Background
Postoperative cognitive impairment, including postoperative delirium (POD) and postoperative cognitive dysfunction (POCD), is a common neuropsychological disorder after surgery among patients [
1]. Although neither POD nor POCD has a formal definition, it is recognized that they do exist [
2]. POD is an acute change of patient’s attention, consciousness, perception or cognition, which occurs in several hours or days after the operation and its duration is usually short (a few days) [
3,
4]; POCD is characterized by short-term disturbances in patients’ memory, executive functioning, personality or sleep, which usually appears in weeks or months after surgery and can last for months or even longer [
5]. POD and POCD are leading to adverse results, including prolonged length of hospital stay, increased mortality and unexpected complications, which results in increased medical costs and decreased the quality of patient’s life [
3,
6‐
8].
The underlying pathophysiology of POD or POCD is multifactor and complicated. Immutable risk factors, such as surgery types, age and baseline cognitive function have been identified [
5,
7]. Although the definitive preventive or therapeutic measure of POD or POCD is still unknown, there are increasing studies shows that hypoperfusion of the brain caused by hypotension during the surgery may be one pathogenic mechanism [
9‐
12].
Intraoperative hypotension, though lack of a widely accepted definition, often appears during anesthesia. It is usually manifested as mean arterial pressure (MAP) below the level of a predefined threshold during surgery [
13,
14]. Hypotensive anesthesia brings a lot of obvious conveniences for some surgeries, including visualized anatomy, dry surgical area, and reduced blood loss during surgery [
15]. Thus, intraoperative hypotension induced by anesthesia is also frequently observed. It seems plausible that the temporary brain perfusion of a patient becomes impaired when experiencing severe and prolonged low blood pressure, leading to cognitive impairment [
16]. However, the specific correlation between intraoperative hypotension and postoperative cognitive function remains unclear and controversial. Evidence has shown a pivotal role for intraoperative hypotension in the development of cognitive impairment after surgery [
9,
12,
17], whereas others have not [
16,
18‐
23]. A single study cannot elucidate all factors while different study designs may cause selection bias.
Therefore, the goal of the current meta-analysis is to evaluate the association between intraoperative hypotension and the incidence of POD or POCD undergoing surgery.
Discussion
In this study, we assessed the correlation of intraoperative hypotension and postoperative cognitive impairment following surgery and anesthesia. For the incidence of POD or POCD, the combined results illustrated no significant difference between the low-target and the high-target participants. TSA analyses showed that there was not enough information to confirm or reject the results, which requires a large number of randomized participants to achieve the boundary line. Furthermore, it demonstrated that intraoperative hypotension prolonged the length of ICU stay. Nevertheless, we did not notice obvious differences in the mortality, the length of hospital stay, and MV time between different groups.
Postoperative cognitive impairment (POD and POCD) is associated with high mortality and increased societal costs, which received increasing attention [
1,
6‐
8]. Common concepts on the etiology are anesthesia-, surgery-, and patient-related factors [
2,
3,
5]. Previous studies have reported that inflammation, neurotransmitter imbalance and sleep deprivation play an essential role in the pathogenesis of cognitive impairment [
4,
6,
7,
32]. Moreover, some studies indicated that intraoperative hypotension was linked to the development of POD or POCD [
9,
12,
17].
In this meta-analysis, we found that the mean ages of patients were more than 50 years old in most of the included researches. The possible explanation is that as population aging, more elderly patients are undergoing the operation, leading to a higher risk of cognitive impairment than younger patients. Furthermore, in this meta-analysis, two studies utilized the CAM-ICU scale [
21] and MMSE scores [
9] to assess the incidence of POD. The CAM-ICU scale had almost 100% sensitivity, specificity and interrater reliability [
33], and MMSE scores had 96% sensitivity and 38% specificity [
34]. For POCD, the incidence of three studies was elevated by neuropsychological tests [
21‐
23], a sensitive method of evaluating the change and detecting beneficial results [
35], while the remaining one used MMSE scores [
9]. According to our study, the incidence of POD and POCD in the high-target group is only 3 and 7%, which were marginally lower than the reported rate in a systematic review (11–43% and 15–25%) [
36]. Possible interpretations for this discrepancy include the considerable difference in test methods, the definition of POD or POCD, the baseline evaluation, and the control groups. Additionally, not only the occurrence of POD and POCD varied widely depending on the surgical variables, demographic as well as the clinical environment, but also increased with advancing age [
3,
4].
Our meta-analysis suggested that intraoperative hypotension has no identified relationship on the incidence of POD, in line with previous studies on cardiac or non-cardiac surgery [
9,
21]. However, two studies about colorectal [
12] and surgical surgery [
11] (a logistic regression and a retrospective cohort analysis) showed that intraoperative hypotension could significantly increase the incidence of POD. Possible reasons for the finding were that the definitions of hypotension used were different, and the above-mentioned two studies were not RCTs. Furthermore, a prospective cohort study on older patients during surgery found that both the degree and duration of intraoperative hypotension were not associated with POD, but fluctuations of intraoperative blood pressure was significantly related to the risk of POD [
18]. Therefore, close monitoring and appropriate intervention of blood pressure during surgery seem to be crucial for preventing POD, which is to be clarified by RCTs with a larger sample size.
Our study concluded that there is no significant correlation between the POCD incidence and intraoperative hypotension. This conclusion is consistent with most studies [
19‐
23,
37‐
39] except for a clinical, randomized study [
9], which found that maintaining mean perfusion during cardiopulmonary bypass surgery at physiological values (80–90 mmHg) is associated with less early POCD. This discrepancy may be attributable to methodological issues concerning POCD: this study assessed it at 48 h after surgery, while others on over 3 months postoperatively; hence, Siepe et al. defined this cognitive impairment as early POCD. Additionally, regarding the effect of postoperative hypotension on postoperative cognitive impairment, no correlation was observed between postoperative hypotension and POCD [
39], and no data was available about the relationship between postoperative hypotension and POD.
The result of our secondary outcomes indicated that intraoperative hypotension significantly prolonged the length of ICU stay, while not being followed by increased mortality, the length of hospital stay, and MV time. Furthermore, we also performed subgroup analysis for the effect of surgery type (cardiac versus non-cardiac surgery) and anesthesia type (general versus epidural anesthesia) on the incidence of POCD, which is consistent with a study revealing that the incidence is not associated with the type of anesthesia and surgery [
40].
Several limitations of this study should be noted. First, the number of eligible studies was relatively small, resulting in a high risk of overestimation effects and a lack of publication bias assessment. Second, many factors such as surgery types, definitions of intraoperative hypotension or POD/POCD, intraoperative hypotension levels, and evaluation tools of POD or POCD varied among included studies. Thus, clinical heterogeneity was relatively high, which may weaken the reliability and precision of our conclusion. Third, given the fact that the incidence of POD and POCD were our primary outcomes, studies that did not contain POD or POCD data were excluded; thus, the application of our secondary outcomes may be limited. Therefore, the results of this meta-analysis should be further confirmed by much more high-quality studies.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.