Comparing vacuum‐assisted closure against conventional approach in severe deep neck infection: A retrospective case‐control study

Abstract

Objectives

To compare the effectiveness of vacuum‐assisted closure (VAC) against traditional drainage technique, using a retrospective case‐control study design, in terms of managing deep neck infections (DNIs).

Methods

Patients presenting to Peking Union Medical College Hospital diagnosed with DNIs were recruited in this study. We analyzed the clinical characteristics of DNI patients and divided them into (a) VAC placement group (26 cases) and (b) traditional drainage group (57 cases) according to whether VAC was placed. The differences in length of stay (LOS), wound healing time, and debridement frequency were compared between the two groups.

Results

Eighty‐three patients had multiple‐space infections, i.e. infection at two or more sites. The debridement frequency of the VAC group was significantly lower than that of the traditional drainage group (p = 0.001). The wound healing time of the traditional drainage group and VAC group was 38 days (a range of 13–98 days) and 40 days (a range of 11–106 days), respectively; the average LOS was 15 days (a range of 2–68 days) and 16 days (a range of 4–35 days), respectively; and the debridement frequencies were one time (a range of 0–3 times) and zero times (a range of 0–2 times), respectively. The two groups did not differ significantly in wound healing time and hospitalization duration (p = 0.319 and 0.937).

Conclusions

VAC treatment of DNIs has significant advantages in reducing the frequency of debridement and patient suffering, but it does not show significant advantages in wound healing. Randomized trials are still needed to demonstrate its efficacy.

Key Points

• Deep neck space infections (DNIs) are potentially life‐threatening conditions, which require consistent wound care during relatively long periods of hospitalization.

• When comparing with the conventional surgical intervention, vacuum‐assisted closure placement in DNIs lowers the frequency of debridement significantly, though little does it contribute to the shortening of wound healing time and hospitalization.

• Streptococcus pharyngitis is the most common pathogen isolated from the drainage in both groups.

INTRODUCTION

Deep neck spaces are real or potential compartments between the cervical fascial layers, which are filled with loose connective tissue. Deep neck infections (DNIs) are suppurative infections occurring in the deep neck spaces, which often include the parotid and masticatory space, submandibular and submental space, para‐pharyngeal space, retropharyngeal space, prevertebral space, and mediastinum. ¹ Those spaces not only extend from the skull base to the mediastinum, but also communicate with each other, providing avenues for infections to spread and progress to life‐threatening events. ² , ³ Common clinical manifestations include sore throat, firm tenderness, and/or trismus, dysphagia, etc. When infections press locally on airway or spread into mediastinum, stridor and dyspnea may appear. Asides from blockage to airway, leading causes of death are sepsis, multiple organ dysfunction in case of severe uncontrollable DNIs. ⁴

The management principles of DNIs include life support, surgical drainage, and the application of appropriate antibiotics, among which surgical drainage is a crucial component ¹ , ² , ³ , ⁴ , ⁵ ; hyperbaric oxygen therapy has also been reported to have adjuvant effects. ⁶ The traditional surgical method is to make extensive incisions and place one (or multiple) drainage tube(s) at a lower position to the abscess, to ensure adequate drainage of the pus driven by the gravity effects. Subsequent dressing changes and/or irrigation are required daily until the infection is under control, and the incision wounds have healed. ⁷ Vacuum‐assisted closure (VAC) is an alternative treatment for such wound. By establishing a local negative pressure environment, VAC suctions out the wound discharge while promoting the formation of repairing granulation tissue. ⁸ VAC is also known as negative pressure wound therapy; the two terms are interchangeable. ⁹ In recent years, the successful application of VAC in managing DNIs has been increasingly reported. ¹⁰ , ¹¹ , ¹² However, the effectiveness of VAC in such scenario was mostly reported by cases. Currently, there is no systemic study reviewing whether VAC has any advantages over conventional surgical drainage in the treatment of DNIs. The objective of our retrospective study was to evaluate if VAC was superior to traditional drainage in treating DNIs.

METHODS

Patients

This was a retrospective study approved by the ethics committee of Peking Union Medical College Hospital (approval number: JS2084). Signed informed consent was collected before surgery. All adult (>18 years old) patients with DNIs who underwent incision drainage treatment in our hospital between January 2015 and December 2020 were recruited. Therapeutic efficacy was evaluated by length of stay (LOS), wound healing time, and debridement frequency.

Traditional drainage methods typically start with extensive cervical incisions. The deep neck spaces were then separated bluntly to fully expose the infected area. Necrotic tissue was debrided followed by repeated, thorough irrigation. Subsequently, one to multiple drainage tubes were placed deep into the cavity at the possibly lowest site. As precautions against airway obstruction, tracheotomy was often performed depending on the risk. Postoperatively, the dressing was changed once or twice daily. In the VAC group, the VAC was placed immediately or sometime after the traditional incision and drainage had been applied. The surgical procedure included (1) placement of an open‐cell foam tailored to the infected area, integrated with traditional debridement and irrigation; (2) adhesion of a film covers to ensure airtightness; (3) connection to the VAC device (V.A.C.®, Kinetic Concepts Inc.) (4) negative pressure was set to 125–150 mmHg. The foam material could be replaced after the operation according to the drainage conditions, with no need to change the dressing otherwise. The purulent exudate sampled during the surgery of both groups was sent for a microbiological culture test.

Indicators for evaluating the therapeutic efficacy

Surgical debridement frequency was defined as the number of debridement in the operating room for the traditional group (the bedside debridement was excluded), and as the total number of debridement and dressing changes in the operating room after implantation of VAC for the VAC group. Besides, LOS, wound healing time, and debridement frequency were also included as evaluation indicators of the therapeutic efficacy. LOS was defined as the time duration between the first visit at either the emergency room or ambulatory center and discharge from Peking Union Medical College Hospital (PUMCH). Wound healing time was calculated as the time lapse, started with the time when the first surgical intervention was performed, ended with the moment when any of these following important events took place: full recovery in the absence of any associated symptoms, complete healing of the incision with suture removed, removal of the drainage tube(s), and decannulation in case of tracheostomy.

Statistical analysis

Statistical analyses were performed using SPSS v17.0 statistical software (SPSS, Chicago, IL). Quantitative data that did not follow a normal distribution were described as median and interquartile range and were compared using the Wilcoxon rank sum test. Probability values less than 0.05 were considered statistically significant.

RESULTS

General information

All 86 patients were referred to PUMCH from other clinics or local hospitals and did not undergo surgical drains before admission. Three patients (3.48%) died during the treatment. The causes of death included multiple organ failure due to airway obstruction in one case, and to disseminated intravascular coagulation (DIC) in the other two cases. These three cases were excluded from the subsequent wound‐healing evaluation. Therefore, a total of 83 patients were included in the study. There were 57 patients in the traditional drainage group and 26 patients in the VAC group. The two groups did not differ significantly in general information. General information included average ages, male‐to‐female ratios, body mass indices, the prevalence of diabetes, other complicated conditions (including coronary heart disease, previous myocardial infarction, renal failure, severe asthma, stroke, liver cirrhosis, autoimmune diseases, Guillain‐Barré syndrome, and pregnancy), performance of tracheotomy, and occurrence of mediastinal infection (Table 1).

Table 1.

General information of deep neck infections patients.

Demographic parameters	Traditional drainage (n = 57)	VAC (n = 26)	p‐Value
Age [year, mean (range)]	59 (18–79)	58 (41–79)	0.996
Male/Female (n)	39/18	21/5	0.298
BMI [kg/m2, mean (range)]	25.95 (19.03–36.23)	25.45 (17.99–35.16)	0.757
Diabetic [n(%)]	25 (43.86)	15 (57.69)	0.811
Other complications [n(%)]	18 (31.58)	10 (38.46)	0.619
Tracheotomy [n(%)]	33 (57.89%)	14 (53.85)	0.631
Mediastinal infection [n(%)]	18 (31.58)	9 (34.61)	0.805

Abbreviations: BMI, body mass indices; VAC, vacuum‐assisted closure.

Spaces involved and bacteriological results

The causes of infections in the patients of this study included odontogenic infections, nonodontogenic infections, uncertain infections, and trauma. Based on the causes, the cases were further divided into four subgroups, with each subgroup containing 22/14/20/1 (traditional drains) and 13/5/8/0 (VAC), respectively. The extent to which the deep neck spaces were affected, was determined by preoperative radiological findings and intraoperative conditions. Infection in all patients involved two or more anatomic sites. The most affected spaces in the traditional drainage group and the VAC group were the para‐pharyngeal space [52 (91.23%), 24 (92.31%)], and the submandibular and submental space [48 (84.21%), 25 (96.15%)], respectively (Table 2).

Table 2.

Spaces involved in traditional drainage groups and VAC group [n(%)].

Cervical spaces involved	Traditional drainage (n = 57)	VAC (n = 26)	p‐Value
Odontogenic/nonodontogenic/uncertain/trauma (n)	22/14/20/1	13/5/8/0
Parotid and masticatory space	19 (33.33)	14 (53.85)	0.094
Submandibular and submental space	48 (84.21)	25 (96.15)	0.160
Para‐pharyngeal space	52 (91.23)	24 (92.31)	0.870
Retropharyngeal space	23 (40.35)	14 (53.85)	0.341
Prevertebral space	9 (15.79)	6 (25.00)	0.540

Abbreviation: VAC, vacuum‐assisted closure.

The bacteriological culture test results were available in more than 90% of the cases, showing that Streptococcus was the most common pathogen. The number of cases with single or multiple bacterial infections of traditional drainage group was 40 (70.18%) and 14 (24.56%), respectively; in VAC group, there were 20(76.92%) and 5(19.23%). No significant difference between the two groups (p = 0.436, 0.741) was observed (Table 3), either. A detailed pathogenic spectrum is described in Table 4.

Table 3.

Bacteriological study results [n(%)].

Number of pathogens	Traditional drainage (n = 57)	VAC (n = 26)	p‐Value
Isolated pathogens	40 (70.18)	20 (76.92)	0.436
Multiple pathogens	14 (24.56)	5 (19.23)	0.741

Abbreviation: VAC, vacuum‐assisted closure.

Table 4.

Pathogenic spectrum of DNIs and distribution of culture samples [n(%)].

Category	Traditional drainage (n = 57)	VAC (n = 26)	p‐Value
Gram‐positive bacteria	37 (64.91)	17 (65.38)	0.13
Streptococcus pharyngitis	19 (51.35)	8 (47.06)
Streptococcus constellatus	10 (27.03)	5 (29.41)
Streptococcus species	5 (13.51)	2 (11.76)
Staphylococcus species	1 (2.70)	2 (11.76)
Lactobacillus	2 (5.41)	0
Gram‐negative bacteria	12 (21.11)	5 (19.23)	0.81
Klebsiella pneumoniae	4 (33.33)	2 (40.00)
Acinetobacter baumannii	3 (25.00)	2 (40.00)
Escherichia coli	2 (16.67)	0
Pseudomonas aeruginosa	1 (8.33)	1 (20.00)
Stenotrophomonas maltophilia	2 (16.67)	0
Fungi	3 (5.26)	2 (7.69)
Candida albicans	3 (100)	2 (100)

Note: The rest of the bacteria culture tests reported as “no growth in 24 or 48 h.”

Abbreviations: DNIs, deep neck infections; VAC, vacuum‐assisted closure.

Efficacy evaluation

The debridement frequency of VAC group was 0 times (0–2 times), and that of traditional drainage group was 2 times (1–4 times); VAC group was significantly better than traditional drainage group (p = 0.001).

The average wound healing time of the traditional drainage group and VAC group was 38 days (a range of 13–98 days) and 40 days (a range of 11–106 days) respectively; the average LOS was 15 days (a range of 2–68 days) and 16 days (a range of 4–35 days), respectively; and the debridement frequencies were one time (a range of 0–3 times) and zero times (a range of 0–2 times), respectively. The two groups did not differ significantly in wound healing time and LOS (p = 0.319, 0.937) (Figure 1).

Figure 1.

Comparison of therapeutic effectiveness between the traditional approach and the vacuum‐assisted closure (VAC) implantation groups. No intergroup differences were found statistically significant in terms of wound healing time (A) and length of stay (B); the latter had a significantly reduced frequency of debridement (C).

However, further grouping the patients into immediate and delayed VAC implantation, allowed for critical insights into the importance of VAC to DNIs. There was a statistically significant difference between these three groups (traditional drainage, immediate, and delayed VAC) as determined by one‐way ANOVA (F(2,72) = 3.9, p < 0.05). The average wound healing time of the VAC immediate implantation group was significantly shorter than the delayed implantation group [22(10.5) against 45(22.3)] in terms of days. Yet, implanting VAC immediately after the traditional approach did not distinguish significantly in shortening average LOS or reducing debridement frequencies (Figure 2).

Figure 2.

Comparison of therapeutic effectiveness among three groups, the traditional drainage approach, the vacuum‐assisted closure (VAC) group with immediate implantation, and the VAC group with delayed implantation. Statistically significant differences were found in only wound healing time (A), yet not in length of stay (B) or frequency of debridement (C).

Severe complications such as mediastinitis occurred in 26 out of 83 patients (31.3%). No significant differences were observed between the two groups, either, in terms of LOS, frequency of debridement, and wound healing time (Table 5).

Table 5.

Evaluation indicators in patients complicated with mediastinitis.

Evaluation indicators	Traditional drainage (n = 17)	VAC (n = 9)	p‐Value
Length of stay [day, mean (SD)]	27.9 (13.4)	20 (9.0)	0.17
Frequency of debridement [times, mean (SD)]	2 (1.44)	2 (0.76)	0.21
Wound healing time [day, mean (SD)]	36.6 (18.9)	47.3 (17.4)	>0.99

Abbreviation: VAC, vacuum‐assisted closure.

DISCUSSION

Mechanistically, VAC can suction out exudates, blood, and wastes from infected wounds driven by local negative pressure, which then promotes reparative granuloma formation and wound healing. ¹³ But these effects have not been corroborated in laboratory findings. ¹⁴ Govea‐Camacho et al. ¹⁵ reported a case‐control study of VAC in the treatment of DNIs and suggested that VAC could shorten wound healing time. However, this study only included 18 patients. Balcı et al. ¹⁰ performed VAC treatment on 11 patients and concluded through observation that VAC could play an important role in the treatment of DNIs, but this study did not include any specific evaluation index. Chen et al. ¹¹ and Cao et al. ¹⁶ reported similar results after treatment and observation of 7 and 12 DNI patients, respectively, concluding that VAC has many advantages and leads to good clinical effects compared to traditional debridement and drainage surgery. Nonetheless, this conclusion derived only from case observation without comparison with the results of the control group. Figures 3, 4, 5 demonstrate two previous DNIs cases treated with VAC, lending support to the clinical effectiveness of VAC approach.

Figure 3.

Vacuum‐assisted closure (VAC) significantly improved the healing of Deep neck space infection wounds. (A) The wound condition of neck infection before treatment. (B) VAC treatment of the wound. (C) The wound condition after the VAC device was removed.

Figure 4.

Vacuum‐assisted closure (VAC) significantly improved the healing of Deep neck space infection wounds. (A) The wound condition of neck infection before treatment. (B) Intraoperative VAC treatment of the wound. (C) The wound condition after the VAC device was removed.

Figure 5.

Depicts imaging findings of patients with multifocal infections in the maxillofacial and neck regions, indicating that administration of VAC therapy has significantly reduced neck infection, gas accumulation, and tissue swelling. (A) Preoperative CT scan shows multifocal abscesses in the maxillofacial region with tissue edema (indicated by the red arrow). (B) One week after VAC placement, the CT scan reveals significant resolution of the abscesses and tissue edema (indicated by the red arrow). (C) Infection in the supraclavicular fossa of neck marked by hypodense area and gas accumulation (indicated by the red arrow). (D) One week after VAC placement, there is significant resolution of infection, gas accumulation, and tissue edema in the supraclavicular fossa (indicated by the red arrow).

Based on these observations, our retrospective analysis should also conclude that VAC can achieve better efficacy in the treatment of DNIs. However, this is not the case. In the comparison of the two indicators of LOS and wound healing time, we did not find any advantage of the 26 patients who underwent VAC implantation over the traditional drainage group on these two indicators. Due to the lack of reports on the treatment of DNIs by VAC, we expanded the scope of analysis to reports on other applications of VAC to explain our finding. The meta‐analyses by Norman et al. ¹⁷ and Shiroky et al. ¹⁸ showed that the application of VAC treatment in closure surgery did reduce the risk of surgical site infection during surgery. Nevertheless, the target disease of these studies was noninfected wounds, and hence, the evidence seemed too feeble to support this conclusion. In the manipulation of infected or contaminated wounds, the meta‐analysis by Iheozor‐Ejiofor et al. ¹⁹ demonstrated evidence of moderate strength that there was no difference in the wound healing rate between VAC treatment and traditional treatment in terms of treating open injuries. The level of evidence presented by Liu et al. ²⁰ was also not sufficiently high, concluding that VAC could shorten the healing time for the treatment of diabetic foot is also low. To sum up, the role of VAC in reducing the healing time of infected wounds, was uncertain, which agreed to the findings of both the prior studies and ours. One possible reason why VAC cannot accelerate the wound healing might be that a closed negative pressure system creates hypoxic environment, which is unfavorable for tissue repair due to less applicable blood supply. This might have undermined some of the advantages brought by VAC.

A remarkable discovery of our study was that the number of debridement in the VAC group was significantly reduced, indicating that VAC implantation is advantageous to relieve patients' pain and discomfort during postoperative care. In a randomized controlled study, Seidel et al. ²¹ found that the pain assessment scores of patients in the VAC treatment group were significantly better than those of the traditional treatment group. Additionally, in a randomized controlled study on suppurative infections in children, Yang et al. ²² found that the pain condition of children in the VAC group was significantly better than that in the traditional treatment group. These findings are consistent with our results.

Our study has some limitations. First, minor misinterpretation of the efficacy of VAC implants might derive from analysis of data without identifying whether VAC was implanted in an immediate or a delayed manner. The subtle reduction of wound healing time encouraged future endeavors into applying the VAC at an earlier stage, which is still in need of further supportive evidence. Plus, although there was a control group in this study, it was limited by the characteristics of retrospective studies, and there may have been bias in the case selection. Furthermore, despite that pathogens of DNIs are often mixtures of aerobic and anaerobic bacteria, a considerably high ratio of anaerobic infections was underreported due to the demanding culture conditions and a lack of special culture medium in the emergency operating room 6 to 8 years ago, when we started to enroll cases. Therefore, we administered metronidazole to each case regardless of the pus culture testresults. Plus, a certain portion of the culture tests failed to provide detailed information on the pathogenic strains, which might lead to a less comprehensive analysis of the results. Finally, we did not include cost‐effectiveness analysis as an observation indicator.

CONCLUSION

The results of our retrospective case‐control study showed that VAC treatment of DNIs had significant advantages in reducing debridement frequency and patient suffering. Yet, no obvious advantage in wound healing was noticed over the traditionally treated group. Randomized trials are still needed to verify the efficacy of VAC in managing DNIs.

AUTHOR CONTRIBUTIONS

Mu Wang and Jian Wang were involved in all aspects of study design, data collection, and manuscript preparation. Rui‐Zhe Yang and Wei Gu were involved in the statistical analysis and manuscript drafting. Rui‐Zhe Yang and Wei Gu assisted with data collection. All authors read and approved the final manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

The study protocol was approved by the local Institutional Review Board (Peking Union Medical College Hospital Ethics Committee, JS2084) in accordance with the Helsinki Declaration. The requirement for obtaining informed consent has been waived.

Wang M, Yang R‐Z, Gu W, Wang J. Comparing vacuum‐assisted closure against conventional approach in severe deep neck infection: A retrospective case‐control study. World J Otorhinolaryngol Head Neck Surg. 2025;11:425‐432. 10.1002/wjo2.219

DATA AVAILABILITY STATEMENT

The data set used and/or analyzed in this current study are available from the corresponding authors upon reasonable request.