Excessive bleeding after cardiac surgery is a severe postoperative complication that is often accompanied by hemorrhagic shock and can occur in up to 12% of patients [1]. Postoperative bleeding has been associated with increased mortality, prolonged stay in the cardiac intensive care unit (CICU) and higher rates of sternal wound infection (SWI) [2, 3, 4]. Re-exploration for bleeding after open-heart surgery has been conventionally performed in the operating room (OR) except for patients in cardiac arrest who most often undergo surgery immediately in the CICU. Returning patients to OR may delay the operation and may result in additional risks to patients due to OR availability and the need for transportation.

Alternatively, conducting the re-exploration in the CICU allows for a more rapid procedure and can save both hospital and patient resources. However, controversies have been raised in conducting such surgery in the CICU due to the relative non-sterile environment [5]. Two previous reports supported the safety of performing chest re-exploration in the CICU [6, 7]. However, neither compared the postoperative outcomes to procedures performed in the OR. Furthermore, these two studies were limited to short-term outcomes and did not mention the long-term results of postoperative re-exploration conducted in the CICU. Therefore, the purpose of this study was to evaluate and compare outcomes of postoperative mediastinal re-explorations for bleeding following cardiac surgery conducted in the CICU versus the OR.

Seventy-two patients (1.4% of all screened patients) including 21 who received a re-operation in the CICU and 51 in the OR were eventually selected for further analysis. The mean age of selected patients was 60.0 $\pm$ 15.1 years. Fifty-two (72.2%) were male. Forty-six (63.9%) patients received an elective operation, 22 (30.6%) received an emergency operation (within 24 h of hospital admission with cardiac symptoms) and 4 (5.6%) received an urgent operation (upon hospital admission due to uncontrolled cardiac symptoms).

As presented in Table 1, there was no significant difference in baseline characteristics between two groups. The average patient age was slightly higher in the CICU group. More female patients underwent re-explorations in the OR. The incidence of hypertension, diabetes, and stroke, the preoperative left ventricular ejection fraction, and the use of anticoagulant therapy prior to surgery were comparable between the two groups. The average international normalized ratio was relatively higher in the OR group (1.1 $\pm$ 0.3 vs 1.0 $\pm$ 0.1; p = 0.016).

Table 2 summarizes the data from the initial cardiac surgery. Two (2.8%) patients received coronary artery bypass, 32 (44.4%) received valve operations, 8 (11.1%) received combined valve and bypass grafting, and 24 (33.3%) patients received aortic operations. The remaining 6 patients (8.3%) received congenital operations, pericardiectomy, or resection of a left ventricular aneurysm, respectively. Additional operative variables including mean cardiopulmonary bypass (CPB) time and cross-clamp time were comparable between two groups. The mean CPB time was significantly prolonged in the CICU group (p = 0.037).

Next, we examined and compared the intra-reoperation variables between two groups. Overall, the peak VIS was 47.0 $\pm$ 27.4 (51.1 $\pm$ 32.2 in the CICU group vs 44.9 $\pm$ 25.0 in the OR group, p = 0.472). 65 re-explorations (90.3%) were performed within 12 h of the primary operation. The median blood loss volume from initial chest closure to re-exploration was 1100 mL. Before re-exploration, the mean systolic blood pressure was 89.4 $\pm$ 9.6 mmHg (88.8 $\pm$ 13.4 in the CICU group vs 89.7 $\pm$ 7.3 in the OR group, p = 0.818), the mean arterial pressure was 67.1 $\pm$ 9.6 mmHg (63.9 $\pm$ 10.3 in the CICU group vs 68.7 $\pm$ 9.0 in the OR group, p = 0.122), and the mean central venous pressure was 12.1 $\pm$ 4.4 cmH${}_2$O (13.7 $\pm$ 5.4 in the CICU group vs 11.4 $\pm$ 3.6 in the OR group, p = 0.152). The average serum lactate concentration was 5.5 $\pm$ 4.1 mmol/L (6.7 $\pm$ 4.9 in the CICU group vs 5.0 $\pm$ 3.6 in the OR group, p = 0.234); 68 patients (94.4%) had lactate levels higher than 2 mmol/L. The mean hemoglobin level was 74.4 $\pm$ 15.4 g/L, and 51 patients (79.8%) had hemoglobin levels lower than 70 g/L (Table 3). No significant differences were found between groups. No patients required CPB support or delayed sternal closure. None of the patients in the CICU group were subsequently transferred to the OR for further re-exploration.

As shown in Table 4, there was no significant difference in post exploration laboratory test results between two groups. In addition, the average CICU stay and hospital stay was comparable between two groups. The incidence of post re-exploration adverse events including SWI, pneumonia, prolonged ventilation, AKI, new-onset hemodialysis, new-onset atrial fibrillation and tracheotomy were similar between two groups. Hospital costs were significantly lower in the CICU group. However, the cost for reoperation was not significantly different between the two groups. In addition, there was no significant difference in the 30-Day mortality between the CICU group and the OR group (14.3% vs 11.8%). Kaplan-Meier curves revealed no significant difference in 30-Day mortality between the two groups (log-rank p = 0.727, Fig. 1). After adjusting for confounders, the hazard ratios for reoperation conducted in the CICU (hazard ratios: 1.304, 95% confidence interval (CI): 0.325–5.232, p = 0.708) were not significantly associated with poor short-term survival.

Fig. 1.

Kaplan-Meier curves for 30-Day mortality in the two group. OR, operation room; CICU, cardiac intensive care unit.

In the univariate analysis, surgical status consisting of emergency and non-emergency, surgical procedure consisting of acute type A aortic dissection (ATAAD) surgery and non-ATAAD surgery were included in the analysis. Eventually, seven parameters were included in the multivariate logistic analysis model. The analysis suggested that emergent surgery, ATAAD surgery, initial cardiac surgery CPB time, and reopening operation time were independent risk factors for developing postoperative major complications. Nevertheless, reoperation conducted in the CICU (odds ratio: 0.958, 95% CI: 0.342–3.071, p = 0.806) was not identified as a risk factor for postoperative major complications (Table 5).

By January 2022, all patients had been followed for a median of 12 months. No SWI event was identified after discharge. Four patients from the OR group died during the follow-up period. However, Kaplan-Meier curves revealed no significant difference in cumulative survival rate between the groups (log-rank p = 0.768, Fig. 2). Multivariate Cox analysis for mortality revealed that reoperation conducted in the CICU (hazard ratios: 1.278, 95% CI: 0.224–6.697, p = 0.772) was not a significant risk factor after adjusting for other major clinical factors.

Fig. 2.

Kaplan-Meier curves for long-term mortality in the two group. OR, operation room; CICU, cardiac intensive care unit.

In this study, we demonstrated that conducting reoperations in the CICU did not result in additional risks such as SWI, hospital stay and mortality as compared to surgeries conducted in the OR. These results support an alternative approach when post-cardiac reoperation is required. To the best of our knowledge, this was the first contemporaneous study describing the outcomes of re-explorations for bleeding performed at different locations in the hospital.

Limited studies had been published to resolve the debate as to whether routine mediastinal re-explorations after cardiac operations can be safely conducted in the CICU. Most previous studies failed to compare the results to the conventional OR setting. As a result, the most efficacious strategy for mediastinal re-exploration remained unidentified. Conducting re-exploration in the CICU also has some advantages such as avoiding patient transfer under unstable hemodynamic conditions. More cardiac intensive care units now have sterile environments and the availability of additional life-saving equipment which are equivalent to that found in the OR. Our results showed that reoperations conducted in the CICU can be performed safely and effectively.

Recent studies suggested that the incidence of re-exploration for bleeding ranges from 2.2% to 5.9% [4, 5, 10, 11]. A total of 1.4% of patients were re-explored in this cohort, which is below the lower end of the range reported in the literature. However, previous results may have been confounded by the inclusion of reoperations performed for emergent conditions. In this study, we excluded patients who received mediastinal re-exploration due to cardiac arrest and cardiac tamponade, which resulted in a lower incidence of reoperations.

It is important to point out that the morbidity and mortality reported in this study for re-explorations was higher than those who received primary cardiac surgery. Based on our experience, this difference was likely attributed to the hemodynamic consequences of excessive bleeding rather than the re-exploration surgery itself. In this study, 5.6% of all patients experienced SWI and 12.5% of patients died after cardiac reoperations for bleeding; which is consistent with other studies [4, 6, 7, 10, 12, 13]. The lower incidence of SWI and mortality reported in some studies might be attributed to the exclusion of aortic dissection patients in their analyses. Reoperation procedures could aggravate the inflammatory response and lead to respiratory or renal dysfunction. The elevated reintubation and new onset dialysis rate after reoperation surgery in our study might be attributed to the augmented inflammatory response.

A primary concern for conducting re-explorations in the CICU is the fear of SWI; which remains a life-threatening complication after cardiac operations. Postoperative hemorrhage, prolonged operation and CPB times as well as hospital stay before the re-operation, internal mammary artery harvesting, immunocompromised states, and diabetes mellitus are considered as predisposing factors for SWI. Early postoperative re-exploration has also been identified as a predisposing factor for SWI [14]. Our data demonstrated that the occurrence of SWI was comparable between the CICU and the OR. This may due to the fact that we employed similar aseptic techniques in the CICU as in OR. Furthermore, only attending cardiac surgeons or senior trainees were eligible to conduct the re-exploration, accompanied by OR trained nursing staff. The mortality rates and occurrence of other postoperative complications were also comparable between two groups. Moreover, as suggested by the logistic regression analysis, the location where the re-exploration was performed was not an independent risk factor for major postoperative complications. This study showed that planned re-explorations conducted in the CICU are associated with comparable outcomes, similar to those that are performed in the OR for bleeding following cardiac surgery.

Hemorrhagic shock is one of the major causes of death in trauma patients [15], and is also commonly seen after cardiac surgery [16, 17, 18]. The main pathophysiological change in hemorrhagic shock is sudden reduction of effective circulating volume which leads to tissue hypoperfusion, increased anaerobic metabolism, lactic acidosis, reperfusion injury, endotoxin translocation, and ultimately leads to multiple organ dysfunction [19]. Rapid recognition, fluid resuscitation, and use of vasopressor drugs are essential in treating hypovolemic shock. A previous study indicated that patients received re-exploration for bleeding after cardiac surgery were at higher risk of experiencing adverse outcomes and this risk was further increased if the time to re-exploration was 12 h or longer [20]. Therefore, prompt re-exploration for bleeding which occurs after cardiac surgery is strongly recommended.

This study has some limitations. First, this was a retrospective study conducted in a single center with a small cohort. Second, the indication for re-exploration was not defined in advance. Third, the similar incidence of adverse events in two groups might be due to the limited number of patients which reduces the statistical power for risk factor analysis. Finally, relatively few patients received re-explorations in the CICU in this study sample (29.2%) which limited statistical modeling efforts and empiric data analysis. Therefore, further prospective multicenter studies are needed to better identify the most effective strategies to improve the prognosis of patients who undergo reoperation for bleeding following cardiac surgery.

In conclusion, our study found that planned re-exploration for bleeding after cardiac surgery can be safely and effectively conducted in the CICU. The CICU can serve as an alternative site to the OR to re-explore these high-risk patients.

OR, operation room; CICU, cardiac intensive care unit; SWI, sternal wound infection; VIS, Vasoactive-Inotropic Score.

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

DJW, ZGW, and YBK designed the research study. ZYW and JFX performed the research. ZGW, DDH, and LFZ analyzed the data and wrote the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

All subjects gave their informed consent for inclusion before participating in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Nanjing Drum Tower Hospital (No. BL2014004).

Not applicable.

This research received no external funding.

The authors declare no conflict of interest.