Based on the findings of the Global Burden of Disease study, there was a significant two-fold increase in cardiovascular disease (CVD) prevalence between 1990 and 2019, with the number of afflicted individuals escalating from 271 million to 523 million [1]. The progressive evolution of cardiac surgical technology has engendered a burgeoning cohort of patients eligible for surgical interventions. Despite the advancements in perioperative management, anesthesia, cardiopulmonary bypass (CPB), and the surgical environment remain vulnerable to postoperative functional changes, which give rise to consequential complications. Among these, the incidence of prolonged mechanical ventilation (PMV) constitutes a substantial proportion, accounting for up to 53.27% [2]. PMV has been correlated with heightened reintubation and prolonged duration of stay in the intensive care unit (ICU), leading to more pulmonary complications [3].

Globally, the number of PMVs has increased. Within the United States, the demand for PMVs reached an estimated 625,000 cases in 2020, while in Taiwan, 92,324 patients required PMVs between 2015 and 2019 [4, 5, 6]. PMV not only results in negative patient experiences but also reduces their quality of life. The complications arising from PMV, encompassing muscle wasting, functional impairment, and diaphragmatic dysfunction, have been found to be associated with extended hospitalization periods and elevated mortality. Consequently, these complications impose considerable financial strain upon afflicted families [7, 8, 9, 10, 11]. The reported incidence of PMV in patients undergoing cardiac surgery ranges from 1.96% to 53.27% [2, 10, 12, 13, 14], and the risk factors for PMV include older age [12, 15], emergency surgery [10, 15, 16], and being female [9, 15]. However, these findings are controversial. The current evidence on PMV prevention and treatment in cardiac surgical patients has yet to be comprehensively synthesized, largely due to variations in PMV definitions and study designs across different investigations. In addition, the effect of PMV on in-hospital mortality remains controversial [8, 14]. Therefore, identifying PMV-related risk factors and prognoses may optimize clinical treatment and decision-making to assist surgeons in surgical planning and postoperative management.

Thus, through a systematic review and meta-analysis, the present synthesis endeavors to investigate the incidence, identify risk factors and analyze in-hospital mortality in patients experiencing PMV after cardiac surgical procedures. The overarching objective of this study is to furnish evidence that informs the prevention and effective management of PMV-associated complications in this specific patient population.

This systematic review and meta-analysis were conducted per the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines (Supplementary material 1) and recorded at PROSPERO.

This systematic review and meta-analysis represent the inaugural endeavor to comprehensively examine the incidence, risk factors, and in-hospital mortality concerning PMV in cardiac surgery patients. The synthesis included 32 studies involving a total of 68,766 patients and yielded the subsequent key findings: PMV incidence was 20%; 15 risk factors were associated with PMV (advanced age, being female, EF 50, higher BMI, BMI $>$28, NYHA $\ge$III, COPD, chronic renal failure, heart failure, arrhythmias, previous cardiac surgery, higher WBC, creatinine, longer CPB time, and CPB time $>$120 min); PMV was associated with increased in-hospital mortality.

PMV is a well-recognized complication of cardiac surgery, with an incidence ranging from 1.96% to 53.27% [12, 14]. The observed variation in reported incidences of PMV may be attributed to the lack of standardized definitions across studies. To address this issue and enhance the homogeneity of the study population, we employed a uniform definition of PMV as a duration equal to or exceeding 24 hours. By implementing this consistent criterion, we aimed to mitigate potential discrepancies in the identification and classification of PMV cases, ensuring a more reliable and comparable analysis of relevant outcomes. CVD patients are typically extubated within 6 h of surgery; 24 h is sufficient to stabilize their hemodynamics and eliminate the negative effects of surgery and CPB [25, 39]. Concurrently, PMV $\ge$24 h also aligns with current clinical practice guidelines recommending early extubation [40]. Of paramount significance is the notable reduction in the potential for underestimating the incidence of PMV.

The pooled incidence of PMV among patients who underwent cardiac surgery was determined to be 20%. Subgroup analysis, according to the study population, revealed that the highest incidence of PMV was 39.4% in aortic surgical patients (patients with aortic dissection and aortic arch disease). Conversely, the lowest incidence of PMV was observed in patients undergoing CABG combined with valve surgery, with a rate of 5.6%. In the subgroup analysis of PMV definitions, when PMV was defined as a duration of 72 hours or longer, the incidence was higher at 27.3%, compared to definitions of 24 h (16.1%) and 48 h (23.0%). Notably, the PMV $\ge$72-hour group primarily consisted of patients undergoing aortic surgery and those with a left ventricular assist device (LVAD), constituting the main study subjects within this particular subgroup. Aortic dissection is a catastrophic cardiovascular clinical event associated with high mortality and surgical risk [41]. Compared to other CVD patients, patients undergoing aortic surgery are more prone to postoperative complications and ischemia-reperfusion injury, which leads to perioperative hemodynamic instability [42, 43, 44]. In the case of patients with LVAD, in addition to the severity of the heart failure syndrome, the considerable burden of comorbidities places them at a high risk of postoperative complications [7]. Furthermore, most studies in the PMV $\ge$72 h group had small sample sizes, which might have resulted in an imprecise assessment of the incidence of PMV. The PMV $\ge$24 h group had the lowest incidence of PMV. Despite the fact that the study population of the six included publications originated from developing countries, we posit that defining PMV as a duration of 24 h or more neither overestimates nor underestimates the incidence of PMV. This assertion is based on several factors, including the substantial sample sizes employed in each study, the current contemporary intensive care management practices, and the notable advancements in clinical care and prognoses observed among patients in developing countries.

Evaluating the risk factors associated with PMV in patients undergoing cardiac surgery is essential for improving patient prognosis. In our synthesis, we identified advanced age, being female, and BMI as demographic characteristics that exhibited a significant association with PMV and served as risk factors. Patients who developed PMV in the included articles were predominantly over 60 years of age. Advanced age would reduce functional reserves and increase comorbidities, which have been associated with PMV. The assessment of other surgical patients provided the same conclusion [45, 46]. Although the relationship between PMV and women continues to receive widespread attention, the findings remain controversial. Epidemiological studies have consistently provided evidence supporting the protective role of estrogen against CVD in women. Notably, before menopause, the incidence of coronary heart disease is lower in women compared to men. This observation is attributed, at least in part, to the beneficial effects of estrogen on various cardiovascular parameters, including lipid metabolism, vascular function, and inflammation [47]. Therefore, we speculated that there may have been a selection bias. In addition, being female has been identified as a possible risk factor for intensive care unit-acquired weakness, which causes diaphragmatic weakness and prolongs the duration of mechanical ventilation [48, 49]. Whether a higher BMI is associated with PMV has been debated. Our results suggest a higher BMI and BMI $>$28 kg/m² are risk factors. However, the evidence for a higher BMI was low; only three publications mentioned a BMI $>$28 kg/m². Coincidentally, the latest meta-analysis on the impact of BMI on patients undergoing cardiac surgery concluded that being underweight was a predictor of worse survival outcomes, whereas a slightly higher BMI was a protective factor [50]. Therefore, further investigation is required to determine whether BMI causally affects PMV. It is imperative to investigate the incidence of PMV across different weight categories, including underweight, normal weight, overweight, and obese patients undergoing cardiac surgery. In summary, a comprehensive evaluation of patient demographics and characteristics before surgery is important to create an individualized care plan and manage the airway, which reduces PMV.

The influence of heart conditions cannot be ignored. Ejection fraction 50, NYHA $\ge$III, heart failure, and arrhythmias were also preoperative risk factors for PMV. A lower EF has a negative impact on the occurrence of PMV, which results in higher ventricular filling pressures and lower cardiac output that cause hemodynamic instability and postoperative complications [13, 51]. However, NYHA $\ge$III, heart failure, and arrhythmias showed low evidence and high heterogeneity. When we performed subgroup analysis based on the PMV definition, the heterogeneity was significantly lower in NYHA $\ge$III than before. Moreover, the OR was higher in the PMV $\ge$24 h group than in the PMV $\ge$48 h group; thus, defining PMV as $\ge$24 h may help identify risk factors more adequately. Furthermore, preoperative optimization of a patient’s cardiac condition plays an important role in preventing the adverse effects of PMV.

In addition, we confirmed that the patient’s medical history and information on other diseases are crucial for preventing PMV. COPD, chronic renal failure, previous cardiac surgery, higher WBC, and creatinine were proved. Comorbidities such as renal injury and pulmonary hypertension are worthy of attention as they increase the risk of severe cardiac dysfunction and limit physiologic reserve [52, 53]. Cardiac patients, such as those with renal impairment, often have other comorbidities (e.g., diabetes and hypertension) that increase their mental, physiological, and economic burden. The same conclusion has been reported in previous studies [54, 55]. While our analysis suggested that hypertension and diabetes mellitus do not significantly elevate the risk of PMV, it is crucial to acknowledge that the evidence supporting this conclusion was of low quality. Consequently, further research with higher-quality evidence is warranted to establish the relationship between these comorbidities and PMV risk definitively.

Regarding intraoperative risk factors for PMV, longer CPB times and CPB $>$120 min were identified. The same results were obtained when different study populations and PMV definitions were used for the subgroup analyses. Extended CPB duration precipitates a cascade of events characterized by a heightened pro-inflammatory response, release of diverse inflammatory cytokines, and activation of the complement system. These processes collectively contribute to an augmented risk of ischemia-reperfusion injury and hemolysis, leading to the destruction of blood cells [15, 56]. In such cases, patients are more prone to pulmonary complications that cause PMV. Nevertheless, many other factors affect CPB time, such as surgeon proficiency and patient condition [57]. The management of CPB time often presents a challenge for doctors and patients, and it is important to improve treatment management, the environment, and patients’ physical and mental conditions.

Finally, we examined the relationship between PMV and in-hospital mortality. We found cardiac surgical patients with PMV had a 14-fold increase in in-hospital mortality compared with that in patients without PMV. Subgroup analyses based on different study populations and PMV definitions revealed that in-hospital mortality significantly increased among PMV patients. Moreover, the relationship between long-term mortality and PMV should be addressed. A large analysis of critically ill patients receiving at least 48 hours of mechanical ventilation showed a 28-day mortality rate of 26.3% [58]. Due to limited evidence, we only analyzed in-hospital mortality. Future studies should explore the effects of PMV on long-term mortality.

We conducted a comprehensive search and rigorous screening of studies for inclusion. NOS was also used to assess the quality. Since the included studies had similar but different clinical settings, we performed subgroup analyses of different study populations and PMV definitions. Given the presence of identical yet distinct clinical settings among the included studies, we conducted subgroup analyses to examine the effects of diverse study populations and variations in PMV definitions. However, our synthesis had some limitations. Firstly, our analysis included only English-language publications, which might have partly caused us to miss relevant non-English publications. Secondly, the number of studies included in each analysis was relatively small. Finally, an exhaustive analysis of other PMV risk factors, such as smoking history and operative time, was not feasible due to their absence or limited coverage in the original studies. Future large-scale clinical trials are required to validate the relationship between those unidentified risk factors and the PMV. In addition, further research should be conducted to determine the weights of these identified risk factors to make them more clinically feasible.

The incidence of PMV was 20%. Fifteen risk factors were associated with a heightened risk of PMV. However, the potential impacts of higher BMI, heart failure, and arrhythmias require further exploration. Our synthesis also revealed that PMV was significantly associated with increased in-hospital mortality in patients undergoing cardiac surgery. Moreover, we support the definition of PMV as a duration greater than or equal to 24 hours, as it would enable earlier identification of cases with PMV. The findings of our synthesis suggest that the timely identification of risk factors for PMV and prompt recognition and intervention are vital for enhancing patient prognosis. Therefore, future PMV management should better emphasize these identified risk factors.

All data points generated or analyzed during this study are included in this article and there are no further underlying data necessary to reproduce the results.

QW, YT, XZ—designed the study, wrote manuscript; SX, YP, LL—conducted literature search, appraisal of study quality, reviewed and revised the manuscript; LC, YL—revised the manuscript, examined study design and findings. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

This study was supported by grant from Fujian provincial health technology project (2021CXA017).

The authors declare no conflict of interest.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/j.rcm2511409.