The oxygen reserve index (ORi) in dogs: validity, reliability, and clinical applicability for estimating hyperoxemia

The oxygen reserve index (ORi) is a non-invasive, continuous parameter measured by multi-wavelength pulse CO-oximetry. Developed in human medicine, it estimates oxygen reserve, defined as arterial partial pressure of oxygen (PaO₂) between 100 and 200 mmHg, with values ranging from 0 to 1. Unlike peripheral oxygen saturation (SpO₂), which plateaus above 100 mmHg and cannot reflect further increases in oxygenation, ORi provides additional information within the hyperoxic range. Despite its potential, ORi has not been extensively validated in veterinary species. This thesis presents findings from five studies evaluating the validity, reliability, and clinical applicability of ORi in dogs. The first prospective study, involving 37 mechanically ventilated dogs, demonstrated a moderate correlation (r² = 0.52) between ORi and PaO₂. ORi showed high sensitivity and specificity in detecting hyperoxemia at thresholds of PaO₂ ≥150 and ≥190 mmHg and was capable of tracking proportional PaO₂ changes. The second study investigated ORi’s utility as an early indicator of oxygen desaturation in 37 apneic dogs. ORi decreased before SpO2 changes. The median time from ORi of 0.5 to SpO2 of 95% was 30–35 sec, and ORi declined faster in obese dogs, indicating that body condition affects oxygen reserve depletion. The third study assessed the relationship between the ORi to inspired fraction of oxygen (FiO₂) ratio and venous admixture (F-shunt) in 44 dogs. A moderate negative correlation (r = −0.59) was observed, and a threshold ORi/FiO₂ ratio of 1.2 was optimal for detecting low shunt levels (≤10%) with good sensitivity, though performance declined for higher shunt fractions. The fourth study explored the influence of the perfusion index (PI) on ORi’s ability to estimate hyperoxemia. It found that PI values above 2 reduced the correlation between ORi and PaO₂, likely due to fluctuations in blood flow perfusing the measurement site. The fifth study compared alternative sensor placements (foot and tail) to the traditional tongue site. While ORi values were strongly correlated (r > 0.89), trend tracking was inconsistent (trending ability: 58.3– 73.9%), and skin pigmentation sometimes interfered with signal quality. Overall, these five studies support the validity and reliability of ORi in dogs. Although ORi does not replace direct PaO₂ measurement, it offers a valuable, non-invasive means of continuously monitoring oxygen reserves in dogs. The findings also highlight important technical and physiological factors that must be considered when interpreting ORi values in clinical practice.