Risk factor analysis and severity assessment of wound hematoma after anterior cervical spine surgery: A retrospective study

We aim to explore the risk factors independently associated with postoperative wound hematoma in patients who have undergone anterior cervical spine surgery. Methods The clinical data of patients with cervical spondylosis or cervical disc herniation who underwent anterior cervical spine surgery by the senior author from January 2011 to December 2017 were evaluated. Multivariate logistic regression was conducted to compare the hematoma group and the no-hematoma group to determine which factors were independently associated with hematoma formation in patients who need evacuation. The Mann-Whitney U test was conducted to compare the Neck Disability Index score in the two groups.


Abstract Background
We aim to explore the risk factors independently associated with postoperative wound hematoma in patients who have undergone anterior cervical spine surgery.

Methods
The clinical data of patients with cervical spondylosis or cervical disc herniation who underwent anterior cervical spine surgery by the senior author from January 2011 to December 2017 were evaluated.
Multivariate logistic regression was conducted to compare the hematoma group and the no-hematoma group to determine which factors were independently associated with hematoma formation in patients who need evacuation. The Mann-Whitney U test was conducted to compare the Neck Disability Index score in the two groups.

Results
A total of 678 patients met the criteria and underwent anterior cervical spine surgery. Thirteen patients undergone hematoma evacuation. Multivariate logistic regression analysis identi ed that history of hypertension (p = 0.039 OR = 4.42 95% CI 1.08-18.07) and therapeutic heparin use (p = 0.020 OR = 4.58 95% CI 1.27-16.59) were independent risk factors for hematoma formation. The t-test showed no signi cant differences between the hematoma group and the no-hematoma group in terms of APTT or PT levels (p > 0.05). The Mann-Whitney U test indicated that there was no difference in NDI scores between the two groups(p > 0.05).

Conclusion
History of hypertension and therapeutic heparin use are risk factors for hematoma formation. Meticulous hemostasis, moderate muscle subtraction, and perioperative airway management are critical for avoiding hematoma development. The Neck Hematoma Scores can quickly determine the severity of a hematoma in the absence of radiographic image evidence.

Background
The past decades have seen increasingly rapid advances in the eld of anterior cervical spine surgery (ACSS). In 1958, Smith and Robinson rst proposed anterior cervical discectomy and fusion (ACDF) [1].
After that, anterior cervical corpectomy and fusion (ACCF) and anterior cervical discectomy and arthroplasty (ACDA) were successively presented. Most clinical studies have con rmed the effectiveness of these procedures [2][3][4]. Moreover anterior cervical surgery is minimally invasive, the postoperative effects are good. However, because of the anatomical complexity of the anterior cervical approach, with the presence of many vital structures, complications occasionally occur, among which wound hematoma, dysphagia, and recurrent laryngeal nerve injury are the most common [5].
The development of postoperative wound hematoma is a critical complication in the clinic that needs to be discovered and treated in time. If not handled properly, this complication can cause serious consequences, such as airway compromise and neurological de cit, and it can even be life-threatening.
To date, several studies have investigated the risk factors for wound hematoma formation and the results are differ [6][7][8]. Therefore, the present research explores the effects of risk factors and the bleeding sites of wound hematoma formation, in order to minimize the occurrence of this dangerous complication.

Methods
This retrospective study consisted of 678 consecutive patients who underwent anterior cervical spine surgery from January 2011 to December 2017. The criteria of inclusion and exclusion were as follows: Inclusion criteria (1) Patients > 18 years old (2) Patients were diagnosed with cervical spondylosis or cervical disc herniation There were 355 males and 323 females included in our study. The patient data recorded in a database included: sex, age, preoperative symptoms, surgical procedures, surgical levels, low-level segment (C6/7 segment), operative time, blood loss, body mass index (BMI), preoperative comorbidities (hypertension, diabetes), smoking history, therapeutic heparin usage, activated partial thromboplastin time (APTT) and prothrombin time (PT) examination before and after surgery. All patients underwent careful preoperative neurological examinations, routine blood examinations, coagulation examinations, anterior and lateral cervical X-rays, CT scans with 3-dimensional reconstruction, and cervical MRI examinations. Patients who had taken antiplatelet or anticoagulant agents daily received heparin replacement therapy(Enoxaparin Sodium (CLEXANE) 4000iu qd) until ve days prior to surgery. Heparin use was discontinued 12 hours before surgery was performed, and antiplatelet or anticoagulant agents were continued two days after surgery.
All surgeries were performed by the senior author, through the right side of the neck, under general anesthesia and endotracheal intubation. All operations were carried out by the senior surgeon, and hemostasis was performed carefully during the operation. Plate xation(SKYLINE, SLIMLOC and VENTURE anterior cervical plate system), cages(n-HA/PA66 cage), and autografts were used in all surgeries except ACDA. An arti cial disc(Prodisc-C, Prestige LP) was used for ACDA. The hemostatic methods included electrocautery and bipolar electrocautery as well as the use of bone wax, which was used when corpectomy was performed, and hemostatic agents (Surgi o Ethicon), which were applied in the intervertebral space and longus colli. At the end of the operation, a subfascial drainage tube was placed through a separate stab incision and was removed the next day after surgery. All patients with postoperative wound hematoma were included in the case group, while the remaining patients were included in the control group.
The occurrence of postoperative wound hematoma is critical and develops rapidly, and the radiographic images are often lacking. Therefore, it is important to diagnose hematoma quickly and determine whether evacuation is necessary. We made the Neck Hematoma Scores (NHS) according to the patient's symptoms and signs to better evaluation. (Table 1). Closed observation for less than 3 points, 3 to 5 points suture line remove is suggested, more than 5 points evacuation is necessary. We followed all patients and used Neck disability index (NDI) scores to evaluate postoperative neck function. We divided the NDI scores of all patients into three groups according to the corresponding time the preoperative group, early postoperative group (> 3 mon < 12 mon), and late postoperative group (> 12 mon).

Statistics
SPSS software version 24.0 was used for statistical analysis. The measurement data in line with a normal distribution were expressed as the mean ± standard deviation, and the rest were expressed as the median (InterQuartile Range); the counting data were expressed as a percentage. Univariate logistic regression analysis was conducted to analyze the risk factors for wound hematoma. Factors of p < 0.05 were added to multivariate logistic regression analysis. Odds ratios (ORs) and 95% con dence intervals (CIs) were determined when statistically signi cant differences (p < 0.05) were found in multivariate logistic regression. T-tests and Mann-Whitney U tests were employed for other measurement data.

Results
A total of 678 patients who met the inclusion and exclusion criteria were included in the study.Of these patients (Table 2), 354 were male(52.2%) and 324 were female(47.8%). The patients' ages ranged between 25 and 84 years, with a mean age (of 53.7 ± 11.2) years. The BMI ranged between 15.1 and 37.3, with an average of (23.7 ± 2.9). There were 307 cases of radiculopathy (45%), 223 cases of myelopathy (33%), and 148 cases of myeloradiculopathy (22%).There were 305 cases who underwent ACDF (45%), 236 ACCF (35%), 79 ACDA (12%) and 58 who underwent ACDF + ACCF (8%). BMI = body mass index; ACDF = anterior cervical discectomy and fusion; ACCF = anterior cervical corpectomy and fusion; ACDA = anterior cervical discectomy and arthroplasty Thirteen cases(1.9%) of hematoma occurred in the study. All patients were graded with NHS and ten patients underwent evacuation ( Table 3). The earliest hematoma occurred 1 hour after surgery, and the latest appeared 44 hours after surgery, with a mean time of 11 (2.5-37.5) hours. Evacuation surgery was performed cautiously by senior author through the original incision and hematomas were proved in the surgery. During the wound exploration, bleeding sites were found in some patients as follows: longus colli in 3 patients, the longus colli and omohyoid in 1 patient, the platysma in 2 patients, the drainage stab incision in 2 patients, and the esophageal wall in 1 patient, and 1 patients had no distinct bleeding sites. A gelatin sponge was placed in front of the plate after adequate hemostasis and a rubber drainage tube was placed at the incision.  (Fig. 1). Sex (p = 0.089) was a possible risk factor but was not statistically signi cant. There was no signi cance difference in age, BMI, symptoms, surgical procedures, operative time, surgical level, low-level segment (C6/7 segment), diabetes, or smoking history.
All patients underwent examinations for activated partial thromboplastin time (APTT) and prothrombin time (PT) before the operation and immediately after operation. A t-test was conducted, and the results showed no signi cant differences between the hematoma group and the no-hematoma group (Figs. 2-3).

Clinical outcome
Neck Disability Index (NDI) scores were recorded to compare the two postoperative times between the hematoma group and the non-hematoma group. The early follow-up time was 3-12 months; and the late follow-up time was > 12 months. A Mann-Whitney U test was conducted, and there was no signi cant difference (Fig. 4).

Discussion
ACSS is commonly used and widely recognized by spine surgeons. Hematomas formation, however, is the most problematic and devastating complication, and cannot be overemphasized. An initial objective of the study was to identify risk factors that would lead to hematoma formation and to detect hematoma quickly.
The incidence of hematoma was 1.9% in our study, which is in accordance with previous reports (0.2-2.4%) [5,6,9,10]. History of hypertension was a risk factor for hematoma formation, and this nding was consistent with that of Palumbo, who mentioned that the surgical stress reaction and anesthesia could cause increased arterial blood pressure, especially in hypertensive patients [11]. Patient 10 in our study had a hematoma after 1 hour in the postanesthesia care unit (PACU) during extubation (Fig. 5). Coughing during extubation can cause increased venous blood pressure, leading to the occurrence of hematoma.
Poor perioperative blood pressure control will signi cantly increase the incidence of hematoma [12]. Patient 8 had a history of chronic pharyngitis, and severe cough after surgery may have resulted in hematoma formation. Therefore, perioperative airway management is also essential.
O'Neill's study concluded that therapeutic heparin usage was a risk factor for hematoma formation, which is in line with the results of our study. But he did not discuss it much, especially in terms of timing and dosage [10]. In our study, all patients who received heparin replacement therapy discontinued heparin until 12 hours before surgery and continued 2 days after surgery. All hematomas occurred within 48 hours, and a possible explanation for this might be that the leading cause of hematoma formation is the preoperative use of heparin. Perhaps the time of withdrawal of heparin should be advanced and the preoperative monitoring of heparin dosage is also important.
The NHS was based on our clinical experience and analysis of recent literature. Patient 10 is diagnosed with hematoma formation in 5 minutes by using NHS, and the total is 7 points(di culty breathing (2 points), wound swelling (3 points), agitation (2 points)). The picture is taken in the operation room prior to the hematoma evacuation and Consent was obtained from the patient for publication.
Due to anticoagulants will effect APTT and PT levels. Therefore the preoperative and postoperative APTT and PT outcomes were collected from all the patients and differences were compared. The results showed no statistical signi cance.
Many studies have suggested that ossi cation of the posterior longitudinal ligament (OPLL) and diffuse idiopathic skeletal hyperostosis (DISH) are risk factors for hematoma formation [10,13]. It is known that meticulous hemostasis is a crucial procedure for preventing hematoma formation [5,10,14]. Compared with general cervical spondylopathy surgery, however, OPLL and DISH surgeries are relatively involved and require higher levels of surgical skills [15,16,17]. Hence, we did not include this type of surgery in our study.
Nine patients were found to have speci c hemorrhage sites during the evacuation procedure. Several studies have reported on the site of hemorrhage, but most have not reported distinct bleeding points during the wound exploration. Nambu concluded that the hemorrhage site during postoperative neck surgery mostly occurred on the muscle surface, and excessive muscle subtraction and too much blunt dissection may be the causes of bleeding [18]. The feeding vessels of the vertebral body are often found on the ventral side of the longus colli. Thus, the stretcher or retractor should be placed on the super cial surface of the longus colli. Kunkel found 1 case where the hematoma bleeding site was platysma [12].
Therefore, more attention should be paid to the hemostasis of the longus colli, platysma, and drainage stab incision. Bipolar electrocautery is indispensable and can achieve su cient hemostasis.
Two patients developed hematoma after drainage removal, and the bleeding site was found at drainage stab point. At present, an increasing number of reports have tended to indicate that the use of drainage cannot prevent the occurrence of hematoma. In contrast, the use of drainage may increase the risk of hematoma formation during removal [10,19,20,21]. These views are consistent with our results, and it is still controversial whether the use of drainage is necessary.
Our study also found that 6 cases of hematoma occurred within 12 hours, and 6 cases of hematoma occurred within 24-48 hours. Monika indicated that hematomas usually occurred within 6-12 hours after surgery [22]. Song mentioned that 67% of hematomas occurred within 24 hours, and the remaining cases occurred within 72 hours [23]. Understanding the timing of hematoma occurrence is valuable to improve the hematoma prevention and management.
Our ndings may be somewhat limited by the lack of radiographic image of hematoma and small sample size to validate the Neck Hematoma Score. The application of heparin dose and preoperative blood pressure should be further quanti ed. The inclusion and exclusion criteria were strictly controlled to exclude confounding factors, but the occurrence of hematoma was still affected by some subjective factors. Therefore, we tried to obtain objective, independent risk factors through this study, and a large sample and multicenter study is still needed in the future.

Conclusion
ACSS is prevalent for treating cervical spine diseases, and postoperative wound hematoma is a severe complication, with a low incidence but with fatal consequences.

Consent for publication
Written informed consent for publication was obtained from all participants.

Availability of data and materials
The data is available from the corresponding author on reasonable request.

Competing interests
There are no known con icts of interest associated with this publication.

Funding
This study did not receive funding support.
Authors' contributions KZ and ZH collected the data. KZ analyzed the data and wrote the manuscript. ZQ and KT contributed to the study design. ZQ, KT, ZH and KZ performed the surgery. All authors read and approved the nal manuscript.