Risk factor of postoperative incision infection after plate internal fixation of calcaneal fractures: a retrospective study

In this study, infection was classified into superficial infection and deep infection according to the US Centers for Disease Control and Prevention criteria [4].

A retrospective analysis was performed for patients with calcaneal fractures who received plate internal fixation surgery via a traditional lateral L-shaped approach at our hospital from January 2012 to June 2018. Inclusion criteria: 1) age ≥ 18 years; 2) patients with Sanders II, III, and IV calcaneal fractures. Exclusion criteria: 1) patients with Sanders I calcaneal fractures; 2) patients who received fixation via other approaches; 3) patients who received other internal fixation; 4) patients with other complicating fractures; 5) patients with severe neurovascular injuries in the operated side lower extremity; 6) patients with local or systemic infection before surgery; 7) patients with liver, renal or autoimmune diseases.

All the enrolled patients were followed up for at least 1 year, with a mean follow-up time of 15.8 months. The patients were divided into an infection group (n = 24; 7.9%) and a noninfection group (n = 278; 92.1%). There were 10 cases of deep infection in the infection group, and their incisions were well healed after thorough debridement and flap coverage without serious complications such as fracture nonunion or talocalcaneal joint fusion. After daily cleaning, dressing changes, red light physiotherapy, and prolonged antibiotic use, 14 patients with superficial infections did not develop deep infections, and their incisions were well healed. The univariate analysis results showed significant differences concerning the incidence of preoperative tension blisters, smoking history, history of diabetes, preoperative haemoglobin, preoperative albumin, time from injury to surgery, and operation time between the two groups (p < 0.05). No significant differences were observed regarding age, sex, cause of injury, fracture classification, anaesthesia method, or intraoperative bone grafting between the two groups (p > 0.05) (Table 1).

A history of diabetes, preoperative haemoglobin, preoperative albumin, the time from injury to surgery, and the operation time were listed as independent variables, and infection was regarded as the dependent variable. The results of binary logistic regression analysis showed that a shorter time from injury to surgery, longer operation time, and lower preoperative albumin were risk factors for postoperative incision infection (Table 2). Forest plots of regression analysis results were plotted to indicate the effect size of each factor (Fig. 1).

Receiver operating characteristic (ROC) curves were plotted to estimate the clinical threshold value of each risk factor. The results demonstrated that the clinical threshold value of the time from injury to surgery was 10.5 days with a sensitivity of 91.7% and specificity of 58.3%, the clinical threshold value of operation time was 84.5 minutes with a sensitivity of 86.0% and specificity of 87.5%. The preoperative albumin was 38.5 g/L with 68.3% sensitivity and 91.7% specificity 91.7% (Fig. 2).

Existing studies have shown that surgical treatment of displaced calcaneal fractures can achieve favourable therapeutic effects [6]. Still, it correlates with an incidence of postoperative incision infection up to 15.4% [7‐9], which is one of the most common complications after internal fixation of calcaneal fractures. Although there have been various surgical approaches for internal fixation of calcaneal fractures in the past 30 years, the lateral L-shaped extension approach remains the most frequently applied method for internal fixation [10]. Through this surgical approach, the soft tissue capsule is opened to expose the fracture site as well as the collapsed and compressed articular surface. Upon poor postoperative incision healing, cracking and exposure of deep structures, including bone, deep fascia, and implants, may develop until infection, resulting in devastating outcomes. The indication for patients with deep infection of whether to remove the internal fixation device is based on the invaded severity of the infection. If a sinus tract is discovered deep to the bone surface, the internal fixation should be removed to avoid residual biofilm infection on the plate surface. Otherwise, the infection can be controlled by adequate debridement and effective antibiotic treatment. Current studies have illustrated nonsurgical factors related to incision infection after internal fixation of calcaneal fractures (higher body mass index, history of diabetes, smoking history, etc.) and surgical-related factors (intraoperative lateral peroneal artery injury, earlier surgical intervention, inexperienced operators, etc.) [11‐14]. Yet, most of them involve univariate analysis. Therefore, we conducted a retrospective study incorporating multiple factors by logistic regression analysis. Our study offered data demonstrating an infection rate of 7.9% following internal fixation of calcaneal fractures. The time from injury to operation < 10.5 days, preoperative albumin value < 38.5 g/L, and operation time > 84.5 minutes were determined to be risk factors for postoperative incision infection (p < 0.05).

Axial violence represents the most common damage mechanism of calcaneal fractures. However, due to the special anatomical structure around the calcaneus, high-energy trauma and relatively thin surrounding soft-tissue coverage can lead to severe soft-tissue damage and swelling, so preoperative soft-tissue management and appropriate timing of surgery are particularly important. Early surgical treatment, especially surgical intervention within 3 days following injury, may lead to aggravation of swelling, poor postoperative incision healing and even infection, and other complications when soft tissue swelling is at its peak. For more swollen patients, heel ice compress and elevation of the affected limbs are applied to alleviate the swelling, and necessary drugs such as 20% mannitol are given to reduce swelling. It should be noted that 1-10% of patients with calcaneal fractures may develop compartment syndrome [15]. If the patient shows persistent and aggravated heel pain, soft-tissue swelling, and loss of skin sensation, early incision decompression is required to reduce sequelae. If tension blisters occur around the incision, we try to retain the blister wall since the blister fluids contain opsonin and immune cells, exerting anti-infection effects; at the same time, the cyst wall is a natural wound dressing that can protect the blister bed from the outside. The fluid in the large blisters should be aspirated after thorough disinfection, and then a sterile elastic bandage should be utilized for pressure dressing. The univariate analysis in the study showed no significant difference in the incidence of tension blisters between the two groups (54.17% vs. 52.16%, p = 0.850), which was closely related to our preoperative aseptic management for blisters. Regarding the timing of surgery, it is currently advocated that surgery performed after the skin folds on the heel causes low rates of complications [16]. However, Ho et al. thought that an experienced surgeon can beperform early surgery for patients who have no potential risk factors for incision infection, which will not increase the incidence of incisional complications [13]. The appropriate time for surgical intervention is the presence of skin folds rather than the complete elimination of soft tissue oedema because it may take a long time to fully eliminate soft tissue oedema, and haematoma mechanization around fracture may affect the operating procedures during the waiting time for detumescence. In this study, the duration from injury to surgery was significantly shorter in the infection group than in the noninfection group (OR = 1.475; 95% CI: 1.024-2.125; p = 0.037). The ROC curves showed 10.5 days as the clinical threshold of the duration from injury to surgery, with a sensitivity of 91.7% and specificity of 58.3%.

The length of operation is closely related to adequate preoperative preparation, operative experience, and intraoperative cooperation. The surgical treatment of calcaneal fractures is also an invasive method that induces certain damage to the soft tissues. The lateral L-shaped extended incision that requires opening the lateral calcaneal flap and full exposure of the fracture site is utilized for patients with Sanders type III or IV fractures to obtain a more intuitive surgical perspective. However, a longer operation time and a longer time of pulling the skin flap will indirectly affect postoperative incision healing, even leading to incision infection. The study conducted by Koski et al. noted that the long duration of surgery is a dangerous factor for postoperative incision infection [17]. Similarly, the work of Al-Mudhaffar et al. showed that the operation time of patients with postoperative incision infection after calcaneal fractures was longer than that of noninfected patients, showing a mean increase of 39 minutes [18]. Zhou et al. suggested that tourniquet use during surgery can shorten the operation time by 4.8 minutes [19]. Consistently, Zhang et al. demonstrated that using a tourniquet leads to a mean reduction of 4.6 min in operation time [20]. Therefore, a lower extremity balloon tourniquet was utilized during complex calcaneal fracture surgery in our hospital to ensure a clear surgical field of vision, reduce intraoperative bleeding, and shorten the operation time. Compared to the noninfection group, the operation time increased by 17.49 minutes in the infection group (OR = 1.511; 95% CI: 1.219-1.874; p < 0.001). Meanwhile, our ROC curves revealed that the clinical cut-off value of the operation time was 84.5 minutes, and its sensitivity and specificity were 86.0 and 87.5%, respectively.

The negative nitrogen balance induced by trauma and surgery is proportional to the degree of injury [21], so normal nutritional status is essential to maintain the healing ability of skin and soft tissues and to prevent infection [22]. The albumin level represents the clinical haematological indicator most directly linked to the nutritional status of patients. Albumin is regarded as one of the major components of human plasma, and malnutrition is usually defined as an albumin level under 3.5 mg/dl [23]. Bohl et al. found that malnutrition evoked by hypoalbuminemia could affect collagen synthesis and reduce fibroblast proliferation, resulting in poor wound healing [24]. Yi et al. considered that hypoalbuminemia would attenuate the inflammatory response against infection, thus affecting incision healing [25]. In this study, the mean albumin value showed a marked reduction of 7.582 g/L in the infection group compared to the noninfection group (OR = 1.559, 95% CI: 1.191-2.041, p = 0.001), and the results of the ROC curves revealed a clinical cut-off value of 38.5 g/L, and its sensitivity and specificity were 68.3 and 91.7%, respectively.

Incision healing is highly correlated with a constant and sufficient oxygen supply. Haemoglobin constitutes the main carrier of oxygen and carbon dioxide in the human body. Oxygen can offer bioenergy, such as adenosine triphosphate, catalyse the production of reactive oxygen species, and mediate collagen components, exerting a crucial role in the incision healing process [26]. High haemoglobin levels can ameliorate hypoxia, reduce hypoxia-induced apoptosis, and promote wound healing, effectively preventing wound necrosis and infection. The univariable analysis indicated a lower haemoglobin level in the infection group (p < 0.001). Yet, the regression analysis exhibited no statistical correlation between the low haemoglobin level and incision infection (p = 0.196).

Recent studies have suggested that diabetic patients may develop pathological stenosis and occlusion in the blood vessels, resulting in poor blood supply in distal limbs, increased blood viscosity, and impaired red blood cell oxygen transport. Additionally, diabetes limits the function of human immune cells, leading to impaired innate and adaptive immune responses, which increases the risk of various infections [27, 28]. Endara et al. revealed a significant association between blood glucose and postoperative incision infection, and a positive correlation between good perioperative glycaemic control and wound healing in diabetic patients who underwent surgery [29]. If the incision does not heal well after calcaneal fractures, there is a great risk of incision infection, inducing devastating outcomes. Therefore, this imposes an extreme challenge on the treatment of diabetic patients with calcaneal fractures. Rammelt et al. proposed severe insulin-dependent diabetes as a contraindication for open reduction and internal fixation [30]. The univariate analysis of this study demonstrated that a history of diabetes was a risk factor for postoperative incision infection (p = 0.03). However, the regression analysis offered data showing no statistical relationship between a history of diabetes and incision infection (p = 0.178), which may be associated with good perioperative glycaemic control. It is of great necessity to control the preoperative fasting blood glucose level below 10.0 mmol/L. If the effect of oral drugs is limited, insulin injection can be adopted for blood glucose control.

The results of this study did not suggest intraoperative implantation of allografts to be a risk factor for incision infection after internal fixation of calcaneal fractures. Intraoperative bone grafting is mainly used to provide mechanical support and induce osteogenesis, but whether intraoperative bone grafting is necessary remains controversial. According to the study of Singh et al., although patients with bone grafts show advantages in terms of Böhler angle and time of full weight-bearing, no significant difference is observed in functional outcomes and complications [31]. A meta-analysis conducted by Pan et al. suggested that intraoperative bone grafting increased the risk of postoperative wound infection and other complications (OR = 1.74, p < 0.01) [32]. Our univariate analysis results showed that allogeneic bone grafting during surgery did not increase the risk of postoperative wound infection (p = 0.172), which potentially correlated with more stable fracture fixation, reduced fracture micromotion, reduced exudation at the fracture site, and a low risk of subcutaneous haematoma after bone grafting. Subcutaneous haematoma is one of the principal causes of postoperative incision infection. Because bone grafting does not increase the incidence of infection in open reduction and internal fixation of calcaneal fractures, we consider that the use of bone grafting during surgery should be determined depending on the actual bone defect.

This study also has some limitations. First, a retrospective study offers a relatively low level of evidence, there may have some inevitable management differences among these cases. In addition, multiple factors are involved in incision infection following calcaneal fracture surgery. However, this study did not discuss other potential risk factors, such as different surgical approaches, incision suture methods, intraoperative bleeding, implant species, etc. Furthermore, differences may exist in the degree of trauma and treatment process for the calcaneal fractures of different Sanders types. Thus, a large-scale, randomized, controlled study io different types of calcaneal fractures is still necessary for clarification of the relevant risk factors.