A Comparison of Negative Pressure and Conventional Therapy in Infected Open Chest Wounds

Abstract

Background:

The role of negative pressure wound therapy (NPWT) in the management of open chest wounds is unclear. Our aim was to determine the safety and efficacy of NPWT compared with conventional therapy for open chest wounds.

Methods:

Ten patients with infected open chest wounds were included in a prospective trial of NPWT after surgical debridement. Their outcomes were compared with those of 11 control patients treated during the same period with surgical debridement and open chest packing only. The control group data were obtained by retrospective review of medical records.

Results:

The median duration of NPWT was eight days (range 2–29 days), with closure in eight patients (80%). Two patients having NPWT had unveiling of occult pleural fistulas leading to early discontinuation. The patients having NPWT had a shorter median time to closure (7 versus 18 days; p = 0.071) and shorter initial (median 6 versus 20 days; p = 0.026) and total (median 6 versus 25 days; p = 0.024) hospital length of stay. Control patients had higher rates of new-onset atrial fibrillation (46% versus 0; p = 0.035) and septic shock (64% versus 10%; p = 0.024). The chest was either closed or healing at the time of the last visit in 100% of the NPWT patients versus 73% of control patients (p = 0.28). The 1-year survival estimates were 90% for the NPWT patients and 80% for the control patients (p = 0.69).

Conclusion:

Negative pressure wound therapy is feasible and safe for open infected chest wounds in selected patients compared with open packing alone and may reduce hospital stay duration and major complication rates.

Thoracic cavity and chest wall infections with or without exposed lung are challenging to manage, often requiring prolonged drainage and intravenous antibiotics. When identified early, such infections often can be managed using percutaneous or minimally invasive techniques [1], but in other instances, open surgical debridement, muscle or omental flaps, and chest packing may be required [2]. Recalcitrant infections or those associated with a chronic broncho-pleural fistula (BPF) typically are managed with a permanent or temporary thoracostomy window for drainage [3]. The complex procedures often necessary for management increase the inherently high morbidity and mortality rate associated with empyema [4].

Negative pressure wound therapy (NPWT), also known as vacuum-assisted closure therapy, is an effective method of managing complicated wounds that do not involve viscera directly [5]. More recent iterations of NPWT devices that integrate non-adherent barriers and fluid instillation capabilities allow even open wounds with exposed viscera to be treated with negative pressure without major adverse events [6,7]. Using consistent negative pressure, NPWT prevents fluid accumulation in the wound bed and pulls the wound edges together gradually. In addition, wound healing is accelerated through promotion of granulation tissue and perfusion [8]. Thus, NPWT is established as safe and effective in the management of deep sternal [9,10] and chest wall infections when lung is not exposed [11,12]. However, its role in the treatment of open chest wounds (OCW) is supported by limited data, mainly by retrospective small case series or case reports [13]; the evidence is even sparser for NPWT and infections with exposed lung or pneumonectomies.

We conducted a prospective study to determine the feasibility and safety of NPWT as an adjuvant to surgical debridement in OCW and compared the outcomes with a control group of patients managed with conventional therapy (CT) of surgical debridement and open packing during the same study period.

Patients and Methods

The Mayo Clinic Institutional Review Board approved both the prospective study of NPWT and the retrospective review of CT through separate applications. We defined OCW as any chest wound where a primary closure of the thoracotomy was not being or could not be performed, resulting in a communication between the chest cavity and the external atmosphere. When an adult (>17 years) patient presented with a chronic or recent chest infection requiring surgical debridement with a high probability of a need for open packing, NPWT was discussed if considered appropriate after review by the surgeon; and informed patient consent was obtained. However, the final decision about the therapeutic pathway was left to the best judgment of the surgeon. Exclusions for NPWT included active BPF (defined as any one or all of the following: Visible air bubbling in the chest cavity under fluid immersion; an audible whistle with coughing; bronchial defect seen by bronchoscopy), active chest malignancies, altered mental status, inability to provide informed consent, need for pharmacologic anti-coagulation, and history of bleeding disorders or coagulopathy. Informed consent was waived for the retrospective review of controls because of the minimal risk. The prospective study data were entered in a Research Electronic Data Capture (REDCap) online database and reviewed periodically for adverse events by the institutional Data Safety Monitoring Board.

Study subjects

A total of 11 patients having NPWT and 11 having CT (control) therapy undergoing OCW management at a Mayo Clinic campus in Jacksonville, Florida, or Rochester, Minnesota, between February 2016 and July 2019 were included in this study. One patient who consented to NPWT was later excluded following intra-operative determination that the infection was not severe enough for open packing. In the CT group, three patients with deep venous thrombosis did not qualify for NPWT because of their need for therapeutic anticoagulation, three for active BPF, one because the first operation included partial resection of the left atrium for tumor involvement, one for trapped lung that required extensive decortication, while the reasons for the remaining three exclusions were unclear.

Information was collected about demographics, co-morbidities, NPWT settings, wound condition, surgery performed, and outcomes. Specific outcome measures of interest were wound status at last visit; total number of operations performed, including dressing changes under anesthesia; initial hospital length of stay (LOS), and total hospital LOS (i.e., including re-admissions) for any reason related to the OCW, re-admissions, and survival after the first operation performed specifically to manage the infection.

Most patients were transferred from other institutions for management of a complicated infection after lung surgery. Limited data were available about the details of the index operation for some patients who had multiple operations for empyema prior to their transfer, and we considered only the first operation done at the Mayo Clinic when calculating LOS and time to closure.

Surgical debridement

During the initial surgical exploration and prior to NPWT or open packing, surgical debridement of infected or necrotic tissues was performed in all cases and included the use of a pulse-irrigation device, curettage, and sharp excision with a knife or scissors. In two patients, visceral pleural decortication also was performed. During subsequent dressing changes, any additional fibrinous exudate or eschar that had developed was debrided.

NPWT technique

The NPWT system used in the study (V.A.C. Therapy System, KCI USA Inc., St. Paul, MN) consisted of a medical-grade non-adherent polyvinyl alcohol white foam applied directly to the infected surface, followed by an open-pore reticulated polyurethane black foam cut to fit the wound and covered by a transparent air-tight adhesive drape. A suction cup with tubing was placed over a small slit on the drape and connected to a suction machine (V.A.C. ULTA Therapy Unit). The target pressures were set ranging from -50 mm Hg to -125 mm Hg, as continuous or intermittent, at the surgeon's discretion based on the intra-operative findings. Scheduled dressing changes were done once every 48 hours. The system included alarms to signal tube blockages, leaks, or low pressure. The total number of foam pieces placed during each dressing change was documented to verify complete removal.

Conventional technique

Patients in the CT group underwent open surgical debridement and packing with gauze rolls soaked in either double-antibiotic (polymyxin and bacitracin) solution (in three), normal saline (in five), povidone–iodine Betadine (in two), or Dakin's solution (sodium hypochlorite) changed to normal saline after two dressing changes (in one patient). The choice of the soaking solution was not based on institutional or society guidelines but on the surgeon's preference. Dressings were changed once or twice daily according to the severity of contamination.

Statistical analysis

Continuous variables were summarized with the sample median and range. Categorical variables were summarized with number and percentage. Survival after the first operation was estimated using the Kaplan-Meier method, where patients who survived were censored on the date of last follow-up. Comparisons of characteristics between NPWT and CT patients were made using a Wilcoxon rank sum test (continuous variables), Fisher exact test (categorical variables), or a log-rank test (survival after first operation). All statistical tests were two-sided. P values <0.05 were considered statistically significant. Statistical analyses were performed using SAS (version 9.4; SAS Institute, Inc., Cary, NC).

Results

A summary of patient characteristics is shown in Table 1. There were no notable differences in the demographics of the two groups except for a preponderance of males in the NPWT group (91% versus 30%; p = 0.008). All patients underwent surgery within 48 hours after consultation except for the two NPWT patients with chronic BPF. No patient demonstrated signs of sepsis or septic shock at the time of initial evaluation or surgery, and none required ventilator or vasoactive support prior to surgery. At the time of consultation, all patients were receiving intravenous antibiotics chosen according to microbial culture and sensitivities, and these were continued post-operatively. Wound and surgery information is displayed in Table 2.

Table 1.

Patient Characteristics

Variable	NPWT (N = 10)	Controls (N = 11)	P^a
Variable	Median (range) or no. (%)	Median (range) or no. (%)	P^a
Demographic information
Age (years)	61 (22, 75)	62 (18, 77)	1.00
Sex (male)	3 (30)	10 (91)	0.008
Admission weight (kg)	64 (48, 110)	85 (53, 116)	0.48
Admission Body Mass Index	24.4 (18.8, 37.8)	26.3 (16.3, 37.9)	0.97
Co-morbidities
History of smoking	5 (50)	7 (64)	0.67
Diabetes mellitus	1 (10)	2 (18)	1.00
Hypertension	4 (40).	7 (64)	0.39
Renal disease	2 (20)	1 ( 9)	0.59
Active malignant disease	2 (20)	4 (36)	0.64
Chronic obstructive pulmonary disease	4 (40)	5 (46)	1.00
Immunosuppression	2 (20)	0	0.21

P values result from a Wilcoxon rank sum test (continuous variables) or Fisher exact test (categorical variables).

Table 2.

Wound and Surgery Information

Variable	NPWT (N = 10)		Controls (N = 11)		P^a
Variable	N	Median (range) or no. (%)	N	Median (range) or no. (%)	P^a
Wound information
Wound (empyema) etiology
Post-operative	10	5 (50)	11	10 (91)	0.063
Spontaneous	10	3 (30)	11	1 (9)	0.31
Traumatic	10	1 (10)	11	0	0.48
Radiation necrosis	10	1 (10)	11	0	0.48
Wound dimensions (cm)
Length	10	13 (0, 27)	0	N/A	N/A
Width	10	11 (0, 21)	0	N/A	N/A
Depth	9	5 (0, 10)	0	N/A	N/A
Length × width × depth (cm³)	9	442 (0, 5400)	0	N/A	N/A
Type of initial surgery	5		10		0.045
Wedge resection		0		1 (10)
Lobectomy		1 (20)		2 (20)
Pneumonectomy		1 (20)		5 (50)
Chest wall resection		2 (40)		0
Abdominal surgery		1 (20)		1 (0)
Repair of radiation-induced tracheoesophageal fistula		0		1 (1)
History of bronchopleural fistula (BPF)	10	3 (30)	11	8 (73)	0.086
BPF closed	3	2 (67)	8	6 (75)	1.00
Flaps used	2	2 (100)	6	6 (100)	N/A
Type of flap	2		6		1.00
Muscle		2 (100)		5 (83)
Omentum		0		1 (17)
Type of muscle flap	2		5		1.00
Latissimus dorsi		1 (50)		2 (40)
Pectoralis major		1 (50)		1 (20)
Transverse rectus		0		2 (40)
Surgery information
First operation performed for chest infection	10		11		0.78
Thoracotomy with washout/debridement		5 (50)		5 (46)
Lung resection		1 (10)		1 (9)
Eloesser window		1 (10)		3 (27)
Other		3 (30)		2 (18)
Number of additional operations	10		11		0.26
0		0		1 (9)
1		1 (10)		0
2		4 (40)		1 (9)
3		3 (30)		1 (9)
4		1 (10)		3 (27)
5		0		1 (9)
6		0		1 (9)
8		1 (10)		0
9		0		1 (9)
10		0		1 (9)
12		0		1 (9)
Type of additional procedures (any of the following)
Thoracotomy with washout and debridement	10	8 (80)	11	8 (73)	1.00
Lung resection	10	1 (10)	11	0	0.48
Eloesser window	10	2 (20)	11	4 (36)	0.64
Modified Clagett	10	1 (10)	11	3 (27)	0.59
Total days from first operation at Mayo Clinic to final closure	9	7 (4, 49)	7	18 (4, 179)	0.071

P values result from a Wilcoxon rank sum test (continuous variables) or Fisher exact test (categorical variables).

Post-operative infections were present in more than 90% (10/11) of control patients compared with 50% (5/10) of the NPWT group (p = 0.063). Three patients having NPWT had spontaneous chest infections: Two rib osteomyelitis cases and one right lung destroyed by Mycobacterium abscessus. Of note, 50% of post-operative infections occurred after pneumonectomy in the control patients compared with 20% (n = 1) in the NPWT group (p = 0.045); a history of BPF was less common in the NPWT group (30% versus 73%, p = 0.086). The median time from the first operation to definitive closure was shorter with NPWT (median 7 versus 18 days; p = 0.071). The number of additional procedures done under anesthesia for each group is shown in Fig. 1.

FIG. 1.

Number of additional operations performed per patient in each group.

The NPWT information is displayed in Table 3. The median length of time from first operation to first application of NPWT was 0 days (range 0–25 days). The median total duration of NPWT was 8 days (range 2–29 days) and resulted in successful closure for 8 patients (80%).

Table 3.

Negative Pressure Wound Therapy (NPWT) Information

Variable	Median (range) or no. (%)
Days from first surgery to first application of NPWT	0 (0, 25)
Negative pressure applied (mm Hg)
-50	3 (30)
-75	3 (30)
-125	4 (40)
Mode of pressure
Continuous	8 (80)
Intermittent	2 (20)
Total duration of NPWT (days)	8 (2, 29)
Non-NPWT used at any time during treatment	4 (40)
NPWT successful in closure	8 (80)
Early discontinuation/failure	2 (20)
Inability to maintain suction, uncovering of fistula	1 (50)
Muscle flap necrosis, persistent air leak from remaining lobe and bronchopleural fistula	1 (50)

A summary of complications is shown in Table 4. Control patients had higher rates of atrial fibrillation (46% versus 0; p = 0.035) and sepsis/septic shock (64% versus 10%; p = 0.024). Other complications were observed relatively rarely. One patient in the NPWT group underwent re-operation for bleeding from the chest wall after debridement; this was considered unrelated to the NPWT. After cauterization, NPWT was re-applied without further bleeding. We tried to analyze for any relation between outcomes and NPWT pressures but did not reach any significant conclusions because of the small sample. Interestingly, the vacuum failures occurred in patients in whom -50 mm Hg was used, perhaps indicating that the surgeon was concerned about tissue fragility during the application.

Table 4.

Complications

Complication	NPWT (N = 10) (%)	Controls (N = 11) (%)	P
Atrial fibrillation	0	5 (46)	0.035
Sepsis/septic shock	1 (10)	7 (64)	0.024
Pneumonia	1 (10)	3 (27)	0.59
Wound dehiscence	1 (10)	0	0.48
Major adverse cardiac event	0	1 (9)	1.00
Air embolism	0	1 (9)	1.00
Death (30-d or in hospital)	2 (20)	1 (9)	0.59
Reintubation	0	2 (18)	0.48
Acute respiratory distress syndrome	0	1 (9)	1.00
Acute kidney failure requiring dialysis	0	1 (9)	1.00
Hemothorax requiring re-operation	1 (10)	1 (9)	1.00

P values result from a Wilcoxon rank sum test (continuous variables) or Fisher exact test (categorical variables).

Statistically significant data are in boldface type.

A comparison of outcomes for the two groups is shown in Table 5. At the time of the last visit, the chest was closed, or open and healing, more commonly in the patients having NPWT than in control patients (100% versus 73%; p = 0.28). The patients undergoing NPWT also had a shorter initial (median 6 versus 20 days; p = 0.026) and total (median 6 versus 25 days; p = 0.024) hospital LOS. The median length of follow-up after the first operation was 14.3 months (range 1.4–37.9 months), during which three patients in each group died. There was no significant difference in survival between the NPWT patients and control patients, with 1-year survival estimates in the two groups of 90% and 80% (p = 0.69), respectively (Table 5).

Table 5.

Outcomes

Variable	NPWT (N = 10)	Controls (N = 11)	P
Variable	Median (range) or no. (%)	Median (range) or no. (%)	P
Wound status at last visit			0.28
Chest closed	8 (80)	6 (55)
Chest open and healing	2 (20)	2 (18)
Chest open without signs of healing	0	3 (27)
Initial hospital length of stay	6 (1, 17)	20 (3, 152)	0.026
Total hospital length of stay (including re-admissions)	6 (1, 43)	25 (3, 152)	0.024
Re-admission	2 (20.0)	5 (46)	0.22
Survival after first operation (%) (95% confidence interval)			0.69
30 d	9 (72, 100)	100 (74, 100)
60 d	90 (72, 100)	10 (72, 100)
90 d	90 (72, 100)	100 (72, 100)
180 d	90 (72, 100)	80 (58, 100)
1 y	90 (67, 100)	80 (56, 100)
2 y	34 (7, 100)	60 (29, 100)

Statistically significant data are in boldface type.

In two patients with BPFs, NPWT was discontinued prematurely because of failure to hold suction. At the time of their last follow up, the wounds in both patients appeared to be healing. The first patient was referred with chronic post-lobectomy empyema of 1-year duration. Although no fistula was seen, she was suspected to have occult pulmonary–pleural fistulae underneath thick fibrinopurulent exudate. Following washout and application of NPWT, she was discharged home after one dressing change but presented a day later with a “whistling sound in the throat” that coincided with the start of the negative pressure cycle and sounding of the device alarm, indicating a failure of the vacuum seal. After removing the dressing, we observed that most of the overlying fibrin on the lung had been debrided, uncovering micro-fistulae. She was subsequently managed with an Eloesser window.

The second patient was referred for total right upper-lobe necrosis and a chronic BPF of nearly 1 year duration after she had undergone high-dose radiation for a lung mass. After multiple surgical debridements, the BPF was closed with a serratus muscle flap, and NPWT was applied. She was placed on veno-venous extracorporeal membrane oxygenation to assist with healing. However, the distal end of the muscle flap became ischemic, leading to reopening of the fistula. Although we planned for additional muscle reinforcement using a transabdominis rectus flap, the patient opted to withdraw from treatment and died in the hospital.

Discussion

Our study demonstrates that NPWT is feasible and safe in the management of selected complex OCWs compared with surgical debridement and wound packing alone. In our analysis, the primary benefits of NPWT were a significant reduction in hospital LOS (both initial and overall), reduced time to closure, and fewer complications. However, long-term survival was not affected. We also observed that some major complications such as atrial fibrillation and septic shock were less frequent in patients treated with NPWT. Although the cause of these differences in complication rates may be multifactorial, it is plausible that the technique of wound management contributed as well. The higher rate of septic shock seen with CT could have resulted from inadequate drainage of the infection; the negative pressure of NPWT constantly clears infective fluid out of the chest cavity, while packing alone potentially creates a damming effect, resulting in a pool of purulent material that can be cleared effectively only by changing the dressings. The nearly double rate of atrial fibrillation in the CT group was surprising but could be the result of their higher rate of sepsis. The incidence of new-onset atrial fibrillation is well known to increase cumulatively with the severity of sepsis [14,15].

The current evidence on the use of NPWT for OCWs is limited but supports our results [12,13,16 –19]. In a 2011 best-evidence review of nine retrospective studies on intra-thoracic NPWT that included 69 patients, eight were either case series or case reports [13]. The only cohort study in this review reported a significantly lower time to closure of open window thoracostomy with NPWT (n = 11) compared with no NPWT (n = 8) (average duration 39 ± 17 days versus 933 ± 1422 days) and better long-term survival in the NPWT group [20]. A single-institutional analysis of NPWT for complex infected thoracic wounds reported successful closure in all 17 patients without a rotational flap, and one recurrence of infection after delayed primary closure [16]. In another retrospective series of 27 patients with post-resectional empyema who were treated with intrathoracic NPWT, chest closure was achieved eventually in all who survived [21]. Of note, the in-hospital mortality rate for this group was 19%, reflecting the severity of the infection and the poor health of the patients. In addition to the therapeutic benefits discussed above, NPWT allows a longer interval between dressing changes than the daily or even more frequent changes needed with CT [21 –23]. This reduces the discomfort and pain associated with dressing changes while saving valuable resources and time [19].

There are valid concerns regarding the universal applicability of NPWT in chest infections, especially the risk of significant bleeding and lung injury when it is applied directly to the lung. Additionally, occult or active BPF can lead to loss of negative pressure, as demonstrated in our study. However, in the absence of a BPF or air leak, we did not observe residual lung in itself to contraindicate NPWT, provided an adequate non-adherent barrier is used, a finding supported by other authors [19,20]. We used the non-adherent white foam, which is easier and less painful to remove than the black foam because granulation in-growth is much less likely. Other non-adherent interfaces that can be used include fenestrated silicone membranes or paraffin-impregnated porous dressings. Another limitation is the presence of a fresh bronchial stump or exposed major pulmonary vessel, which may rupture or erode when in direct contact with the dressing sponge. Hence, the use of a pedicled, well-vascularized, protective layer of omentum, pericardial fat, or extra-thoracic muscle flap to cover such vulnerable structures is critical prior to application of NPWT [11]. In a retrospective study of 21 patients with post-pneumonectomy empyema treated with NPWT, the chest was closed in 90% (n = 19) of all surviving patients with a Clagett-type procedure [12].

The duration of NPWT for complex thoracic wounds depends on the size and location of the wound, the indication for use, and patient condition; it may range from <1 week to months as observed by us and other authors [13,19 –21]. The NPWT does not necessarily eliminate the need for additional surgical procedures to close large wounds but may decrease the time required for such secondary closures, as shown in our study. The presence of good granulation tissue in the wound bed appears to be a key determinant of the timing for definitive closure, rather than negative microbiology findings [12]. In a prospective study of 10 empyema patients treated with open window thoracostomy and NPWT, the authors reported successful secondary closure without surgery in all surviving patients (n = 9) over a duration of 2–7 months [19].

Thus, NPWT is a powerful adjunct to wound healing but does not prevent recurrent infection over time [16,17,23], especially if a dead space remains after closure. Patients who develop post-operative empyema typically are immunocompromised for multiple reasons and thus prone to recurrent infections. Regardless of how clean the wound space appears grossly at the time of closure, sterilization is unlikely to be achieved [12,24]; and any remaining dead space provides a favorable substratum for microbial growth. In particular, infections with refractory or multi–drug-resistant organisms or in immunocompromised patients are more likely to recur [25], as in the case of one of our patients who developed a recurrent M. abscessus infection within 30 days of closure and subsequently died from sepsis. For such infections, when the dead space cannot be obliterated, a permanent thoracostomy window may be a better option than definitive closure for long-term management.

There are obvious limitations to our non-randomized and non-blinded study, which was conducted in a small number of patients. The treatment selection was based on the best judgment of the surgeon and potentially was subject to many biases. The different soaking solutions used for packing highlight the variability among surgeons and the lack of standard management techniques for OCW. Although a randomized controlled study would be ideal to study the value of NPWT, it still would depend on the surgeon's operative assessment because of the variable nature of empyema and likely would jeopardize enrollment in and completion of the study. Our study population reflects some of the heterogeneity of OCWs but is by no means exhaustive. Our purpose was not to determine if NPWT will be able to replace traditional therapies but rather to understand its appropriateness and the limitations of its use as adjunctive therapy. The current NPWT technology is not feasible with active BPF or air leaks, thus limiting the relevance of the results of our study to a selected population of patients with empyema.

Conclusion

A NPWT is feasible and safe in selected patients with OCWs compared with traditional surgical management. The presence of an active or occult BPF remains a contraindication to NPWT in our experience. Larger randomized controlled studies will be helpful to determine if NPWT is superior to CT but may be difficult to complete because of the challenges of maintaining surgeon equipoise.

Footnotes

Funding Information

This project was funded by a peer-reviewed grant from the Thoracic Surgery Foundation through Acelity, Inc.

Author Disclosure Statement

The authors have no other conflicts of interest to disclose.

References

Ben-Nun

, Soudack

, Best

. Non-surgical treatment for post pneumonectomy empyema. Interact Cardiovasc Thorac Surg, 2003; 2:616–619.

Schneiter

, Cassina

, Korom

, et al. Accelerated treatment for early and late postpneumonectomy empyema. Ann Thorac Surg, 2001; 72:1668–1672.

Shamji

, Ginsberg

, Cooper

, et al. Open window thoracostomy in the management of postpneumonectomy empyema with or without bronchopleural fistula. J Thorac Cardiovasc Surg, 1983; 86:818–822.

Gharagozloo

, Margolis

, Facktor

, et al. Postpneumonectomy and postlobectomy empyema. Thorac Surg Clin, 2006; 16:215–222.

Anghel

, Kim

. Negative-pressure wound therapy: A comprehensive review of the evidence. Plast Reconstr Surg, 2016; 138:129S–137S.

Zheng

, Lu

, Zhang

, et al. Novel abdominal negative pressure lavage-drainage system for anastomotic leakage after R0 resection for gastric cancer. World J Gastroenterol, 2019; 25:258–268.

Alvarez

, Betancourt

, Fernandez

. Negative pressure wound therapy with instillation in the septic open abdomen utilizing a modified negative pressure therapy system. Ann Med Surg (Lond), 2018; 36:246–251.

Huang

, Leavitt

, Bayer

, Orgill

. Effect of negative pressure wound therapy on wound healing. Curr Probl Surg, 2014; 51:301–331.

Falagas

, Tansarli

, Kapaskelis

, Vardakas

. Impact of vacuum-assisted closure (VAC) therapy on clinical outcomes of patients with sternal wound infections: A meta-analysis of non-randomized studies. PLoS One, 2013; 8:e64741.

10.

Salica

, Weltert

, Scaffa

, et al. Negative pressure wound treatment improves Acute Physiology and Chronic Health Evaluation II score in mediastinitis allowing a successful elective pectoralis muscle flap closure: Six-year experience of a single protocol. J Thorac Cardiovasc Surg, 2014; 148:2397–2403.

11.

Schneiter

, Grodzki

, Lardinois

, et al. Accelerated treatment of postpneumonectomy empyema: A binational long-term study. J Thorac Cardiovasc Surg, 2008; 136:179–185.

12.

Perentes

, Abdelnour-Berchtold

, Blatter

, et al. Vacuum-assisted closure device for the management of infected postpneumonectomy chest cavities. J Thorac Cardiovasc Surg, 2015; 149:745–750.

13.

Haghshenasskashani

, Rahnavardi

, Yan

, McCaughan

. Intrathoracic application of a vacuum-assisted closure device in managing pleural space infection after lung resection: Is it an option?. Interact Cardiovasc Thorac Surg, 2011; 13:168–174.

14.

Klein Klouwenberg

, Frencken

, Kuipers

, et al. Incidence, predictors, and outcomes of new-onset atrial fibrillation in critically ill patients with sepsis. A cohort study. Am J Respir Crit Care Med, 2017; 195:205–211.

15.

Moss

, Calland

, Enfield

, et al. New-onset atrial fibrillation in the critically Ill. Crit Care Med, 2017; 45:790–797.

16.

O'Connor

, Kells

, Henry

, Scalea

. Vacuum-assisted closure for the treatment of complex chest wounds. Ann Thorac Surg, 2005; 79:1196–1200.

17.

Sziklavari

, Ried

, Neu

, et al. Mini-open vacuum-assisted closure therapy with instillation for debilitated and septic patients with pleural empyema. Eur J Cardiothorac Surg, 2015; 48:e9–e16.

18.

de Abreu

, Pontes

, Tamagno

, et al. Treatment of thoracic wounds with adapted vacuum therapy. Asian J Surg, 2014; 37:49–52.

19.

Begum

, Papagiannopoulos

. The use of vacuum-assisted wound closure therapy in thoracic operations. Ann Thorac Surg, 2012; 94:1835–1839.

20.

Palmen

, van Breugel

, Geskes

, et al. Open window thoracostomy treatment of empyema is accelerated by vacuum-assisted closure. Ann Thorac Surg, 2009; 88:1131–1136.

21.

Saadi

, Perentes

, Gonzalez

, et al. Vacuum-assisted closure device: A useful tool in the management of severe intrathoracic infections. Ann Thorac Surg, 2011; 91:1582–1589.

22.

Groetzner

, Holzer

, Stockhausen

, et al. Intrathoracic application of vacuum wound therapy following thoracic surgery. Thorac Cardiovasc Surg, 2009; 57:417–420.

23.

Al-Mufarrej

, Margolis

, Tempesta

, et al. Outpatient management of post-pneumonectomy and post-lobectomy empyema using the vacuum-assisted closure system. Surg Today, 2010; 40:711–718.

24.

Diefenbeck

, Mennenga

, Guckel

, et al. [Vacuum-assisted closure therapy for the treatment of skin and soft-tissue infections. Are wound specimens of use in planning secondary wound closure?] (Ger). Z Orthop Unfall, 2011; 149:324–329.

25.

Osmani

, Sotello

, Alvarez

, Odell

, Thomas

. Mycobacterium abscessus infections in lung transplant recipients: 15-Year experience from a single institution. Transpl Infect Dis, 2018; 20:e12835.