Early clinical outcomes of two regimens of prophylactic antibiotics in cardiac surgical patients with delayed sternal closure

Abstract

BACKGROUND

Delayed sternal closure (DSC) can be a lifesaving approach for certain patients who have undergone cardiac surgery. The value of the type of prophylactic antibiotics in DSC is still debatable.

AIM

To investigate clinical outcomes of different prophylactic antibiotic regimens in patients who had DSC after cardiac surgery.

METHODS

This was a retrospective observational single-center study. Fifty-three consecutive patients who underwent cardiac surgery and had an indication for DSC were included. Patients were subjected to two regimens of antibiotics: Narrow-spectrum and broad-spectrum regimens.

RESULTS

The main outcome measures were length of hospital and intensive care unit (ICU) stay, duration of mechanical ventilation, and mortality. Of the 53 patients, 12 (22.6%) received narrow-spectrum antibiotics, and 41 (77.4%) received broad-spectrum antibiotics. The mean age was 59.0 ± 12.1 years, without significant differences between the groups. The mean duration of antibiotic use was significantly longer in the broad-spectrum than the narrow-spectrum group (11.9 ± 8.7 vs 3.4 ± 2.0 d , P < 0.001). The median duration of open chest was 3.0 (2.0-5.0) d for all patients, with no difference between groups (P = 0.146). The median duration of mechanical ventilation was significantly longer in the broad-spectrum group [60.0 (Δ interquartile range (IQR) 170.0) h vs 50.0 (ΔIQR 113.0) h, P = 0.047]. Similarly, the median length of stay for both ICU and hospital were significantly longer in the broad-spectrum group [7.5 (ΔIQR 10.0) d vs 5.0 (ΔIQR 5.0) d, P = 0.008] and [27.0 (ΔIQR 30.0) d vs 19.0 (ΔIQR 21.0) d, P = 0.031]. Five (9.8%) patients were readmitted to the ICU and 18 (34.6%) patients died without a difference between groups.

CONCLUSION

Prophylactic broad-spectrum antibiotics did not improve clinical outcomes in patients with DSC post-cardiac surgery but was associated with longer ventilation duration, length of ICU and hospital stays vs narrow-spectrum antibiotics.

Key Words: Cardiac surgery; Delayed sternal closure; Intensive care; Open chest; Prophylactic antibiotics

Core Tip: In cardiac surgical patients with delayed sternal closure the use of prophylactic broad-spectrum antibiotics did not affect the clinical outcomes in terms of length of mechanical ventilation and length of intensive care stay.

INTRODUCTION

Delayed sternal closure (DSC) after cardiac surgery is a therapeutic approach that can be lifesaving in certain situations such as severe hemorrhage, myocardial edema, arrhythmias, and low cardiac output[1-4]. The procedure can relieve excessive heart compression and allows easy access to control bleeding until the patient’s hemodynamic condition has been stabilized[2]. DSC is more popular in pediatric cardiac surgery[5]. The guidelines for antibiotic prophylaxis in cardiac surgery are well-defined[6,7]. However, there is a lack of evidence regarding the use of prophylactic antibiotics in DSC cases, and the existing literature does not address this specific matter in adult patients. Concerns have been raised about the potential susceptibility to bacterial infections when the chest is left open[8]. As a result, some surgeons have been reluctant to adopt this approach.

The existing evidence supports increased sternal infection rates in patients with DSC[9]. In a retrospective analysis performed by Das et al[10] in a pediatric population, it was found that DSC could pose a risk for blood stream infection. The antibiotic prophylaxis approaches, i.e. use of narrow-spectrum or broad-spectrum antibiotics, in DSC is highly dependent on the attending physician’s discretion. Thus, it is important to examine the outcomes of such approaches to justify the use of broad-spectrum antibiotics for prophylaxis in DSC patients.

We hypothesized that the use of broad-spectrum antibiotics may affect the outcomes differently when compared with the use of narrow-spectrum antibiotics in patients with DSC. The objective of this study was to investigate the outcomes of different prophylactic antibiotic regimens in patients who had DSC after cardiac surgery.

MATERIALS AND METHODS

Study design and setting

A single-center retrospective observational study was conducted in the cardiothoracic intensive care unit (CTICU) at a tertiary cardiac center between January 1, 2013 and December 31, 2020. The study was approved by the Corporation’s Medical Research Center with a waiver of informed consent given the retrospective design of the study, which was conducted in full agreement with the Declaration of Helsinki’s principles.

Study population

The study included all patients who were transferred to the CTICU with an open chest after cardiac surgeries during the study period. Patients were excluded if they were febrile, had evidence of sepsis, underwent thoracic surgery, received corticosteroid therapy, or died within 48 h after the decannulation of extracorporeal membrane oxygenation (where applicable). The cardiac surgical procedures in our study included coronary artery bypass grafting (CABG), valve replacement, combined CABG/valve surgery, and aortic dissection. All patients had received standard general anesthesia with standard cardiac monitoring as per our center’s protocol. Cardiopulmonary bypass (CPB) was conducted with roller pump and membrane oxygenator, and cardiac protection was maintained during CPB as per our center protocols.

Data collection

The data of all patients were collected from the electronic medical records [Dendrite Clinical Systems (London, United Kingdom) and Cerner (United States)] and included demographics, past medical history, laboratory blood tests, peri-procedure-specific data, and outcomes. Peri-procedure data included, but was not limited to, the durations of CPB and aortic cross clamp, inotropes and vasopressor use, blood loss details, and transfusion requirements. The main outcome measures included the length of hospital and intensive care unit (ICU) stay, duration of mechanical ventilation, and mortality. Other outcome measures included postoperative atrial fibrillation (POAF), perioperative myocardial infarction, the need for mechanical circulatory support, and infections (wound and nosocomial). Sample size calculation was not performed as all the patients who were transferred to the CTICU with an open chest during the study period were included.

DSC at heart hospital

The indications of DSC in our center include severe hemodynamic instability, persistent bleeding without an obvious source, and coagulopathy. The decision for DSC is made by the operating surgeon according to the requirements. The subsequent closure of the chest is also decided by the surgeon once the reason to initially leave the chest open is resolved. The choice of prophylactic antibiotics is selected by the treating CTICU clinical team.

Technique of DSC

The DSC protocol at our center is inserting bilateral chest drains after opening both pleurae, packing the heart and the great vessels, and applying a metal stent to keep the sternum open. The drains are connected to a low-pressure high-flow suction device.

Definitions

POAF was defined as the occurrence of new onset atrial fibrillation in the early postoperative period, i.e. within 7 d[11]. Acute kidney injury was defined according to the Kidney Disease Improving Global Outcomes criteria, which includes an increase in serum creatinine by ≥ 26.5 µmol/L (≥ 0.3 mg/dL) within 48 h or serum creatinine > 1.5 times the baseline value within the last 7 d after the insult, or reduced urine output < 0.5 mL/kg/h for six hours[12]. Postoperative myocardial infarction was defined as a rise in high-sensitivity troponin-T levels to 3466 ng/L, along with electrocardiographic, echocardiographic, or angiographic changes[13]. The duration of ventilation was defined as the time from CTICU admission after surgery to extubation as per our institution’s guidelines. Early mortality was defined as death within 28 d as per our corporation’s criteria. In surgical cases we prescribed the routine prophylactic antibiotic according to guidance in cardiac surgery from the thoracic surgeons practice guidelines series[7]. We adopted the definition of broad-spectrum antibiotics from Bratzler et al[14] as antibiotics with extended spectrum when compared with the routine and active against resistant microbes.

Statistical analysis

Sample size calculation was not performed as all the patients who were transferred to the CTICU with an open chest during the study period were included. Patients were divided into two groups based on the antibiotic regimen: Narrow-spectrum and broad-spectrum. Normally distributed continuous variables were presented as mean ± standard deviation and skewed variables as median and interquartile range (IQR). Continuous variables were compared using student’s t-test or Mann-Whitney U test, as appropriate. Fisher’s exact test or χ² were used to compare the categorical variables. The priori alpha level was set at < 0.05 (two-tailed). Data was managed using Excel and IBM SPSS Statistics (version 29.0.1.0 (171), IBM Corp. 2023) programs.

RESULTS

Patient characteristics

Fifty-three patients were included in the analysis. Of them, 12 (22.6%) received narrow-spectrum antibiotics and 41 (77.4%) received broad-spectrum antibiotics. The mean age and body mass index for all patients were 59.0 ± 12.1 years and 26.7 ± 4.9 kg/m², respectively, without significant differences between the groups. Most patients were males (85.0%), of an Arab ethnicity (45.3%), and non-smokers (45.1%). The most common comorbidities were hypertension (54.9%) and diabetes (43.4%) with no significant difference between the groups. Table 1 presents the collected baseline variables. In-hospital laboratory blood tests were unremarkable and did not differ between the groups as demonstrated in Table 2.

Table 1 Baseline characteristics among the studied population.

Variable	All patients, n = 53	Narrow-spectrum antibiotics, n = 12	Broad-spectrum antibiotics, n = 41	P value
Demographics
Age in yr	59.0 ± 12.1	60.0 ± 13.4	58.7 ± 11.8	0.366
BMI in kg/m2	26.7 ± 4.9 (N = 52)	25.7 ± 3.7 (n = 12)	26.7 ± 5.2 (n = 40)	0.227
Sex				0.175
Male	45/53 (85.0)	12	33
Female		0	8
Ethnicity				0.011
African	3/53 (5.7)	3	0
Arab	24/53 (45.3)	4	11
Asian from Indian subcontinent	20/53 (37.7)	3	17
Asia	4/53 (7.5)	0	4
Western	2/53 (3.8)	1	1
Smoking status				0.402
Non-smoker	23/51 (45.1)	3	20
Smoker	7/51 (13.7)	1	6
Ex-smoker	21/51 (41.2)	6	15
Comorbidities
Hypertension	28/51 (54.9)	4	24	0.086
Diabetes				0.305
Type II	16/53 (30.2)	4	12
Type I	7/53 (13.2)	0	7
COPD	2/52 (3.8)	0	2	1.00
CKD	8/53 (15.1)	1	7	0.665
On dialysis	2/53 (3.8)	0	2	1.00
PVD	2/53 (3.8)	0	2	1.00
Statin use	24/52 (46.2)	5	19	0.722
ECHO
LVEF as %	43.4 ± 16.9	45.8 ± 12.9	42.7 ± 18.0	0.578
LVEF groups				0.309
Good, > 50%	23/53 (43.4)	4	19
Moderate, 31%-50%	16/53 (30.2)	6	10
Poor, 21%-30%	10/53 (18.9)	2	8
Very poor, ≤ 20%	4/53 (7.5)	0	4

Data are total N/subset n (%) or mean ± standard deviation. BMI: Body mass index; CKD: Chronic kidney disease; COPD: Chronic obstructive pulmonary disease; ECHO: Echocardiography; LVEF: Left ventricular ejection fraction; PVD: Peripheral vascular disease.

Table 2 Laboratory test measurements among the studied population.

Variable	All patients, n = 53	Narrow-spectrum antibiotics, n = 12	Broad-spectrum antibiotics, n = 41	P value
Creatinine in micromole/L	95.0 (72.0-110) (N = 51)	79.0 (ΔIQR 24.0) (n = 11)	95.0 (ΔIQR 46.0) (n = 40)	0.688
Highest creatinine in micromole/L	110 (85.0-175) (N = 51)	95.0 (ΔIQR 29.0) (n = 11)	107.5 (ΔIQR 106) (n = 40)	0.457
Bilirubin in micromole/L	9.0 (6.9-17.1) (N = 47)	8.6 (ΔIQR 26.5) (n = 10)	9.5 (ΔIQR 6.7) (n = 37)	0.755
SGPT in U/L	26.0 (18.0-40.0) (N = 47)	18.0 (ΔIQR 13.5) (n = 9)	25.5 (ΔIQR 22.3) (n = 38)	0.185
HbA1c as %	6.3 ± 1.6	5.9 ± 1.7	6.3 ± 1.5	0.243
Hb in mg/dL	12.9 ± 2.0 (N = 52)	13.2 ± 1.9 (n = 12)	12.7 ± 2.0 (n = 40)	0.239
WBC as 103/U	10.8 ± 5.2 (N = 49)	6.8 (ΔIQR 4.0) (n = 10)	8.8 (ΔIQR 3.0) (n = 39)	0.076

Data are median (IQR) or mean ± standard deviation. Hb: Hemoglobin; HbA1c: Glycated hemoglobin; IQR: Interquartile range; SGPT: Serum glutamate pyruvate transaminase; WBC: White blood cells.

Antibiotic regimens

The narrow-spectrum antibiotics used were cefazolin and vancomycin. Ten (83.3%) patients were administered cefazolin alone, whereas the remaining two patients were given a combined cefazolin and vancomycin regimen. However, 23 (56.1%) patients in the broad-spectrum antibiotic group were initially administered a narrow-spectrum antibiotic regimen, and then therapy was escalated. The most used antibiotics in the broad-spectrum group were piperacillin-tazobactam followed by meropenem. Piperacillin-tazobactam was used as a single agent in 22% of patients or combined with another agent with Gram-positive coverage spectrum (e.g., vancomycin, teicoplanin) in 31.7% of patients. The respective percentages for meropenem use were 2.4% (single) and 22.0% (combined). Sequential antibiotic use comprised various regimens; piperacillin-tazobactam was the initial agent in 17.1% of patients and meropenem in 4.8% of patients (Figure 1). The estimated duration of antibiotic regimen in all patients was a median of 7.5 (IQR 3.0-14.0) d. The mean duration of antibiotic use was significantly longer in the broad-spectrum group than the narrow-spectrum group (11.9 ± 8.7 d vs 3.4 ± 2.0 d, P < 0.001).

Figure 1  Broad-spectrum antibiotic regimens among the studied group.

Peri-procedural findings

CABG procedure was performed in 58.5% of patients, followed by valve replacement (20.8%) and combined surgeries (11.3%). The mean CPB duration for all patients was 165 ± 75.3 min. Patients in the narrow-spectrum antibiotics group had a significantly shorter mean CPB duration than those in the other group (127.3 ± 67.9 min vs 174.4 ± 74.9 min, P = 0.037). The aortic cross clamp time was 85.0 ± 45.7 min for all patients without a difference between the groups. The median durations of open chest were 3.0 (2.0-5.0) d for all patients, 2.0 (ΔIQR 2.0) d for the narrow-spectrum group and 3.0 (ΔIQR 2.0) d for the broad-spectrum group (P = 0.146). Most patients were transfused intraoperatively (78.8%) and postoperatively (96.1%). Total or at-any-time interval blood loss did not differ significantly between the groups. Pharmacological (97.9%) and mechanical (50.9%) circulatory support did not differ between the groups (Table 3).

Table 3 Peri-procedural findings.

Variable	All patients	Narrow-spectrum antibiotics, n = 12	Broad-spectrum antibiotics, n = 41	P value
Procedure data				0.398
CABG	31/53 (58.5)	7	24
Valve replacement	11/53 (20.8)	4	7
Combined CABG/valve surgery	6/53 (11.3)	0	6
Aortic dissection	5/53 (9.4)	1	4
Epicardial wires	16/52 (30.8)	4	12	1.00
CPB in min	165 ± 75.3 (N = 50)	127.3 ± 67.9 (n = 10)	174.4 ± 74.9 (n = 40)	0.037
ACC in min	85.0 ± 45.7 (N = 41)	83.2 ± 50.5 (n = 9)	85.4 ± 45.1 (n = 32)	0.451
Open chest duration in d	3.0 (2.0-5.0)	2.0 (ΔIQR 2.0)	3.0 (ΔIQR 2.0)	0.146
Intraoperative transfusion	41/52 (78.8)	10	31	1.00
Postoperative transfusion	49/51 (96.1)	12	37	1.00
Blood loss in mL
At 6 h	700 (375-1350) (N = 45)	900 (ΔIQR 1050) (n = 11)	1050 (ΔIQR 1250) (n = 34)	0.864
At 12 h	1225 (500-2050) (N = 44)	1750 (ΔIQR 1000) (n = 11)	1575 (ΔIQR 2088) (n = 33)	0.881
At 24 h	1700 (775-2800) (N = 45)	1900 (ΔIQR 2500) (n = 11)	2175 (ΔIQR 3213) (n = 34)	0.8443
Total loss	2400 (1362-4700) (N = 44)	2050 (ΔIQR 4375) (n = 11)	3325 (ΔIQR 4300) (n = 33)	0.492
Support
Inotropes/vasopressors use
Agent(s) use	46/47 (97.9)	11	35	1.00
Number of agent(s) used				0.385
0	2/49 (4.1)	0	2
1	3/49 (6.1)	2	1
2	16/49 (32.7)	4	12
3	20/49 (40.8)	5	15
4	8/49 (16.3)	1	7
Number of agents used	2.6 ± 0.9 (N = 49)	2.4 ± 0.9 (n = 12)	2.6 ± 1.0 (n = 37)	0.24
IABP use	27/53 (50.9)	6	21	0.941
Defibrillation	13/53 (24.5)	3	10	1.00

Data are total N/subset n (%), median (IQR) or mean ± standard deviation. ACC: Aortic cross clamp; CABG: Coronary artery bypass grafting; CPB: Cardiopulmonary bypass; IABP: Intra-aortic balloon pump; IQR: Interquartile range.

Clinical outcomes

Table 4 demonstrates clinical outcomes. The proportion of patients who experienced new nosocomial infection was 40.0%. Sepsis was reported in 50 patients, while data was missing for 3 patients. The mean time to sepsis was 4.8 ± 2.6 d for all patients, with similar times in both groups. Almost half of the patients (46.2%) experienced POAF and 13.5% of them experienced heart block. Only 2 patients had myocardial infarction post-surgery, and 8 patients experienced a stroke.

Table 4 Clinical outcomes.

Variable	All patients	Narrow-spectrum antibiotics, n = 12	Broad-spectrum antibiotics, n = 41	P value
New nosocomial infection	20/50 (40.0)	5	15	0.736
Time to sepsis/septic shock postoperatively in d	4.8 ± 2.6 (N = 50)	3.0 (ΔIQR 4.0) (n = 11)	3.5 (ΔIQR 2.0) (n = 39)	0.586
POAF	24/52 (46.2)	5	19	0.722
New ECG changes	39/51 (76.5)	9	30	1.00
Heart block	7/52 (13.5)	2	5	0.656
LV dysfunction	7/52 (13.5)	0	7	0.181
Postoperative MI	2/52 (3.8)	0	2	1.00
New postoperative stroke				0.576
Transient	3/50 (6.0)	0	3
Permanent	5/50 (10.0)	1	4
Duration of ventilation in h	70.0 (39.0-190)	50.0 (ΔIQR 113)	60.0 (ΔIQR 170)	0.047
LOS-ICU in d	8.0 (3.25-15.5) (N = 52)	5.0 (ΔIQR 5.0) (n = 12)	7.5 (ΔIQR 10.0) (n = 40)	0.008
LOS-hospital in d	27.0 (10.5-47.0) (N = 44)	19.0 (ΔIQR 21.0) (n = 11)	27.0 (ΔIQR 30.0) (n = 33)	0.031
Mortality	18/52 (34.6)	4	14	1.00
Readmission to ICU	5/51 (9.8)	1	4	1.00

Data are total N/subset n (%), median (IQR) or mean ± standard deviation. ECG: Electrocardiographic; ICU: Intensive care unit; IQR: Interquartile range; LOS: Length of stay; LV: Left ventricular; MI: Myocardial infarction; POAF: Postoperative atrial fibrillation.

Mechanical ventilation: The median duration of mechanical ventilation was 70.0 (IQR 39.0-190) h, with a significantly longer duration in the broad-spectrum antibiotic group [60.0 (ΔIQR 170.0) h vs 50.0 (ΔIQR 113.0) h, P = 0.047].

Length of stay and mortality: The median length of stay for both ICU and hospital were 8.0 (3.25-15.5) d and 27.0 (10.5-47.0) d, with significantly longer durations in the broad-spectrum antibiotic group [7.5 (ΔIQR 10.0) d vs 5.0 (ΔIQR 5.0) d, P = 0.008] and [27.0 (ΔIQR 30.0) d vs 19.0 (ΔIQR 21.0) d, P = 0.031], respectively. Five (9.8%) patients were readmitted to the ICU and eighteen (34.6%) patients died without a difference between the groups.

DISCUSSION

The aim of this study was to determine the outcomes related to the use of different regimens of prophylactic antibiotics in adult patients with DSC after cardiac surgery. The salient findings of this research were: (1) A high rate of infection-related complications (40.0%); (2) patients in the broad-spectrum antibiotic group required more prolonged antibiotic course; (3) the incidence of POAF was exceptionally high (46.2%); (4) the length of stay for both ICU and hospital were significantly longer in the broad-spectrum antibiotic group; (5) the rates of ICU readmission and mortality were only numerically higher with broad-spectrum antibiotic use; and (6) the use of broad-spectrum antibiotics did not provide an outcome-related benefit.

The DSC management strategy after cardiac surgery has been widely adopted in the pediatric population for possible avoidance of compression of the swollen myocardium and lungs with increased fluid collection. Improvement in hemodynamics is mainly notable in the diastolic functions[15,16] or variable in adult cardiac surgery[8,17,18].

Infection rate

In our study we reported a high rate of new nosocomial infections (40.0%). Sepsis was reported in 50 patients. The mean time to sepsis was 4.8 d for all patients, with similar times between the groups. Similarly, Das et al[10] reported a high rate of blood stream infection in patients with DSC (30% vs 9% in the control group).

Antibiotics regimens

While prophylactic antibiotic guidance follows a clear guideline in routine cardiac surgery, i.e. cefazolin or vancomycin[6], the antibiotic regimen in patients with DSC remains unsettled. In our study, the choice of antibiotics for patients with DSC was dependent on the discretion of the managing team. Prophylactic narrow-spectrum and broad-spectrum antibiotics were used in 22.6% and 77.4% of patients, respectively. The published literature on the use of prophylactic antibiotics was diverse; in pediatric populations a simple antibiotic regimen[19-21] or broad-spectrum antibiotics were used[8]. In adult cardiac surgery, the use was inconsistent as well[8-10,22,23].

The estimated duration of the antibiotic regimen was 7.5 d in all patients, with a significantly longer use in broad-spectrum group (11.9 d vs 3.4 d). The optimal duration of prophylactic antibiotics in DSC was not previously addressed in the published literature[7,14], which showed variable durations of therapy ranging from a median of 7.73 d total and 3.1 d until sternal closure[23].

Outcome measures

The median length of stay for both ICU and hospital were 8.0 d and 27.0 d, respectively, with significantly longer durations in the broad-spectrum group. The ICU length of stay reflects patient morbidity and resource utilization and remains a popular outcome measure despite its limitations[24]. The median duration of mechanical ventilation was 70.0 h, and it was significantly longer in the broad-spectrum group. Previously published studies in the adult population were heterogenous, of limited number of patients, and did not report ICU length of stay or duration of ventilation use in DSC patients. The incidence of POAF was numerically higher in patients treated with broad-spectrum antibiotics (19 patients vs 5 patients, P = 0.722), which may be attributed to the prolonged ICU and hospital course as reported in previous studies[25].

Broad-spectrum coverage

The usage of broad-spectrum antibiotics did not improve clinical outcomes in our study. The concerns of acquiring wound infection and possible mediastinitis may have influenced the choice of broad-spectrum antibiotics in many cases. Eckard et al[23] did not find a significant advantage of extended-spectrum or broad-spectrum coverage on wound infection or mediastinitis with trends towards increased complications in adults with DSC. On the other hand, Boeken et al[17] routinely started broad-spectrum antibiotic prophylaxis with piperacillin/tazobactam and continued it until closing the sternum with a low incidence of deep sternal and superficial wound infections (5.3% and 4.8%, respectively). Shalabi et al[8] used a combination of cephalosporin and aminoglycoside for prophylaxis and reported high rates of superficial wound infection (22%). The authors did not report delayed sternal infection or mediastinitis. The broad-spectrum antibiotic regimen that was advocated by some investigators in the pediatric cardiac surgery population[26] did not provide benefit as well.

Limitations

The study was conducted in a single center with a limited number of patients. Despite the 7-year recruitment period in our center with over 500 cardiac surgeries per year, we reported outcomes of only 53 patients with DSC. This may indicate that DSC is rarely required in our patient population. Given the small sample size and the retrospective nature of the design, it may be difficult to draw a solid conclusion regarding the study questions. Larger, randomized multicenter studies with a longer follow-up period are needed to further confirm our results.

CONCLUSION

Prophylactic broad-spectrum antibiotics use did not improve clinical outcomes in patients with DSC after cardiac surgery. It was associated with longer ventilation duration and length of ICU and hospital stay compared with narrow-spectrum antibiotics use. Further randomized controlled trials are required to confirm the results.

ACKNOWLEDGEMENTS

We thank all members of the Cardiothoracic Surgery Department.