Blood pressure targets during general anaesthesia for noncardiac surgery

Editor, Observational cohort studies have demonstrated an association between intraoperative hypotension and postoperative complications and mortality. Unfortunately, such observational studies are at a high risk of bias, primarily due to confounding, and it is unclear whether hypotension is causally related to these outcomes. Further, these studies examine the actual blood pressure achieved and not a given blood pressure target. We performed this systematic review to identify clinical trials testing various blood pressure targets during general anaesthesia. This review was part of a larger review project including clinical trials of adult patients undergoing noncardiac surgery with general anaesthesia. This manuscript focuses on trials assessing various blood pressure targets. Details on the methodology are provided in the protocol and a previous manuscript.1 We conducted a comprehensive search of PubMed and Embase on March 8, 2021, to identify relevant trials reporting postoperative outcomes. A total of 13 trials, including a total of 2466 patients, were identified (eFigure 1, https://links.lww.com/EJA/A705, Table 1).2–14 Given the heterogeneity between the trials, it was not feasible to perform meta-analyses or GRADE evaluation. An overview of the trials is provided in Table 1 and eTable 1, https://links.lww.com/EJA/A705. We chose to describe six outcomes (Table 1), as the remaining outcomes were defined heterogeneously and reported in a limited number of trials (eTable 2, https://links.lww.com/EJA/A705). Table 1 - Overview of trials on blood pressure targets Blood pressuresa Trial Year of inclusion Type of surgery n Low target (achieved mean [SD]) High target (achieved mean [SD]) Reported outcomesb Bias assessmentc Individualised targets Thompson, 19784d NR Ortho. 30 MAP 50 (NR) ±20% MAP (NR) Delirium, renal complications, cardiac complications, intraoperative bleeding. Intermediate Shapira, 19977e NR Ortho. 16 MAP 50 (53 [4]) ±20% MAP (82 [13]) Intraoperative bleeding Intermediate Han, 20062f NR Ortho. 56 MAP 55–65 (60 [8]) ±20% MAP (84 [8]) Delirium Intermediate Jin, 20166g NR Eye-Nose 45 -40% MAP (NR) -20% MAP (NR) Delirium High Futier, 20175 2012–16 GIh 292 SBP > 80 (116 [24]) ±10% SBP (123 [25]) Mortality, hospital length of stay, delirium, renal complications, cardiac complications, intraoperative bleeding. Intermediate Langer, 20193 2014–16 Multiple 101 No target (85 [11]) > -10% of MAP (92 [9]) Mortality, hospital length of stay, delirium Intermediate Xu, 20208i 2019 Ortho. 156 -10–20% MAP (75 [1]) > -10% of MAP (91 [1]) Hospital length of stay, delirium Intermediate Fixed targets Morrison, 201113 2007–09 Trauma 90 MAP 50 (65 [12]) MAP 65 (69 [13]) Mortality, intraoperative bleeding Intermediate Carrick, 201612 2007–13 Trauma 180 MAP 50 (66 [12]) MAP 65 (69 [14]) Mortality, renal complications, cardiac complications, intraoperative bleeding Intermediate Wu 201711j 2015–16 GI 678 MAP 65–79 (72 [5]) MAP 96–110 (100 [6]) Mortality, renal complications, intraoperative bleeding Intermediate Hu, 20219 2016–20 NR 322 MAP 60–70 (NR) MAP 90–100 (NR) Mortality, Hospital length of stay, delirium, renal complications, cardiac complications, intraoperative bleeding Intermediate Qui, 202110 2016–17 Lumbar 42 MAP 65 (65 [3]) MAP 75 (75 [2]) Hospital length of stay, Renal complications Intermediate Wanner, 202114 2016–19 Multiple 458 MAP > 60 (NR)k MAP > 75 (NR)k Mortality, renal complications, cardiac complications Intermediate GI, Gastrointestinal; MAP, Mean arterial pressure; n, Sample size; NR, not reported; Ortho, Orthopaedic; SBP, Systolic blood pressure.All blood pressures are in mmHg.aFor the individualised targets, the targets are expressed as a percentage of the baseline blood pressurebOutcomes described in results. Additional reported outcomes are provided in eTable 3, https://links.lww.com/EJA/A705.cOverall bias assessment, for more detailed bias assessment, see eTable2, https://links.lww.com/EJA/A705.dThree arms, two MAP 50 groups achieved by either nitroprusside or halothane.eLow blood pressure group: MAP 50 mmHg and haemodilution.fHaemodilution in both groups.gThree arms, intermediate group 30% < MAP, mean NRh95% GI, 5% other surgeryiThree arms, intermediate group 0–10% <MAP, mean 84jThree arms, intermediate group MAP 80–95, mean 88kNumerical achieved mean not reported but illustrated in Figure 2E in the manuscript. Approximately mean MAP of 77 mmHg in the low target group and 83 mmHg in the high target group. Seven trials used preoperative mean or systolic arterial blood pressure to determine individualised targets during anaesthesia.2–8 These trials generally compared an individualised higher intraoperative blood pressure target to a lower individualised or fixed target (Table 1). Three of the trials reported a difference in the actual measured mean blood pressures of at least 10 mmHg between the groups.2,7,8 Two trials did not report the achieved blood pressures and two trials did not report a blood pressure difference of at least 10 mmHg (Table 1).3–6 All trials except one were assessed as having an intermediate risk of bias (eTable 3, https://links.lww.com/EJA/A705). Two trials reported mortality and three reported hospital length of stay. None reported a difference in these outcomes.3,5,8 Three trials found a protective effect on delirium of a higher individualised blood pressure target,5,6,8 whereas one trial found no significant difference.3 Two other trials reported very few delirium events.2,4 Two trials reported renal complications. Thompson et al. found no significant difference in postoperative creatinine-clearance, whereas Futier et al. found a significant difference in renal dysfunction in favor of a higher individualised blood pressure target.4,5 The same two trials reported cardiac complications and found no difference between groups. Thompson et al. and Shapira et al. found significantly greater bleeding and transfusion need in the individualised high blood pressure target group, whereas Futier at al. found no significant difference.4,5,7 Six trials compared one higher fixed mean arterial pressure (ranging from 65 to 110 mmHg) to another lower fixed mean arterial pressure (ranging from 50 to 79 mmHg). Two trials did not achieve a blood pressure difference of at least 10 mmHg between groups.12,13 Two other trials succeeded in achieving a blood pressure difference of at least 10 mmHg, and two trials did not report the achieved blood pressures (Table 1).9–11,14 All trials were assessed as having an intermediate risk of bias (eTable 2, https://links.lww.com/EJA/A705). Morrison et al. found a decrease in 24-h mortality in the lower target group, but the difference in 30-day mortality was nonsignificant.13 Four other trials found no difference in mortality.9,11,12,14 Two trials reported hospital length of stay and found no difference.9,10 Hu et al. was the only trial reporting delirium and found a significantly lower incidence in the high blood pressure target group.9 Five trials reported renal complications. Only Wu et al. found increased postoperative creatinine in the hypotensive group.9–12,14 Three trials reported cardiac complications and found no significant differences between groups.9,12,14 Four trials reported data on intraoperative bleeding with no differences between groups.9,11–13 In conclusion, there is very limited evidence from randomised clinical trials on optimal blood pressure targets during general anaesthesia. Only 13 trials including a total of 2466 patients were identified, and the trials were very heterogeneous in terms of blood pressure targets, patient population, type of surgery, and in the reported outcomes. Given this heterogeneity, it was not feasible to perform meta-analyses. Moreover, many of the trials did not achieve a difference in achieved blood pressures between the groups, making the results difficult to interpret. Our findings are similar to a recent review focused on cardiac surgery, where trials were also few, heterogenous, and with relatively small sample sizes.15 In that review, there were no clear differences in clinical outcomes when comparing different blood pressure targets.15 Given the limitations of observational studies within this area and the lack of randomised clinical trials, there is a need for additional large clinical trials to guide clinical practice.