Original Article

Evaluation of Oxygenation in Low- and High-Flow Anesthesia Applications by Oxygen Reserve Index: A Randomized Prospective Study


  • Huseyin Oztoprak
  • Gamze Kucukosman
  • Bengu Gulhan Aydin
  • Rahsan Dilek Okyay
  • Ozcan Piskin
  • Hilal Ayoglu

Received Date: 07.09.2021 Accepted Date: 17.01.2022 Med Bull Haseki 2022;60(2):104-112


While arterial blood gas (ABG) analysis is invasive, intermittent, and costly, the oxygen reserve index (ORI) is a new method that can be non-invasive and continuous measurement aimed at providing information about the patient’s O2 status in the moderately hyperoxic range. In our study, the ORI to PaO2 relationship in different fresh gas flows was evaluated.


This randomized prospective study was conducted between November 2018 and November 2019. All patients were ventilated for the first 10 min after intubation with 50% O2/air and 6 L/min fresh gas flow. Then, the flow rate was randomly set to 4 L/min for high-flow anesthesia (group H) or 1 L/min for low-flow anesthesia (group L). ABG’s were taken before preoxygenation, intraoperative 60th min, and at the end of surgery, and simultaneous ORI and SpO2 were recorded.


The study was completed with 70 patients. Mean PaO2 values were higher in group H, apart from before preoxygenation (p<0.05). Mean ORI values differed between groups except before preoxygenation and the intraoperative 10th min (p<0.05). A statistically significant, positive and weak correlation was identified between ORI and PaO2. According to the regression analysis, the ORI value was approximately 0.2 when the PaO2 value was ≥100 mmHg at the intraoperative 60th min and at the end of the surgery, and 0.3 when the PaO2 was ≥150 mmHg.


ORI may be an alternative to PaO2 in monitoring the oxygen status of intraoperative patients.

Keywords: General anesthesia, hyperoxia, low flow anesthesia, oxygen reserve index


Low-flow anesthesia is based on the principle of returning at least 50% of the exhaled gases to the patient via a breathing system after the elimination of carbon dioxide in the anesthesia circuit (1). In low-flow anesthesia applications, the difference between the amount of oxygen (O2) and the O2 centration in the gas composition provided to the patient increases as the fresh gas flow decreases. Because the oxygen-depleted gas mixture takes up a large amount of space in the rebreathing volume, the O2 concentration delivered to the patient may drop significantly, increasing the risk of hypoxia (1,2).

Peripheral O2 saturation (SpO2) and arterial partial oxygen pressure (PaO2) are monitored to detect hypoxia during anesthesia applications. Oxygen reserve index (ORI) is a measurement technique derived from hemoglobin sensors, which shows the O2 reserve in arterial blood and can instantly evaluate tissue oxygenation. It is an index with a unitless scale between 0 and 1, which is a relative indicator of changes in PaO2, especially in the mild hyperoxic range (PaO2 100-200 mmHg) (3). The ORI reportedly gives an early warning during a possible oxygenation impairment-before any change in SpO2 occurs-and shows the response to the oxygen administration (4,5). Recently, it has been reported that ORI can be used to prevent hyperoxia during general anesthesia and after surgery (6,7). This study evaluates the ORI-PaO2 relationship in different fresh gas flows.


Compliance with Ethical Standards  and Study Design

The study was conducted at the Zonguldak Bulent Ecevit University between November 2018 and November 2019, with the approval of the Zonguldak Bulent Ecevit University Clinical Research Ethics Committee (date: 07.11.2018, protocol no: 2018-221-07/11) and the written consent of the patients. The trial was registered before subject enrollment at ClinicalTrials.gov (Ref: NCT04698863; principal investigator: G.K.; date of registration: January 6, 2021). The flow diagram of the study according to the Consolidated Standards of Reporting Trials (CONSORT) 2010 is presented in Figure 1 (8).

Patient Population

In this prospective study, randomization was performed using the sealed-envelope method. Patients between the ages of 18 and 65 who were in the ASA I-II risk group and who were scheduled for a tympanomastoidectomy under elective conditions, with a minimum operation time of 1.5 h, were included in the study. Patients with morbid obesity, a history of malignant hyperthermia, opioid sensitivity, alcohol or drug abuse problems, congestive heart failure, coronary artery disease, significant anemia, liver and kidney disease, pregnant or lactating women, and patients allergic to the drugs used in the study were excluded.

Before each patient, the seal of the anesthesia circuits was checked, the gas monitors were calibrated, and the lower limit of the inspired O2 fraction (FiO2) was set to 30%. The carbon dioxide (CO2) absorbent was replaced after each patient.

Oxygen Reserve Index and Arterial Blood Gas Measurements

Routine hemodynamic monitoring was performed on patients in the operating room. Additionally, the ORI sensor (ORi™, Masimo Corp., Irvine, CA, USA) was placed on the fourth finger of the arm without a blood pressure cuff. The sensor was covered to avoid light exposure, and it was connected to the oximeter device (Root® platform Pulse CO-Oximetre, Masimo Corp., Irvine, CA, USA) and monitored. The vascular access of all patients without premedication was opened with an 18 gauge (G) granule and the saline infusion was started at a 10 mL/kg/h rate. The Allen test was performed for arterial blood gas (ABG) analysis, and 20 G granules were placed in the radial artery in the non-dominant hand, if possible. Hemodynamic parameters before preoxygenation, ORI values, ​​and ABG samples were taken, and the PaO2 values ​​in the blood gas device (I-STAT®1 Analyzer, Abbott, California, USA) were recorded.

Anesthesia Management

The patients were preoxygenated by administering 100% O2 with a face mask for 5 min. Standard anesthesia induction was applied. Afterwards, a remifentanil infusion was started. All patients were ventilated with the same anesthesia device. The respiratory rate was adjusted so that the tidal volume was 8 mL/kg and the end-tidal carbon dioxide (EtCO2) was between 35 and 45 mmHg.

In terms of the maintenance of anesthesia, the patients were divided into two groups: high-flow anesthesia (group H, n=35) and low-flow anesthesia (group L, n=35). For all patients, the fresh gas flow was provided with 50%/50% O2/air (3/3L/min) and 6% desflurane for the first 10 minutes after intubation. After the first 10 min, the anesthesia was maintained with flow rates of 4 L/min in group H and 1 L/min in group L.

It was planned to increase the fresh gas flow to 4 L/min if the FiO2 fell below 30%, the EtCO2 above 45 mmHg, or the SpO2 fell below 92%. It was planned to change the remifentanil dosage if the mean arterial pressure (MAP) increased more than 20% compared to the control value, administer 5 mg of ephedrine iv if it decreased more than 20%, and administer 0.5 mg of atropine iv if the heart rate (HR) value fell below 50 beats/min.

Ten minutes before the end of the operation, the fresh gas flow was increased to 6 L/min. Anesthetic agents were discontinued at the end of the surgery. Patients with routine wake-up protocols were extubated and taken to the recovery unit.

HR, MAP, SpO2 and ORI values ​​of all patients were recorded before preoxygenation and intraoperatively at 10., 15., 30., 45., 60 minute (min), and then at 30-min intervals until the end of surgery. Also, three ABG samples were taken from all patients before preoxygenation, intraoperative 60 min, and at the end of the surgery, just before starting ventilation with 6 L min, 100% O2, and the PaO2 values ​​and simultaneous ORI and SpO2 values ​​were recorded. In the study, PaO2 of more than 100 mmHg was determined as hyperoxia, and <60 mmHg as hypoxia.

All patients were administered 1 mg/kg tramadol iv and 10 mg/kg acetaminophen infusion intravenously for postoperative pain control 15 min before the end of the operation, and 10 mg/kg metoclopramide iv for nausea and vomiting prophylaxis.

Statistical Analysis

Data was analyzed using the Statistical Package for the Social Sciences version 23.0 (IBM SPSS Inc. Chicago, IL, USA) program. Considering the mean ORI values, it was determined that there should be a minimum of 35 patients in each group, with 95% confidence and 99.99% test power (9). The normality of data distribution was analyzed with the Kolmogorov-Smirnov test. An independent sample t-test was used for the normally distributed data. A paired samples t-test was used for normally distributed data to examine two time-dependent changes. Repeated measures variance analysis was used to examine three or more time-dependent changes. Pearson’s correlation coefficient was used to measure the relationships between normally distributed quantitative variables. The effect of the PaO2 parameter on ORI was analyzed by linear regression. For quantitative data, the results were presented as mean and standard deviation, and for categorical data, as frequency (percentage).The significance level was determined as p<0.05.


The study was completed with 70 patients. The groups were similar in terms of demographic characteristics, ASA risk groups, and duration of surgery (p>0.05) (Table 1).

A comparison of PaO2 values between and within groups is presented in Table 2. It was observed that PaO2 values increased in both groups in the intraoperative period (Table 2).

A comparison of mean values ​​of ORI between and within groups is presented in Table 3. Intraoperative ORI values ​​increased during surgery in both groups (Table 3).

There was a statistically significant, positive weak correlation found with the correlation analysis performed between the mean ORI and PaO2 values ​​obtained at the intraoperative 60th minute and at the end of the surgery, excluding the preoxygenation values ​​between the groups (Table 4).

When the effect of the independent PaO2 variable on the ORI value measured at the 60th min intraoperative and at the end of surgery was examined by linear regression analysis, the regression models established by the time were found to be statistically significant (fintraoperative 60th min=5,326, p-intraoperative 60th min=0.024, FEnd= 6,084, pEnd=0.016). With the regression model established for the 60th min after intubation, it was found that when PaO2 increased by one unit, the ORI value increased by 0.002 units, and with the regression model established for the time measured at the end of the surgery, there was an increase of 0.002 units in the ORI value for one unit increase in PaO2 value (Table 5).

If PaO2 was 100 mmHg, the ORI value at the intraoperative 60th minute was 0.244 (~0.2), and the ORI at the end of the surgery was 0.205 (~0.2), according to the linear regression analysis in which PaO2 was determined as the independent variable and ORI as the dependent variable. Accordingly, when PaO2 was ≥100 mmHg, ORI >0.20 was found in 62.8% of patients at the intraoperative 60th min and in 72.8% at the end of surgery (Figures 2 and 3).

According to the regression analysis, when PaO2 was 150 mmHg, ORI was 0.344 (~0.3) at the intraoperative 60th minute and was 0.305 (~0.3) at the end of surgeries. Similarly, when PaO2 was ≥150 mmHg, ORI >0.30 was found in 63.3% of patients at the intraoperative 60th min and in 75.7% at the end of surgery (Figures 4 and 5).


In this study, in which the effectiveness of ORI in oxygenation monitoring in low- and high-flow anesthesia applications was evaluated, it was concluded that ORI could be used as a non-invasive method, especially in determining hyperoxia due to its positive correlation with PaO2 and because hyperoxia, which is a feared situation in low-flow anesthesia applications, can also be seen in addition to hypoxia.

There are concerns, especially among anesthetists using low-flow techniques, that oxygen cannot reach the tissues in sufficient amounts (10,11). In the study of Çukdar et al. (12), in which low-flow [1 L/min (O2/N2O=0.5/0.5)] and high-flow [4.4 L/min (O2/N2O=1.4/3)] desflurane anesthesia with FiO2 greater than 30% was compared, it was reported that in none of the cases, FiO2 fell below 30% and SpO2 below 97%. Similarly, the fact that the SpO2 value did not fall below 97% in any of the patients in this study and that hypoxia findings were not found in the ABG analysis suggests that this study is compatible with the literature.

Although oxidative stress caused by hypoxia has been a well-known fact for many years, interest in the negative effects of hyperoxia has been increasing recently (13). There have been studies suggesting pathological changes in alveolar cells subjected to hyperoxia (FiO2: 80-90%) for 48 h (14). In the animal study by Clerch and Massaro (15) it was reported that pleural effusion and pulmonary edema appeared in the 48th-60th hours with exposure to intense oxygen (FiO2: >95%). Animals started to die at the 60th hour, and most of them were dead at the 72nd. It is common practice to use 100% O2 in general anesthesia induction and before extubation. However, studies have shown that even a short time (~20-30 minutes) of high O2 use causes atelectasis (16-18). Using computed tomography, Benoit et al. (16) reported that using 100% O2 at the end of general anesthesia (<2.5 h) triggers the formation of atelectasis and low FiO2 levels prevent it. Rothen et al. (19) reported that after the “recruitment” maneuver they applied during general anesthesia, it took 40 min for atelectasis to occur in the group ventilated with low FiO2 (0,4), while atelectasis occurred in the 5th minute in the group getting high FiO2 (1,0) (19). In a multicenter study with 14,441 intensive care patients, it was reported that severe hyperoxia (PaO2>200 mmHg) was associated with higher mortality rates, and weaning was more difficult than both mild hyperoxia (PaO2: 120-200 mmHg) and normoxia (PaO2<100) (20). The optimal PaO2 level is not yet clearly defined (3,21-23). Elmer et al. (22) define hypoxia as a PaO2 of 60 mmHg, normoxia as a PaO2 of 60-100 mmHg, moderate hyperoxia as a PaO2 of 101-299 mmHg, and severe hyperoxia as a PaO2 of >300 mmHg (21). In this study, in which PaO2<60 mmHg was determined as hypoxia and PaO2>100 mmHg as hyperoxia, all the patients were at a normoxic level before preoxygenation, but it was found that PaO2 values ​​at the intraoperative 60th minute and at the end of surgery were PaO2>100 mmHg for all patients. This suggests that hyperoxia rather than hypoxia should be considered in low-flow anesthesia.

Supplying high concentrations of oxygen (>50%) for more than 48 h is reported to possibly cause O2 toxicity. Therefore, O2 application of more than 50% should be limited to 48 h (24). In this study, O2 was used at a concentration of 50% during an average of 5 hours of surgery in both groups.

The most common method for monitoring oxygenation under anesthesia is the pulse oximeter, which has become a universal standard of care and can be measured continuously (25). Reports of respiratory complications during anesthesia have decreased significantly since the use of the pulse oximeter (26). SpO2 has a major limitation when it comes to evaluating hypoxia or hyperoxia in patients receiving oxygen therapy. That is, due to the sigmoid shape of the Oxyhemoglobin dissociation curve, a small change in PaO2 in the vertical part of the curve causes a large difference in SpO2. Consequently, when SpO2 is ≥97%, the PaO2 level can be anywhere between 90 and 600 mmHg. Therefore, monitoring SpO2 alone cannot exclude unwanted hyperoxia in patients receiving O2 treatment. In the impending hypoxia state, the SpO2 decrease is slower than the PaO2 decrease. SpO2 may not decrease before PaO2 is <70 mmHg and SpO2 may be insufficient for the impending danger (27). In cases where SpO2 is insufficient in oxygenation, PaO2 gives precise information, but being costly and invasive, waiting for results, and causing blood loss are the weaknesses of this method (28). In this study, no clinical findings of hypoxia or any decrease in SpO2 and PaO2 values ​​were detected in either group. However, when SpO2 was more than 97%, PaO2 was found to be in a wide range, such as 70-240 mmHg. According to the literature on the subject, the researchers of this study believe that SpO2 cannot be used as a guide in intraoperative O2 management and hyperoxia prevention. Because hyperoxia in ABG increased gradually in both fresh gas flows during the prolonged surgical period and SpO2 remained limited in this regard, it is necessary to be more careful with O2 applications, especially at high currents, and routine ASA monitoring may be insufficient for intraoperative hyperoxia detection.

Studies have reported that ORI can provide an early warning when arterial oxygenation is impaired without any change in peripheral O2 saturation (9,29,30). In the study by Szmuk et al. (29) with 25 healthy children, it was found that during the induction of anesthesia, ORI detected the approaching desaturation (on mean) 31.5 seconds before any change in SpO2 and that this represents a clinically important warning period. This can give clinicians enough time to intervene. Tsymbal et al. (31) also showed that the added warning time provided by the ORI was 46.5 s in obese and 87.0 s in normal BMI patients. Fleming et al. (32) found that the ORI allowed an added warning time of 48.4 s compared to the SpO2 in cardiac surgical patients. In the study of Applegate et al. (9), where they examined the relationship between the intraoperatively measured 1594 ORI value in 106 patients and PaO2 values ​​in the ABG samples taken when clinically necessary, PaO2 was found to be 100 mmHg in all measurements for ORI >0.24, while when ORI>0.55, SpO2 was >96% and PaO2 was ≥150 mmHg (3). In the same study, during the decrease of PaO2 from 500 mmHg to 100 mmHg in 30 min, ORI also decreased, but the response of SpO2 to such various O2 changes was monitored only as from 99% to 96%. Saraçoğlu et al. (33) showed in their study investigating the effect of ORI-guided oxygen titration on morbidity in one lung ventilation with low fresh gas flow, that adjusting ORI with SpO2 and blood gas analysis can prevent hyperoxemia in patients under low-flow or high-flow anesthesia. In this study, SpO2 did not fall below 97% in any patient, and no decrease ​​was observed in ORI values. As reported in the literature, the researchers of this study also believe that ORI can be used as a guide in determining the hyperoxia that may occur at different gas flows, in addition to its role of providing an early warning when arterial oxygenation is impaired without any change in SpO2. According to the linear regression analysis, ORI was 0.244 (~0.2) at the intraoperative 60th min, when PaO2 was 100 mmHg, and ORI was determined to be 0.344 (~0.3) when PaO2 was 150 mmHg.

Accordingly, considering the values ​​at the intraoperative 60th minute, it was observed that although PaO2 was ≥100 in all patients, while the ORI values ​​were >0.244 in 62.8% of them, this rate increased to 72.8% at the end of the surgery. Similar rates were found for the ORI values ​​determined for ≥150 mmHg of PaO2. This suggests that the relationship between ORI and PaO2 gets stronger with the duration of surgery.

Study Limitations

There are some limitations in our study. The limitations of this study can be listed as the inclusion of only the patients who were in the ASA I-II risk group without any problems with their lung mechanics and not being able to evaluate the effectiveness of ORI for situations that may be affected by peripheral perfusion, due to the absence of a condition (such as severe hypotension, etc.). Despite these limitations, our study is valuable as it is one of the few studies showing that ORI is effective in detecting hyperoxia in different fresh gas streams and contributes to the literature.


It is the strong belief of the researchers of this study that ORI, which can be useful in every stage of monitoring oxygenation starting from preoxygenation, can be a guide in terms of titrating the O2 levels in long operations with its relationship with PaO2. Considering the defining power of the model designed in the study (R2=0.059, R2=0.069), it can be evaluated whether the identification power of the model increases by increasing the number of participants. By evaluating variables other than PaO2 that affect the ORI value, regression models with higher definition powers can be developed with the inclusion of these variables. The usefulness of the ORI values ​​determined for Pa100 mmHg and 150≥mmHg of PaO2 should be tested with different clinical studies.


We are grateful to the  Zonguldak Bulent Ecevit University Scientific Researches Projects Coordination Unit for supporting the study.


Ethics Committee Approval: This study approved by the Zonguldak Bulent Ecevit University Clinical Research Ethics Committee (date: 07.11.2018, protocol no: 2018-221-07/11).

Informed Consent: Written informed consent was obtained from the patients.

Authorship Contributions

Concept: H.O., G.K., B.G.A., R.D.O., O.P., H.A., Design: H.O., G.K., B.G.A., R.D.O., O.P., H.A., Data Collection and/or Processing: H.O., G.K., B.G.A., O.P., H.A., Analysis and/or Interpretation: H.O., G.K., B.G.A., O.P., H.A., Literature Research: H.O., G.K., H.A., Writing: H.O., G.K., B.G.A., H.A.

Conflict of Interest: No conflict of interest was declared by the authors.

Funding: This study was supported by  Zonguldak Bulent Ecevit University Scientific Researches Projects Coordination Unit.

  1. Eger EI II. Uptake and distribution. In: Miller RD, editor. Miller’s Anesthesia. 6th ed. Philadelphia: Elsevier Churchill Living Stone; 2005. p. 131-53.
  2. Baum JA, Aitkenhead AR. Low-flow anaesthesia. Anaesthesia 1995;50 Suppl:37-44.
  3. Scheeren TWL, Belda FJ, Perel A. The oxygen reserve index (ORI): a new tool to monitor oxygen therapy. J Clin Monit Comput 2018;32:379-89.
  4. Szmuk P, Steiner JW, Olumu PN, Curuz JD, Sessler D. Oxygen reserve index: a new, noninvasive method of oxygen reserve measurement. Presented at the American Society of Anesthesiologists Annual Meeting 2014;121:681-9.
  5. Bateman NT, Leach RM. ABC of oxygen. Acute oxygen therapy. BMJ 1998;317:798-801.
  6. Yoshida K, Isosu T, Noji Y, et al. Adjustment of oxygen reserve index (ORi™) to avoid excessive hyperoxia during general anesthesia. J Clin Monit Comput 2020;34:509-14.
  7. Kumagai M, Kurihara H, Ishida K, Komatsu H, Suzuki K. The Oxygen Reserve Index as a determinant of the necessary amount of postoperative supplemental oxygen. Minerva Anestesiol 2021;87:439-47.
  8. Schulz KF, Altman DG, Moher D; CONSORT Group. CONSORT 2010 Statement: Updated guidelines for reporting parallel group randomised trials. J Clin Epidemiol 2010;63:834-40.
  9. Applegate RL, Dorotta IL, Wells B, Juma D, Applegate PM. The Relationship Between Oxygen Reserve Index and Arterial Partial Pressure of Oxygen During Surgery. Anesth Analg 2016;123:626-33.
  10. Baum JA, Anaesthetic methods with reduced fresh gas flows. In: Baum JA, editor. Low flow anaesthesia. The theory and practise of low flow, minimal flow and closed system anaesthesia. 2th ed. Oxford, Butterworth Heinemann, 2001.
  11. Hanci V, Yurtlu S, Ayoğlu H, et al. Effect of low-flow anesthesia education on knowledge, attitude and behavior of the anesthesia team. Kaohsiung J Med Sci 2010;26:415-21.
  12. Çukdar G, Özkoçak Turan I, Ayoğlu H, Hancı V, Yurtlu S, Özer Y. Comparison of the Effects of Low and High Flow Desflurane Anesthesia on Hemodynamics and Anesthetic Gas Consumption. Turk J Anaesthesiol Reanim 2008;36:222-9.
  13. Kallet RH, Branson RD. Should Oxygen Therapy Be Tightly Regulated to Minimize Hyperoxia in Critically Ill Patients? Respir Care 2016;61:801-17.
  14. Roan E, Wilhelm K, Bada A, et al. Hyperoxia alters the mechanical properties of alveolar epithelial cells. Am J Physiol Lung Cell Mol Physiol 2012;302:L1235-41.
  15. Clerch LB, Massaro D. Tolerance of rats to hyperoxia. Lung antioxidant enzyme gene expression. J Clin Invest 1993;91:499-508.
  16. Benoit Z, Wicky S, Fischer JF, et al. The effect of increased FIO(2) before tracheal extubation on postoperative atelectasis. Anesth Analg 2002;95:1777-81.
  17. Edmark L, Kostova-Aherdan K, Enlund M, Hedenstierna G. Optimal oxygen concentration during induction of general anesthesia. Anesthesiology 2003;98:28-33.
  18. Edmark L, Auner U, Enlund M, Ostberg E, Hedenstierna G. Oxygen concentration and characteristics of progressive atelectasis formation during anaesthesia. Acta Anaesthesiol Scand 2011;55:75-81.
  19. Rothen HU, Sporre B, Engberg G, Wegenius G, Högman M, Hedenstierna G. Influence of gas composition on recurrens of atelectasis after a reexpansion maneuver during general anesthesia. Anesthesiology 1995;82:832-42.
  20. Helmerhorst HJ, Arts DL, Schultz MJ, et al. Metrics of arterial hyperoxia and associated outcomes in critical care. Crit Care Med 2017;45:187-95.
  21. De Graaff AE, Dongelmans DA, Binnekade JM,  de Jonge E. Clinicians’ response to hyperoxia in ventilated patients in a Dutch ICU depends on the level of  FiO2. Intensive Care Med 2011;37:46-51.
  22. Elmer J, Scutella M, Pullalarevu R, et al. The association between hyperoxia and patient outcomes after cardiac arrest: analysis of a high-resolution database. Intensive Care Med 2015;41:49-57.
  23. Del Castillo J, López-Herce J, Matamoros M, et al. Hyperoxia, hypocapnia and hypercapnia as outcome factors after cardiac arrest in children. Resuscitation 2012;83:1456-61.
  24. Tinits P. Oxygen therapy and oxygen toxicity. Ann Emerg Med 1983;12:321-8.
  25. Hanning CD, Alexander-Williams JM. Pulse oximetry: A practical review. BMJ 1995; 311:367-70.
  26. Pedersen T, Nicholson A, Hovhannisyan K, Møller AM, Smith AF, Lewis SR. Pulse oximetry for perioperative monitoring. Cochrane Database Syst Rev 2014;2014:CD002013.
  27. Fu ES, Downs JB, Schweiger JW, Miguel RV, Smith RA. Supplemental oxygen impairs detection of hypoventilation by pulse oximetry. Chest 2004;126:1552-8.
  28. Davis MD, Walsh BK, Sittig SE, Restrepo RD. AARC clinical practice guideline: blood gas analysis and hemoximetry: Respir Care 2013;58:1694-703.
  29. Szmuk P, Steiner JW, Olomu PN, Ploski RP, Sessler DI, Ezri T. Oxygen reserve index: a novel noninvasive measure of oxygen reserve-a pilot study. Anesthesiology 2016;124:779-84.
  30. Simpao AF, Galvez JA. When seconds count, buy more time: the oxygen reserve index and its promising role in patient monitoring and safety. Anesthesiology 2016;124:750-1.
  31. Tsymbal E, Ayala S, Singh A, Applegate RL, Fleming NW. Study of early warning for desaturation provided by Oxygen Reserve Index in obese patients. J Clin Monit Comput 2021;35:749-56.
  32. Fleming NW, Singh A, Lee L, Applegate RL. Oxygen Reserve Index: utility as an early warning for desaturation in high-risk surgical patients. Anesth Analg 2021;132:770-6.
  33. Saraçoğlu A, Yamansavci Şirzai E, Yildizeli B, Yüksel M, Aykaç ZZ. Oxygen reserve index guided oxygen titration in one lung ventilation with low fresh gas flow. Turk J Med Sci 2021;51:2413-9.