Material in this chapter has been reproduced with permission from Elfeky et al.2 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.
Search results
The sensitive search of eight electronic databases and other sources in September 2021 retrieved 5892 references. In total, 4072 titles and abstracts were screened after deduplication. Of these, 3784 were excluded for reasons including duplicates, studies that were not systematic reviews of RCTs, studies where the target group was not surgical patients and studies where the intervention of interest was not perioperative oxygen therapy (see Appendix 7).
The remaining 288 articles were obtained for full-text screening. Of these, a further 175 were excluded, for reasons similar to those quoted above, and 59 reviews were selected for final inclusion. Thirteen additional RCTs were identified and were used to update the analyses in relevant anchoring reviews.
The PRIOR diagram describing each of these stages is presented in Figure 2.
FIGURE 2 PRIOR flow diagram (in relation to PRIORs). Reproduced with permission from Elfeky et al.2 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)
Mapping of included reviews
We included 59 reviews in this systematic overview: 29 reviews compared high versus low FiO2;53,67–74, 133–152 13 reviews compared HFNO versus COT62,75,76,94–96,153–158 and 17 reviews compared NIV versus COT.63,65,66,91,159–171 Of the 29 reviews comparing high versus low FiO2, 16 used an FiO2 of 80% oxygen in the group with a high FiO2 during anaesthesia; 5 used more than 80% oxygen; 3 used more than 60% oxygen and 1 used 60–90% oxygen.
Considerable overlapping between the primary studies included in the systematic reviews was observed (see Appendix 2, Tables 15–17).
All the included reviews examined the effectiveness of different oxygenation strategies in the adult population. No systematic reviews have attempted to synthesise the evidence in the paediatric population.
No systematic reviews in this overview assessed the effectiveness of oxygenation strategies at the preoperative stage. At the intraoperative stage, no evidence was identified on the effectiveness of NIV or HFNO compared with COT. Likewise, no evidence was identified on the effectiveness of postoperative administration of high FiO2 on patient outcomes compared with low FiO2.
We selected five anchoring reviews62,72,133,134,171 which provided comprehensive up-to-date evidence on the effectiveness of perioperative oxygen therapy.
Three anchoring reviews assessed the effectiveness of high (80%) versus low (30–35%) FiO2 in patients undergoing surgery.72,133,134 de Jonge et al. performed a systematic review of RCTs comparing the effect of 80% versus 30–35% FiO2 on the incidence of SSI.72 Lim et al. conducted a systematic review and meta-analysis to determine whether high FiO2 compared with low FiO2 has more deleterious postoperative clinical outcomes in patients undergoing non-thoracic surgery under general anaesthesia.133 Markwei et al. assessed the effect of 80% versus 30% oxygen concentration on the incidence of postoperative nausea and/or vomiting (PONV).134
Two anchoring reviews examined the effectiveness of postoperative non-invasive respiratory support. Chaudhuri et al. assessed the effect of HFNO in the immediate postoperative period compared to COT.62 Hui et al. assessed the effectiveness of routine postoperative non-invasive respiratory support after elective surgery.171
Table 3 outlines the characteristics of chosen anchoring reviews.
| Review ID | Included RCTs (participants), n | Care stage | Patient population (age group) | Intervention/comparator | Relevant outcomes | Summary of findings | Reasons for being selected as anchoring review |
|---|---|---|---|---|---|---|---|
| de Jonge et al. 201972 | 21 RCTs (17 RCTs in final analysis, 7817 participants) | Intra- and postoperative | Patients undergoing surgical procedures (adults) | FiO2: high (80%), low (30–35%) | SSI | There was no evidence that high FiO2 reduces SSI (RR: 0.89, 95% CI 0.73 to 1.07). There was a beneficial effect of high FiO2 in intubated patients (RR 0.80, 95% CI 0.64 to 0.99) but not in non-intubated patients (RR: 1.20, 95% CI 0.91 to 1.58; test of interaction; p = 0.048). | Most comprehensive review on the effects of high vs. low FiO2 on SSI |
| Lim et al. 2021133 | 26 RCTs (4991 patients) | Intra- and postoperative | Patients undergoing non-thoracic surgery (adults) | FiO2: high (80–100%), low (30–40%) | Primary outcome: mortality within 30 days. Secondary outcomes: atelectasis, pneumonia, respiratory failure, PPCs, ICU admissions, and length of hospital stay | Mortality in the high FiO2 group did not differ from that in the low FiO2 group (RR 0.91, 95%, CI 0.42 to 1.97), nor were there any significant differences between the groups in such outcomes as pneumonia, respiratory failure, PPCs, ICU admission and length of hospital stay. High FiO2 was associated with postoperative atelectasis more often (RR 1.27, 95% CI 1.00 to 1.62). In subgroup analysis of BMI > 30 kg/m2, these parameters were similarly affected between the groups | Most up-to-date review on the effect of high FiO2 on mortality and pulmonary outcomes in non-thoracic surgery |
| Chaudhuri et al. 202062 | 11 RCTs (2201 patients) | Postoperative | Patients undergoing surgical procedures (adults) | HFNO/COT or NIV | Reintubation, escalation of respiratory support, mortality, ICU and hospital LOS, postoperative hypoxemia, and treatment complications | HFNO use in the postoperative period was associated with a lower reintubation rate (RR 0.32, 95% CI 0.12 to 0.88; moderate certainty) and decreased escalation of respiratory support (RR 0.54, 95% CI 0.31 to 0.94; very low certainty). Post hoc subgroup analysis suggested that this effect was driven by patients who were obese and/or at high risk (subgroup differences, p = 0.06). No differences were found in any of the other stated outcomes between HFNO and COT | Most comprehensive review on the effect of HFNO in the immediate postoperative period |
| Hui et al. 2022171 | 38 RCTS (9782 patients) | Postoperative | Adult patients undergoing major elective surgery | Non-invasive respiratory support, including CPAP, NIV, and HFNO/standard postoperative care | The primary outcome was pneumonia, and the secondary outcome was PPCs | No difference was found in pneumonia (RD −0.01, 95% CI −0.02 to 0.00) and PPCs (RD −0.11, −0.23 to 0.01) in patients receiving non-invasive respiratory support compared with standard care. Subgroup analyses did not identify a benefit of CPAP, NIV or HFNO in preventing pneumonia | The largest and most comprehensive evidence synthesis to date of whether preventive postoperative non-invasive respiratory support, including CPAP, NIV, and HFNO, can reduce the incidence of pneumonia after elective surgery |
| Markwei et al. 2022134 | 15 trials with a total of 7170 patients. 10 trials were included in meta-analysis (6773 patients) | Intraoperative | Adult patients undergoing surgery | 80% vs. 30% FiO2 | The incidence of postoperative nausea and vomiting | The meta-analysis of 10 trials did not find a difference in postoperative nausea and vomiting: RR 0.97 (95% CI 0.86 to 1.08) | Most comprehensive review on the effect of 80% vs. 30% oxygen concentration on the incidence of postoperative nausea and/or vomiting |
Appendix 3, Table 18 highlights the characteristics of non-anchoring reviews. Characteristics of additional RCTs identified from the updated search are provided in Table 4.
| Study | Country | Patients randomised (n) | Type of surgery | Care stage | Intervention vs. control | Relevant outcomes |
|---|---|---|---|---|---|---|
| Alvandipour et al. 2019172 | Iran | 85 | Colorectal surgery | Intraoperative +2 hours postoperatively | FiO2 80% vs. 30% | SSI, atelectasis, pneumonia, respiratory failure, length of hospital stay |
| Ferrando et al. 2020173 | Spain | 740 | Abdominal surgery | Intraoperative +3 hours postoperatively | FiO2 80% vs. 30% | Mortality within 30 days, mortality within 180 days, SSI, pulmonary complications |
| Holse et al. 2022174 | Denmark | 600 | Major non-cardiac surgery | Intraoperative +2 hours postoperatively | FiO2 80% vs. 30% | SSI, pneumonia, respiratory failure, mortality |
| Lin et al. 2021175 | China | 652 | Laparoscopic gastric and colorectal malignancies surgery | Intraoperative | FiO2 80% vs. 40% | SSI, pneumonia, atelectasis |
| Mayank et al. 2019176 | India | 94 | Colorectal surgery | Intraoperative +6 hours postoperatively | FiO2 80% vs. 33% | SSI |
| Reiterer et al. 2021177 | Austria | 260 | Major abdominal surgery | Intraoperative +2 hours postoperatively | FiO2 80% vs. 30% | Mortality, unplanned ICU admission, respiratory failure |
| Yerra 2021178 | India | 178 | Abdominal surgery | Intraoperative +2 hours postoperatively | FiO2 80% vs. 30% | SSI, postoperative nausea and vomiting, respiratory complications |
| Ferrando et al. 2019179 | Spain | 64 | Laparoscopic bariatric surgery | First 3 hours postoperatively | HFNO at 60 l/minute vs. Venturi mask 15l/minute | Atelectasis, hypoxaemia, pneumonia, ARDS, respiratory failure, length of hospital stay |
| Pibul et al. 2021180 | Thailand | 67 | Cardiac surgery with midline sternotomy | First 24 hours postoperatively | HFNO 20–60 l/minute vs. face mask/nasal cannula | Reintubation rate, length of hospital stay, length of ICU stay |
| Theologou et al. 2021181 | Greece | 99 | Cardiac surgery | First 48 hours postoperatively | HFNO 60 or 40 l/minute vs. venturi mask | Reintubation, mortality, escalation of respiratory support |
| Twose et al. 2019182 | UK | 20 | Major head and neck surgery | Postoperatively at the point when mechanical ventilation stopped | HFNO of 60 l/minute vs. standard O2 | Pneumonia, ARDS, length of hospital stay, ICU length of stay |
| Vourc’h et al. 2020183 | France | 98 | Cardiac surgery | First 48 hours postoperatively | HFNO 45 L/minute vs. high-flow face mask 15 l/minute | Reintubation, escalation of respiratory support, ICU length of stay |
Methodological quality of included reviews
We assessed the risk of bias in included reviews with the ROBIS tool. The ROBIS assessment is organised in three phases:
- Phase 1: assessment of relevance
- Phase 2: identification of concerns within the review process. It is split into four domains: study eligibility criteria, study selection, data collection and appraisal of included studies, synthesis and findings
- Phase 3: overall risk of bias in the interpretation of findings and whether any limitations identified in phase 2 were considered by study authors
Questions included in phase 3 of the assessment are related to how the original review authors presented and interpreted their findings. As we updated the meta-analyses of the anchoring reviews and made our own interpretation of the updated findings, we focused on and presented findings related to risk of bias concerns identified in phase 2 of the assessment.
All reviews achieved a low risk of bias rating for specification of study eligibility criteria. Three reviews were judged to be at low risk of bias for the identification and selection of studies;72,133,171 one review was rated as unclear134 and one was at high risk.62 All reviews achieved a low risk of bias rating for domain 3 which assessed concerns regarding methods used to collect data and appraise studies. All reviews were judged to be at low risk of bias in domain 4, regarding the synthesis and findings.
Figure 3 presents the results of the phase 2 ROBIS rating according to the four different domains.
FIGURE 3 ROBIS assessment for included anchoring reviews. Reproduced with permission from Elfeky et al.2 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)
Risk of bias in newly identified randomised controlled trials
Six of the 13 newly identified RCTs were considered at low risk of bias for all risk of bias domains.172–174,177,180,184 Five RCTs had some cause for concern.175,179,181–183 Two trials reporting on SSI were considered at high risk of bias.176,178 The domain most frequently assessed at high risk or of some concern of bias was deviations from the intended interventions.
Risk of bias judgements for important outcomes across comparisons are presented in Figure 4.
FIGURE 4 Risk of bias assessment for additional RCTs. Reproduced with permission from Elfeky et al.2 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)
Risk of bias reassessment
The anchoring review (de Jonge et al. 2019) assessed risk of bias in included RCTs using the original Cochrane risk of bias tool. Therefore, the risk of bias for included RCTs was reassessed for the relevant outcome (SSI) using the revised ROB2 tool. Details of RoB2 ratings are provided in Appendix 4, Figure 40.