastic left heart syndrome depends on success in achieving several therapeutic goals: (1) maintain patency of the ductus arteriosus, (2) assure adequate mixing of blood at the atrial level, and (3) establish and maintain a balance between systemic and pulmonary blood flow at or near unity. In accomplishing that final goal, various ventilatory strategies have been used to alter the physiologic modifiers of pulmonary vascular resistance and thus maintain balanced circulation, including ventilation with gas of subatmospheric oxygen concentration. However, no data on this subject have been published in the scientific literature, and commercial oxygen analyzers are specified for use within the range of 0.21 to 1.0 fraction of inspired oxygen (F(IO)(2)), leaving the accuracy of hypoxic gas delivery somewhat uncertain. We evaluated the performance of oxygen analyzers below F(IO)(2) 0.21.

METHODS

Two commercially available analyzers were studied: the TED-190 (Teledyne) and the Mini-OX III. Five new analyzers of each model were tested. After a 2-point calibration (F(IO)(2) 1.0 and 0.21), all 5 analyzers of the same model were simultaneously exposed to precision-blended gases at 6 different concentrations of oxygen in nitrogen. Steady state was maintained for at least 2 min at each concentration before readings were recorded. Calibration was verified at F(IO)(2) 0.21 between each level.

RESULTS

The mean +/- SD error was 0.0013 +/- 0.0021 for the Mini-OX III analyzers and -0.0004 +/- 0.0009 for the Teledyne analyzers. The upper and lower limits of the 95% confidence interval were 0.39% and -0.13% for the Mini-OX III analyzers and 0.07% and -0.15% for the Teledyne analyzers. The maximum difference between measured and known oxygen concentrations was 1% of full scale.

CONCLUSIONS

The Mini-OX III and the Teledyne TED-190 provide accurate and reliable F(IO)(2) readings between 0 and 0.21 that are within the manufacturers' specifications for maximum error. These 2 analyzers are therefore acceptable for use in delivering subambient oxygen concentrations. The Mini-OX III displays oxygen concentration to the nearest 0.1% and may be more appropriate for precise control.