Abstract
Background
Postoperative fever is common after aortic surgery and often leads to concerns about infection. However, fever may also arise from noninfectious inflammatory responses. The aim of this study was to identify the risk factors and implications of noninfectious fever in patients undergoing surgery for acute type A aortic dissection (AD).
Methods
This retrospective analysis examined 636 patients who underwent aortic replacement for acute type A AD between 2003 and 2024. Patients were excluded if they had pre- or postoperative infections. The remaining 545 patients were divided into a “Non-fever” group and a “Noninfectious fever” group. Patient demographics, surgical details, and outcomes were compared. Multivariable logistic regression was performed to identify independent predictors of noninfectious fever.
Results
Patients with noninfectious fever were younger (63.3±14.8 vs. 55.7±16.0 years, P<0.001) and had longer cardiopulmonary bypass (CPB) and operation times (P=0.005 and P=0.04, respectively). In the multivariable analysis, independent risk factors for noninfectious fever were age [odds ratio (OR): 0.973, 95% confidence interval (CI): 0.959–0.988, P<0.001], longer CPB time (OR: 1.008, 95% CI: 1.003–1.012, P=0.002), and complete false lumen thrombosis (OR: 2.169, 95% CI: 1.074–4.007, P=0.01). Noninfectious fever was not associated with increased risk of 30-day mortality (OR: 0.537, 95% CI: 0.222–1.297, P=0.17).
Conclusions
Younger age, longer CPB time, and complete false lumen thrombosis were independent risk factors for noninfectious postoperative fever following type A AD surgery. Recognizing this benign response could help to avoid unnecessary testing and antibiotics, which would improve postoperative care and resource utilization.
Keywords: Aortic dissection (AD), false lumen thrombosis, noninfectious fever, postoperative complications
Highlight box.
Key findings
• Younger age, longer cardiopulmonary bypass (CPB) time, and complete false lumen thrombosis were independent predictors of noninfectious postoperative fever after acute type A aortic dissection (AD) surgery.
• Noninfectious postoperative fever did not increase 30-day mortality, intensive care unit stay, hospital stay, or readmission rates, indicating a benign and self-limiting inflammatory response.
What is known and what is new?
• Postoperative fever is common after major surgery and often triggers concern for infection, leading to extensive diagnostic testing and antibiotic use. In many surgical fields, early postoperative fever is frequently noninfectious and associated with inflammatory responses rather than true infection.
• This study provides a comprehensive evaluation of noninfectious postoperative fever in a large cohort of patients undergoing surgery for acute type A AD.
• The study identifies distinct risk factors and clarifies that noninfectious fever does not adversely affect early postoperative outcomes.
What is the implication, and what should change now?
• Clinicians should recognize that postoperative fever in this population is often noninfectious and benign, especially in younger patients or those with long CPB times or complete thrombosis.
• Once infectious etiologies are reasonably excluded, unnecessary diagnostic tests and antibiotic therapy should be avoided, supporting better resource utilization and antimicrobial stewardship.
• Adoption of more structured evaluation algorithms for postoperative fever may reduce overtreatment and improve postoperative management after AD surgery.
Introduction
Background
Postoperative fever is a frequent occurrence after surgery, with prevalence rates ranging from 14% to 74% depending on the procedure and reporting institution (1-4). Postoperative fever may suggest infection but can also result from benign inflammatory responses (1,5). Distinguishing between infectious and noninfectious causes remains challenging and often leads to unnecessary diagnostic tests, longer hospital stays, increased healthcare costs, and excessive antibiotic use (2,3,5,6).
These issues are particularly important in aortic dissection (AD) surgery, where misinterpreting fever can lead to serious consequences (7). Among patients undergoing surgery for acute type A AD, postoperative fever may indicate severe complications such as pneumonia, graft infection, mediastinitis, or thrombotic events like stroke and visceral infarction (8,9). As in many other surgical contexts, noninfectious postoperative fever also occurs in cases of AD, which poses challenges for decision making and raises the risk of overtreatment.
Rationale and knowledge gap
Noninfectious postoperative fever has been studied in general and orthopedic surgeries and has often been found to be more common in younger and obese patients. However, there are limited data about its prevalence and risk factors in the context of AD surgery (4,10-12). Given the complexity and risks associated with these procedures, better understanding of noninfectious fever in this context would help to guide postoperative management, avoid unnecessary interventions, and optimize resource utilization.
Objective
Therefore, the aim of this study is to compare patients who underwent aortic surgery for acute type A AD and developed noninfectious fever to those who did not experience postoperative fever. Key risk factors and clinical outcomes were evaluated to facilitate more accurate identification of noninfectious fever, thereby supporting improved postoperative management in this high-risk population. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2439/rc).
Methods
The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Seoul National University Bundang Hospital (No. B-2403-891-102). The consent requirement was waived due to the retrospective nature of the study.
Participant enrollment
The study initially examined 636 patients who underwent aortic replacement surgery for acute type A AD at our center from September 2003 to September 2024. The focus was solely on the relationship between aortic surgery and fever, so patients who were diagnosed with active infection prior to surgery were excluded. Patients were also excluded if they developed infections with identifiable foci postoperatively.
Diagnostic laboratory tests and medical records were reviewed to identify infection foci, including urine, sputum, blood, and wound swab cultures, as well as chest X-rays and computed tomography (CT) examinations. Any documented evidence of infection in the medical records was also considered. Pneumonia was diagnosed based on lung infiltrates on imaging or two positive cultures of tracheal tube aspirate samples. Mediastinitis, complicated fluid collections, and graft infections were defined according to the CT that prompted surgery, intraoperative findings, or positive cultures intraoperatively. Diagnosis of bacteremia, wound infection, and urinary tract infection required two positive cultures of the same organism.
Other causes included pressure ulcers, colitis, cellulitis, herpes zoster, and acute cholecystitis. Pressure ulcers were defined based on two positive swab cultures. Colitis was diagnosed based on active symptom and physical examination. A diagnosis of cellulitis was made when it was clinically confirmed and documented in the medical records with the affected area specified by the attending physician. Herpes zoster was diagnosed based on confirmation by a dermatologist.
After exclusion, the remaining patients were grouped according to the presence of postoperative fever, which was defined as a body temperature greater than 38.0 ℃ for two consecutive days or a single occurrence of a temperature exceeding 38.5 ℃. Those without fever events were assigned to the “Non-fever” group, while who experienced postoperative fever were assigned to the “Noninfectious fever” group. The sample size was determined by including all consecutive eligible patients during the study period. Due to the retrospective character of the study, no formal sample-size calculation was performed.
This study adhered to standard diagnostic criteria and rigorous review processes to ensure accurate classification of infectious and noninfectious postoperative fever cases.
Surgical strategy
All patients underwent surgery via median sternotomy. Our operative strategy is based on a tear-oriented surgery. Total or partial arch replacement was actively performed when an arch aneurysm was present or when the primary intimal tear was located within the aortic arch. Likewise, aortic root replacement was aggressively pursued in patients with a concomitant root aneurysm or when the dissection extended toward the aortic root.
For patients undergoing hemiarch replacement, the distal anastomosis was performed under hypothermic circulatory arrest, followed by approximately 3 minutes of brief retrograde cerebral perfusion (RCP), after which cardiopulmonary bypass (CPB) was resumed. In cases requiring total or partial arch replacement, hypothermic circulatory arrest was initiated to secure the arch vessels and distal anastomotic site, followed by approximately 3 minutes of brief RCP, after which bilateral selective antegrade cerebral perfusion was immediately established using balloon-tipped catheters.
Cerebral oximetry using near-infrared spectroscopy was performed in all patients to continuously monitor intraoperative cerebral oxygenation. In most cases, the sequence of vascular anastomoses was distal, proximal, and subsequently supra-aortic vessel reconstruction. During surgery, the core body temperature was reduced to 25 ℃, and rewarming was initiated after completion of the distal anastomosis. Core temperature was meticulously monitored using bladder or rectal probes, supplemented by esophageal temperature monitoring. For myocardial protection during cardiac arrest, cold blood cardioplegia or Del Nido cardioplegia was administered at the surgeon’s discretion.
Postoperative antibiotic use and decision-making
During postoperative course, decision to initiate antibiotic therapy was made at the discretion of the attending physician, based on the patient’s febrile status, laboratory findings, and consultation with the infectious disease service when indicated. Antibiotic use was determined in a relatively liberal, patient-tailored manner according to overall clinical condition, and no strict institutional protocol was uniformly applied.
Data collection
Most of the demographic characteristics were obtained from our prospectively collected database. Medical records were reviewed to determine the presence of any preoperative malperfusion or cardiac tamponade, intraoperative data, and postoperative treatment details. The intraoperative data included the extent of surgery, root replacement, use of conventional elephant trunk (CET) or frozen elephant trunk (FET) technique, graft type, CPB time, aortic cross clamp (ACC) time, and operation time. Postoperative variables comprised peak white blood cell (WBC) count, peak C-reactive protein (CRP), chest tube drainage volume within the first 24 hours, chest tube duration days, length of intubation, length of stay in the intensive care unit (ICU), length of hospital stay, and postoperative complications. Tympanic temperatures were measured at 4- or 8-hour intervals in the general ward depending on the patient’s condition, as well as every hour during ICU stay. All postoperative body temperature measurements were downloaded from the health information system and reviewed to clean up the data. 30-day readmission, 30-day mortality and overall mortality were also recorded.
CT scans were typically performed during the 5th to 7th days postoperatively. All scans were reviewed to assess the extent of false lumen thrombosis. The CT reviewer was blinded to the group assignments. Based on the extent of thrombosis, patients were divided into three groups: “patent” (no thrombosis), “partial thrombosis” (proximal to left upper pulmonary vein), and “complete thrombosis” (distal to left upper pulmonary vein with full thickness on axial view). Patients without residual false lumen, such as DeBakey type II AD or intramural hematoma, were assigned to the “patent” group due to a lack of evidence of newly formed thrombosis.
Statistical analysis
Statistical analyses were conducted to compare “Non-fever” group and “Noninfectious fever” group. Continuous variables were analyzed using an independent sample t-test, while categorical variables were assessed using either Pearson’s Chi-squared test or Fisher’s exact test as appropriate. Continuous variables were used in either a continuous or categorical form based on their data distribution, the goodness-of-fit of the overall model, and the existence of clinically relevant threshold values. Data were expressed as the mean with the standard deviation for continuous variables and as a frequency and percentage for categorical variables.
To identify independent risk factors associated with noninfectious postoperative fever, a multivariable logistic regression analysis was performed with noninfectious fever as the dependent variable. Variables with a P value of less than 0.20 in the univariable analyses were considered candidates for inclusion in the multivariable model. After excluding clinically redundant or highly correlated variables, backward selection process was then applied to identify variables with statistical significance and evaluate the independent association of potential risk factors with noninfectious postoperative fever.
An additional multivariable analysis was performed with 30-day mortality as the dependent variable to investigate whether the postoperative course differed significantly between patients with and without noninfectious fever. Among variables with a P value of less than 0.20 in the univariable analyses, clinically redundant or highly correlated variables were excluded based on clinical relevance and model stability. Noninfectious fever was then added to the remaining variables to comprehensively assess its independent association with 30-day mortality.
All statistical analyses were two-sided, a P value <0.05 was considered statistically significant, and all analyses were conducted using SPSS 30.0 software (IBM SPSS, Armonk, NY, USA). No imputation was performed because key variables had no missing data. All demographic, operative, and postoperative variables were available from the institutional database, and temperature measurements were reviewed and cleaned before analysis.
Results
Patient demographics
Three patients who were diagnosed with active infection prior to surgery were excluded, including 1 patient with active tuberculosis and 2 patients with coronavirus disease 2019 (COVID-19). The “Non-fever” group comprised 423 (66.5%) patients without fever. Among the 210 (33.0%) patients who met the fever criteria, 88 (13.8%) patients were excluded due to identifiable infection. There were 61 cases of pneumonia, 27 cases of bacteremia, 11 cases of wound infection or mediastinitis, 6 cases of complicated fluid collection or graft infection, 4 cases of urinary tract infection, 5 cases of colitis, 3 cases of pressure ulcer infection, 2 cases of cellulitis, 2 cases of acute cholecystitis, and 1 case of herpes zoster. Multiple infections in an individual patient were counted separately. Ultimately, a total of 122 (19.2%) patients were assigned to the “Noninfectious fever” group. In this group, the mean day of fever onset was postoperative day 4.1±3.2. Flow diagram of the patient grouping is presented in Figure 1.
Figure 1.
Patient classification flowchart for study cohort. A total of 636 patients who underwent surgery for acute type A aortic dissection were included in the study. Three patients with preoperative active infections were excluded. Among the remaining 633 patients, 88 had confirmed infections and were excluded. Among remaining 545 patients, 423 without postoperative fever were assigned to the “Non-fever” group, and 122 patients with postoperative fever but no identifiable infection were assigned to “Noninfectious fever” group. Multiple infections occurring in a single patient were counted as separate events in the infection subcategories.
Preoperative patient demographics are summarized in Table 1. Males were more prevalent in the “Noninfectious fever” group (48.7% vs. 59.8%, P=0.03). The mean age was lower in the “Noninfectious fever” group (63.3±14.8 vs. 55.7±16.0 years, P<0.001). The two groups had similar rates of underlying medical conditions, including hypertension, diabetes mellitus, hypercholesterolemia, cerebrovascular accident, chronic renal failure, and history of coronary artery disease (CAD). Marfan syndrome was more prevalent in the “Noninfectious fever” group (6.4% vs. 13.9%, P=0.02). The incidence of any preoperative malperfusion was comparable between the two groups, and this finding remained consistent across subgroups of arch vessel, coronary artery, or lower extremity malperfusion. However, visceral malperfusion was more prevalent in “Noninfectious fever” group (7.1% vs. 13.1%, P=0.04). The incidence of preoperative cardiac tamponade was also comparable between the two groups.
Table 1. Baseline patient and disease characteristics in non-fever and noninfectious fever groups.
| Characteristics | Non-fever (n=423) | Noninfectious fever (n=122) | P value |
|---|---|---|---|
| Patient characteristics | |||
| Sex (male) | 206 (48.7) | 73 (59.8) | 0.03 |
| Age (years) | 63.3±14.8 | 55.7±16.0 | <0.001 |
| Hypertension | 295 (69.7) | 74 (60.7) | 0.06 |
| Diabetes mellitus | 50 (11.8) | 10 (8.2) | 0.26 |
| Hypercholesterolemia | 67 (15.8) | 18 (14.8) | 0.77 |
| Cerebrovascular accident | 60 (14.2) | 11 (9.0) | 0.14 |
| Chronic renal failure | 23 (5.4) | 6 (4.9) | 0.82 |
| History of CAD | 41 (9.7) | 5 (4.1) | 0.050 |
| Marfan syndrome | 27 (6.4) | 16 (13.1) | 0.02 |
| Disease characteristics | |||
| Any preoperative malperfusion | 98 (23.2) | 33 (27.0) | 0.38 |
| Arch vessel malperfusion | 51 (12.1) | 12 (9.8) | 0.50 |
| Coronary artery malperfusion | 9 (2.1) | 3 (2.5) | 0.83 |
| Visceral malperfusion | 30 (7.1) | 16 (13.1) | 0.04 |
| Peripheral malperfusion | 30 (7.1) | 12 (9.8) | 0.32 |
| Preoperative cardiac tamponade | 83 (19.6) | 26 (21.3) | 0.68 |
Data are presented as n (%) for categorical variables and mean with standard deviation for continuous variables. CAD, coronary artery disease.
The operative characteristics are shown in Table 2. There was no statistically significant difference between two groups in the extent of surgery, which was divided into three categories: ascending or hemiarch replacement, partial arch replacement, and total arch replacement. Aortic root replacement was more frequently performed in the “Noninfectious fever” group than the “Non-fever” group (18.7% vs. 28.7%, P=0.02). The use of CET or FET technique was comparable between the two groups. CPB time, ACC time and operation time were longer in the “Noninfectious fever” group (CPB time: 160±47 vs. 176±57 min, P=0.005; ACC time: 110±42 vs. 125±53 min, P=0.001; operation time: 292±72 vs. 310±90 min, P=0.04).
Table 2. Operative characteristics in non-fever and noninfectious fever groups.
| Operative characteristics | Non-fever (n=423) | Noninfectious fever (n=122) | P value |
|---|---|---|---|
| Surgical extent | 0.29 | ||
| Ascending, hemiarch replacement | 140 (33.1) | 33 (27.0) | – |
| Partial arch replacement | 127 (30.0) | 35 (28.7) | – |
| Total arch replacement | 156 (36.9) | 54 (44.3) | – |
| Conventional elephant trunk | 87 (20.6) | 32 (26.2) | 0.18 |
| Frozen elephant trunk | 35 (8.3) | 15 (12.3) | 0.18 |
| Root replacement | 79 (18.7) | 35 (28.7) | 0.02 |
| Gelatin-impregnated graft | 85 (20.1) | 18 (14.8) | 0.18 |
| CPB time (min) | 160±47 | 176±57 | 0.005 |
| ACC time (min) | 110±42 | 125±53 | 0.001 |
| Operation time (min) | 292±72 | 310±90 | 0.04 |
| Transfusion | |||
| Intraoperative RBC (pkg) | 7.6±3.8 | 7.0±3.7 | 0.10 |
| Postoperative 24 h RBC (pkg) | 9.8±8.6 | 11.3±8.4 | 0.09 |
Data are presented as n (%) for categorical variables and mean with standard deviation for continuous variables. ACC, aortic cross clamp; CPB, cardiopulmonary bypass; RBC, red blood cell.
The postoperative outcomes are summarized in Table 3. The distribution of false lumen thrombosis differed significantly between the two groups (P=0.01). Complete thrombosis was observed in 9.9% of the “Non-fever” group and 17.2% of the “Noninfectious fever” group. Chest tube drainage volume within the first 24 hours and chest tube duration days were comparable between two groups. Peak WBC counts were also similar; however, peak CRP levels were higher in the “Noninfectious fever” group (14.8±8.7 vs. 16.6±6.9 mg/dL, P=0.045). Additionally, no significant differences were observed between the groups in terms of intubated hours, length of ICU stay, 30-day readmission rate, or 30-day mortality.
Table 3. Postoperative course comparison between non-fever and noninfectious fever groups.
| Postoperative course | Non-fever (n=423) | Noninfectious fever (n=122) | P value |
|---|---|---|---|
| False lumen thrombosis | 0.01 | ||
| Patent | 278 (65.7) | 64 (52.5) | |
| Partial thrombosis | 63 (14.9) | 28 (23.0) | |
| Complete thrombosis | 42 (9.9) | 21 (17.2) | |
| Peak WBC count (×103) | 18.0±5.3 | 18.5±5.3 | 0.45 |
| Peak CRP level (mg/dL) | 14.8±8.7 | 16.6±6.9 | 0.045 |
| Postop 24 h chest tube drain (mL) | 1,599±2,028 | 1,838±1,552 | 0.23 |
| Chest tube duration (days) | 5.2±3.2 | 5.6±2.9 | 0.16 |
| Intubated time (hours) | 37.9±83.4 | 48.7±86.9 | 0.21 |
| ICU stay (hours) | 93.1±158.5 | 100.3±107.4 | 0.48 |
| Hospital stays (days) | 15.4±19.7 | 15.2±9.4 | 0.90 |
| Postop new onset dialysis | 28 (6.6) | 9 (7.4) | 0.77 |
| 30 days readmission | 19 (4.5) | 3 (2.5) | 0.44 |
| 30 days mortality | 34 (8.0) | 7 (5.7) | 0.40 |
Data are presented as n (%) for categorical variables and mean with standard deviation for continuous variables. CRP, C-reactive protein; ICU, intensive care unit; Postop, postoperative; WBC, white blood cell.
Risk factor analysis
The univariable analyses identified several risk factors associated with noninfectious fever (Table 4). Based on the results of univariable analyses and exclusion of clinically redundant variables, the following characteristics were included in the multivariable analysis: sex, age, hypertension, cerebrovascular accident, history of CAD, Marfan syndrome, visceral malperfusion, CET or FET use, root replacement, Gelatin-impregnated graft use, CPB time, intraop red blood cell (RBC) transfusion, extent of false lumen thrombosis, peak CRP, and chest tube duration days. The final multivariable analysis identified the following as independent predictors of noninfectious fever: age [odds ratio (OR): 0.973, 95% confidence intervals (CIs): 0.959–0.988, P<0.001], CPB time (OR: 1.008, 95% CI: 1.003–1.012, P=0.002), and the presence of complete false lumen thrombosis (OR: 2.169, 95% CI: 1.174–4.007, P=0.01). The results of the multivariable logistic regression were illustrated in a forest plot (Figure 2A), which showed effect sizes and CI that were consistent with the values in Table 4.
Table 4. Univariable and multivariable analyses for risk factors associated with noninfectious postoperative fever.
| Variables | Univariable analysis | Multivariable analysis | |||
|---|---|---|---|---|---|
| P value | OR (95% CI) | P value | OR (95% CI) | ||
| Sex (male) | 0.03 | 1.569 (1.042, 2.363) | – | – | |
| Age | <0.001 | 0.968 (0.955, 0.981) | <0.001 | 0.973 (0.959, 0.988) | |
| Hypertension | 0.06 | 0.669 (0.440, 1.016) | – | – | |
| Diabetes mellitus | 0.26 | 0.666 (0.327, 1.356) | – | – | |
| Hypercholesterolemia | 0.77 | 0.920 (0.523, 1.617) | – | – | |
| Cerebrovascular accident | 0.14 | 0.600 (0.305, 1.180) | – | – | |
| Chronic renal failure | 0.82 | 0.900 (0.358, 2.262) | – | – | |
| History of CAD | 0.058 | 0.398 (0.154, 1.031) | – | – | |
| Marfan syndrome | 0.02 | 2.214 (1.151, 4.260) | – | – | |
| Any preoperative malperfusion | 0.38 | 1.230 (0.777, 1.946) | – | – | |
| Arch vessel malperfusion | 0.500 | 0.796 (0.410, 1.545) | – | – | |
| Coronary artery malperfusion | 0.83 | 1.160 (0.309, 4.352) | – | – | |
| Visceral malperfusion | 0.04 | 1.977 (1.039, 3.763) | – | – | |
| Peripheral malperfusion | 0.32 | 1.429 (0.708, 2.884) | – | – | |
| Preoperative tamponade | 0.68 | 1.109 (0.676, 1.821) | – | – | |
| Surgical extent | |||||
| Ascending, hemiarch | Ref | Ref | |||
| Partial arch | 0.57 | 1.169 (0.686, 1.992) | – | – | |
| Total arch | 0.12 | 1.469 (0.900, 2.396) | – | – | |
| Conventional elephant trunk | 0.18 | 1.373 (0.861, 2.191) | – | – | |
| Frozen elephant trunk | 0.18 | 1.554 (0.818, 2.952) | – | – | |
| Root replacement | 0.02 | 1.752 (1.103, 2.781) | – | – | |
| Gelatin-impregnated graft | 0.19 | 0.688 (0.395, 1.198) | – | – | |
| CPB time | 0.002 | 1.006 (1.002, 1.010) | 0.002 | 1.008 (1.003, 1.012) | |
| ACC time | 0.002 | 1.007 (1.003, 1.011) | – | – | |
| Operation time | 0.02 | 1.003 (1.001, 1.006) | – | – | |
| Lowest temperature (rectal) | 0.74 | 1.013 (0.939, 1.092) | – | – | |
| Intraoperative RBC | 0.10 | 0.955 (0.903, 1.010) | – | – | |
| False lumen thrombosis | |||||
| Patent | Ref | Ref | Ref | Ref | |
| Partial thrombosis | 0.01 | 1.931 (1.146, 3.252) | 0.13 | 1.520 (0.992, 2.621) | |
| Complete thrombosis | 0.01 | 2.172 (1.204, 3.918) | 0.01 | 2.169 (1.174, 4.007) | |
| Peak WBC | 0.45 | 1.015 (0.977, 1.054) | – | – | |
| Peak CRP | 0.07 | 1.023 (0.998, 1.048) | – | – | |
| Postoperative 24 h chest tube drain | 0.26 | 1.000 (1.000, 1.000) | – | – | |
| Chest tube duration | 0.16 | 1.044 (0.983, 1.109) | – | – | |
| Intubated time | 0.22 | 1.001 (0.999, 1.003) | – | – | |
| ICU stay | 0.48 | 1.000 (0.999, 1.002) | – | – | |
| Hospital stays | 0.90 | 0.999 (0.988, 1.011) | – | – | |
| Postoperative new-onset dialysis | 0.77 | 1.124 (0.515, 2.450) | – | – | |
Odds ratio with 95% confidence intervals are provided alongside P values. ACC, aortic cross clamp; CAD, coronary artery disease; CI, confidence interval; CPB, cardiopulmonary bypass; CRP, C-reactive protein; ICU, intensive care unit; OR, odds ratio; RBC, red blood cell; WBC, white blood cell.
Figure 2.
Forest plots of multivariable logistic regression analyses. (A) Risk factors associated with noninfectious postoperative fever following surgery for acute type A aortic dissection. Younger age, longer CPB time, and complete false lumen thrombosis were identified as independent predictors. (B) Factors associated with 30-day mortality. Any preoperative malperfusion and CPB time were independently associated with increased 30-day mortality. ORs and 95% confidence intervals are shown for each variable. *, variables identified as independent predictors in the final model. CPB, cardiopulmonary bypass; OR, odds ratio.
Postoperative course analysis
The univariable analyses identified several risk factors associated with 30-day mortality (Table 5). Based on the results of univariable analyses, as well as exclusion of variables considering clinical relevance and model stability, multivariable logistic regression analysis was performed including any preoperative malperfusion, CPB time, and the presence of noninfectious fever. Any preoperative malperfusion and CPB time were associated with 30-day mortality (any preoperative malperfusion: OR 3.934, 95% CI: 2.026–7.638, P<0.001; CPB time: OR 1.010, 95% CI: 1.004–1.015, P<0.001). The occurrence of noninfectious fever was not associated with an increased risk of 30-day mortality (OR: 0.537, 95% CI: 0.222–1.297, P=0.17). The forest plot of the results (Figure 2B) showed effect sizes and CI that were consistent with the values in Table 5.
Table 5. Univariable and multivariable logistic regression analyses for risk factors associated with 30-day mortality.
| Variables | Univariable analysis | Multivariable analysis | |||
|---|---|---|---|---|---|
| P value | OR (95% CI) | P value | OR (95% CI) | ||
| No fever vs. noninfectious fever | 0.40 | 0.696 (0.301, 1.613) | 0.17 | 0.537 (0.222, 1.297) | |
| Sex | 0.51 | 1.238 (0.652, 2.350) | – | – | |
| Age | 0.73 | 1.004 (0.983, 1.025) | – | – | |
| Hypertension | 0.19 | 0.647 (0.338, 1.238) | – | – | |
| Diabetes mellitus | 0.10 | 0.189 (0.025, 1.397) | – | – | |
| Hypercholesterolemia | 0.14 | 0.406 (0.122, 1.348) | – | – | |
| Cerebrovascular accident | 0.43 | 1.415 (0.602, 3.327) | – | – | |
| Chronic renal failure | 0.20 | 2.071 (0.685, 6.268) | – | – | |
| History of CAD | 0.75 | 1.189 (0.404, 3.498) | – | – | |
| Marfan syndrome | 0.46 | 0.579 (0.135, 2.485) | – | – | |
| Any preoperative malperfusion | <0.001 | 3.372 (1.765, 6.444) | <0.001 | 3.934 (2.026, 7.638) | |
| Arch vessel malperfusion | 0.04 | 2.344 (1.062, 5.172) | – | – | |
| Coronary artery malperfusion | 0.003 | 6.703 (1.928, 23.298) | – | – | |
| Visceral malperfusion | 0.20 | 1.697 (0.750, 3.839) | – | – | |
| Peripheral malperfusion | 0.27 | 1.753 (0.649, 4.734) | – | – | |
| Surgery extent | – | – | |||
| Ascending, hemiarch | Ref | Ref | – | – | |
| Partial arch | 0.27 | 0.620 (0.263, 1.458) | – | – | |
| Total arch | 0.84 | 0.928 (0.449, 1.917) | – | – | |
| Conventional elephant trunk | 0.45 | 0.721 (0.311, 1.669) | – | – | |
| Frozen elephant trunk | 0.89 | 1.076 (0.367, 3.154) | – | – | |
| Root replacement | 0.87 | 1.069 (0.495, 2.309) | – | – | |
| Gelatin-impregnated graft | 0.47 | 0.719 (0.294, 1.758) | – | – | |
| CPB time | <0.001 | 1.010 (1.005, 1.015) | <0.001 | 1.010 (1.004, 1.015) | |
| ACC time | 0.40 | 1.003 (0.996, 1.009) | – | – | |
| Operation time | 0.002 | 1.006 (1.002, 1.009) | – | – | |
| Lowest temperature (rectal) | 0.52 | 0.962 (0.856, 1.082) | – | – | |
| False lumen thrombosis | – | – | |||
| Patent | Ref | Ref | – | – | |
| Partial thrombosis | 0.35 | 0.369 (0.047, 2.920) | – | – | |
| Complete thrombosis | 0.56 | 0.535 (0.067, 4.258) | – | – | |
Results of univariable and multivariable logistic regression analyses assessing potential risk factors for 30-day mortality. Odds ratio with 95% confidence intervals are provided alongside P values. ACC, aortic cross clamp; CAD, coronary artery disease; CI, confidence interval; CPB, cardiopulmonary bypass; OR, odds ratio.
Discussion
This retrospective study analyzed 545 patients who underwent emergency surgery for acute type A AD. Patients with noninfectious fever were compared to those without fever. Noninfectious postoperative fever was independently associated with younger age, longer CPB time, and complete false lumen thrombosis.
Younger age is a known risk factor for postoperative noninfectious fever in various types of surgery. Li et al. reported this association in patients with type B acute AD who underwent thoracic endovascular aneurysm repair (TEVAR) (11), while Garibaldi et al. and Walid et al. observed similar trends in general and spinal surgery (5,12). This may result from a stronger systemic inflammatory response in younger individuals. Dieleman et al. reported reduced inflammatory responses in cases of advanced age after cardiac surgery (13), which is consistent with age-related decline in immune responsiveness (14,15). A strong inflammatory stimulus results from tissue injury and CPB in aortic surgery, which may make younger patients more susceptible to noninfectious fever (16-18).
The association of longer CPB time with noninfectious fever is consistent with previous studies. Clark et al. reported that patients undergoing on-pump coronary artery bypass grafting (CABG) experienced more hyperthermia than those undergoing off-pump CABG, which suggested that CPB has a role in the development of fever (19). CPB-induced systemic inflammation likely has similar contributions in aortic surgery (16,17,20). Prolonged CPB time also reflects longer operation duration, which itself has been associated with postoperative fever (11).
The impact of false lumen thrombosis was a key focus of our analyses. A residual patent false lumen in AD is associated with aortic expansion and mortality, whereas complete thrombosis is associated with better long-term outcomes (21-24). Acute thrombosis has been frequently associated with fever in various clinical contexts, and postoperative false lumen thrombosis detected in CT prior to discharge may represent a similar process (25-27). In our study, the incidence of noninfectious fever did not differ significantly between patients with patent and partially thrombosed false lumens. However, patients with complete thrombosis had a significantly higher rate of noninfectious fever, which suggests that the extent of thrombosis may contribute to postoperative inflammatory responses.
Studies have presented varying findings on the clinical implications of noninfectious postoperative fever. Li et al. reported that patients who developed noninfectious fever after TEVAR had longer hospital stays (11), whereas Keijmel et al. found no significant difference in hospital stay or overall clinical course following aortic surgery (4). Discharging patients with unexplained fever remains challenging, and the decision often depends on institutional protocols or the preference of individual surgeons. In our study, we directly compared 30-day mortality, 30-day readmission rates, hospital stay, and ICU stay duration between patients with and without noninfectious fever. None of these parameters differed significantly between groups, which suggests that noninfectious fever does not adversely affect short-term outcomes.
These findings highlight the importance of distinguishing infectious fever from noninfectious fever as both groups showed comparable postoperative trajectories. These conditions often lead to unnecessary evaluation. Li et al. reported that patients developing fever after shoulder arthroplasty were often subjected to unnecessary tests (6), while Kendrick et al. similarly found that postoperative fever after gynecologic surgery yielded low diagnostic value despite high associated costs (2). In the context of aortic surgery, careful evaluation of postoperative fever is warranted, but once an infectious cause has been reasonably excluded, clinicians should avoid excessive diagnostic workups. This approach is especially relevant in cases of younger patients, longer CPB time, and complete false lumen thrombosis.
There are no standardized guidelines for antibiotic use specifically after aortic surgery, although the principles of antimicrobial stewardship principles emphasize selecting appropriate agents, optimizing dosage, and limiting treatment duration. Nevertheless, inappropriate or prolonged antibiotic use remains common (28-30). Recognizing noninfectious fever can help to prevent unnecessary antimicrobial therapy, which would improve postoperative management and reduce the risk of antibiotic resistance.
Several limitations of this study warrant consideration. First, the single-center retrospective design may limit the generalizability of the findings, particularly for institutions with different perioperative management protocols. There is no universally accepted criterion for defining postoperative fever, and we defined noninfectious fever based on our institution’s treatment threshold and the fact that tympanic measurement is standard in both ICU and general ward settings. However, the use of a uniform fever definition and standardized temperature monitoring would have improved the consistency of fever assessment. Secondly, although postoperative WBC and CRP levels were analyzed, the inclusion of additional inflammatory biomarkers and more comprehensive preoperative laboratory data might have enabled a more refined assessment of postoperative inflammatory responses. In addition, propensity score–based adjustment could have strengthened the robustness of the findings, but these analyses were limited by data availability. Furthermore, while preoperative imaging was available for all patients, the analysis did not explicitly distinguish newly developed postoperative false lumen thrombosis from pre-existing completely thrombosed false lumen or intramural hematoma, which may differ in their inflammatory profiles. This distinction should be addressed in future studies. Lastly, the postoperative management was not protocolized, and antibiotics therapy was frequently initiated based on clinical judgement in the early postoperative period. This may have influenced the recognition, duration, and classification of postoperative fever.
Despite these limitations, this study provides a comprehensive evaluation of noninfectious postoperative fever in a large cohort of patient undergoing surgery for acute type A AD. The inclusion of 636 patients provided a robust sample size for evaluating the incidence of fever and associated risk factors. The large cohort enabled the identification of key predictors of noninfectious fever and offers valuable insights into postoperative management. Prospective multicenter studies are warranted to validate our findings and optimize resource utilization.
Conclusions
This study highlights the distinct clinical features and postoperative course of noninfectious fever following surgery for acute type A AD. Younger age, longer CPB time, and complete false lumen thrombosis were identified as independent predictors. Importantly, noninfectious fever was not associated with adverse outcomes, which suggests that it is benign and self-limiting. These findings underscore the importance of recognizing noninfectious fever to avoid excessive testing and antibiotic use.
Supplementary
The article’s supplementary files as
Acknowledgments
None.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Seoul National University Bundang Hospital (No. B-2403-891-102). The consent requirement was waived due to the retrospective nature of the study.
Footnotes
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2439/rc
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2439/coif). The authors have no conflicts of interest to declare.
Data Sharing Statement
Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1-2439/dss
References
- 1.Abdelmaseeh TA, Azmat CE, Oliver TI. Postoperative Fever. StatPearls. Treasure Island (FL): StatPearls Publishing, Copyright © 2025, StatPearls Publishing LLC.; 2025. [Google Scholar]
- 2.Kendrick JE, Numnum TM, Estes JM, et al. Conservative management of postoperative Fever in gynecologic patients undergoing major abdominal or vaginal operations. J Am Coll Surg 2008;207:393-7. 10.1016/j.jamcollsurg.2008.04.001 [DOI] [PubMed] [Google Scholar]
- 3.Seo J, Park JH, Song EH, et al. Postoperative Nonpathologic Fever After Spinal Surgery: Incidence and Risk Factor Analysis. World Neurosurg 2017;103:78-83. 10.1016/j.wneu.2017.03.119 [DOI] [PubMed] [Google Scholar]
- 4.Keijmel SP, Zwartkruis IMM, Jongenotter J, et al. Postoperative Inflammation and Fever After Elective Aortic Valve and Aortic Root Replacement: A Retrospective Cohort Study. Open Forum Infect Dis 2023;10:ofad015. 10.1093/ofid/ofad015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Garibaldi RA, Brodine S, Matsumiya S, et al. Evidence for the non-infectious etiology of early postoperative fever. Infect Control 1985;6:273-7. 10.1017/s0195941700061749 [DOI] [PubMed] [Google Scholar]
- 6.Li M, Lyu C, Fang Y, et al. Characteristics and risk factors of fever after total joint arthroplasty: a single-center retrospective study. BMC Musculoskelet Disord 2022;23:979. 10.1186/s12891-022-05940-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gawinecka J, Schönrath F, von Eckardstein A. Acute aortic dissection: pathogenesis, risk factors and diagnosis. Swiss Med Wkly 2017;147:w14489. 10.4414/smw.2017.14489 [DOI] [PubMed] [Google Scholar]
- 8.Sayed A, Munir M, Bahbah EI. Aortic Dissection: A Review of the Pathophysiology, Management and Prospective Advances. Curr Cardiol Rev 2021;17:e230421186875. 10.2174/1573403X16666201014142930 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Trimarchi S, Nienaber CA, Rampoldi V, et al. Contemporary results of surgery in acute type A aortic dissection: The International Registry of Acute Aortic Dissection experience. J Thorac Cardiovasc Surg 2005;129:112-22. 10.1016/j.jtcvs.2004.09.005 [DOI] [PubMed] [Google Scholar]
- 10.García-Romo E, López-Medrano F, Llovet A, et al. Fever due to inflammation in acute aortic dissection: Description and proposals for diagnostic and therapeutic management. Rev Esp Cardiol 2010;63:602-6. 10.1016/s1885-5857(10)70123-7 [DOI] [PubMed] [Google Scholar]
- 11.Li X, Wu S, Huang Y, et al. Clinical Characteristics and Risk Factors of Noninfectious Fever after Thoracic Endovascular Aortic Repair of Acute Type B Aortic Dissection. Ann Vasc Surg 2023;91:145-54. 10.1016/j.avsg.2022.11.017 [DOI] [PubMed] [Google Scholar]
- 12.Walid MS, Woodall MN, Nutter JP, et al. Causes and risk factors for postoperative fever in spine surgery patients. South Med J 2009;102:283-6. 10.1097/SMJ.0b013e31819676a4 [DOI] [PubMed] [Google Scholar]
- 13.Dieleman JM, Peelen LM, Coulson TG, et al. Age and other perioperative risk factors for postoperative systemic inflammatory response syndrome after cardiac surgery. Br J Anaesth 2017;119:637-44. 10.1093/bja/aex239 [DOI] [PubMed] [Google Scholar]
- 14.Butcher SK, Lord JM. Stress responses and innate immunity: aging as a contributory factor. Aging Cell 2004;3:151-60. 10.1111/j.1474-9728.2004.00103.x [DOI] [PubMed] [Google Scholar]
- 15.Goronzy JJ, Weyand CM. Understanding immunosenescence to improve responses to vaccines. Nat Immunol 2013;14:428-36. 10.1038/ni.2588 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Day JR, Taylor KM. The systemic inflammatory response syndrome and cardiopulmonary bypass. Int J Surg 2005;3:129-40. 10.1016/j.ijsu.2005.04.002 [DOI] [PubMed] [Google Scholar]
- 17.Levy JH, Tanaka KA. Inflammatory response to cardiopulmonary bypass. Ann Thorac Surg 2003;75:S715-20. 10.1016/s0003-4975(02)04701-x [DOI] [PubMed] [Google Scholar]
- 18.Sönmez E, Jalalzai İ, Arslan Ü. Inflammation-Based Cell Ratios Beyond White Blood Cell Count for Predicting Postimplantation Syndrome After EVAR and TEVAR. Int J Mol Sci 2025;26:9753. 10.3390/ijms26199753 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Clark JA, Bar-Yosef S, Anderson A, et al. Postoperative hyperthermia following off-pump versus on-pump coronary artery bypass surgery. J Cardiothorac Vasc Anesth 2005;19:426-9. 10.1053/j.jvca.2005.01.042 [DOI] [PubMed] [Google Scholar]
- 20.Mitchell JD, Grocott HP, Phillips-Bute B, et al. Cytokine secretion after cardiac surgery and its relationship to postoperative fever. Cytokine 2007;38:37-42. 10.1016/j.cyto.2007.04.009 [DOI] [PubMed] [Google Scholar]
- 21.Fattori R, Bacchi-Reggiani L, Bertaccini P, et al. Evolution of aortic dissection after surgical repair. Am J Cardiol 2000;86:868-72. 10.1016/s0002-9149(00)01108-5 [DOI] [PubMed] [Google Scholar]
- 22.Halstead JC, Meier M, Etz C, et al. The fate of the distal aorta after repair of acute type A aortic dissection. J Thorac Cardiovasc Surg 2007;133:127-35. 10.1016/j.jtcvs.2006.07.043 [DOI] [PubMed] [Google Scholar]
- 23.Kimura N, Tanaka M, Kawahito K, et al. Influence of patent false lumen on long-term outcome after surgery for acute type A aortic dissection. J Thorac Cardiovasc Surg 2008;136:1160-6, 1166.e1-3. [DOI] [PubMed]
- 24.Tsai MT, Wu HY, Roan JN, et al. Effect of false lumen partial thrombosis on repaired acute type A aortic dissection. J Thorac Cardiovasc Surg 2014;148:2140-2146.e3. 10.1016/j.jtcvs.2014.02.003 [DOI] [PubMed] [Google Scholar]
- 25.Assouline B, Belli G, Dorgham K, et al. Fever following extracorporeal membrane oxygenation decannulation: Infection, thrombosis or just physiology? J Crit Care 2025;85:154945. 10.1016/j.jcrc.2024.154945 [DOI] [PubMed] [Google Scholar]
- 26.Barba R, Di Micco P, Blanco-Molina A, et al. Fever and deep venous thrombosis. Findings from the RIETE registry. J Thromb Thrombolysis 2011;32:288-92. 10.1007/s11239-011-0604-7 [DOI] [PubMed] [Google Scholar]
- 27.Stein PD, Afzal A, Henry JW, et al. Fever in acute pulmonary embolism. Chest 2000;117:39-42. 10.1378/chest.117.1.39 [DOI] [PubMed] [Google Scholar]
- 28.Ludwig KA, Carlson MA, Condon RE. Prophylactic antibiotics in surgery. Annu Rev Med 1993;44:385-93. 10.1146/annurev.me.44.020193.002125 [DOI] [PubMed] [Google Scholar]
- 29.Wang Q, Cao M, Tao H, et al. Evidence-based guideline for the prevention and management of perioperative infection. J Evid Based Med 2023;16:50-67. 10.1111/jebm.12514 [DOI] [PubMed] [Google Scholar]
- 30.Barlam TF, Cosgrove SE, Abbo LM, et al. Implementing an Antibiotic Stewardship Program: Guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis 2016;62:e51-77. 10.1093/cid/ciw118 [DOI] [PMC free article] [PubMed] [Google Scholar]