Incidence and Risk Factors of postoperative Dysphagia following Cervical Spine Surgery

Background. Dysphagia is a known complication of anterior cervical spine surgery and may be prolonged or occasionally serious. However, the association between posterior cervical spine surgery and the development of dysphagia is rare discussed. Furthermore, the risk factors of postoperative dysphagia are not well understood. Methods. Three hundred and fty four patients were reviewed in this study, including 172 patients undergoing AC procedure, and 182 patients with PC procedure between June 2007 and May 2012. All patients were given the same postoperative management. The presence and severity of dysphagia were reported during the telephone interviews performed at 1, 3, 6, and 12 months after the procedure. The incidence and risk factors of dysphagia after cervical spine surgery were studied. Results. Of all 354 patients, 12.8% AC and 9.4% PC patients reported dysphagia at early postoperative assessment. The overall incidence for dysphagia at 1, 3, 6 and 12 months were decreased to 9.3%, 7.0%, 4.6% and 3.5% in AC group, and 6.0%, 4.4%, 2.2% and 1.6% in PC group, respectively. Long term use of Neck Collar after surgery may lead to postoperative dysphagia. Other risk factors such as age, gender, BMI, operative time, blood loss, procedure type, revision surgery and operative levels did not signi ﬁ cantly inuence the incidence of postoperative dysphagia. incidence dysphagia both decrease time. surgery related four after

related of postoperative dysphagia. The neck collar should be worn no longer than four weeks after surgery. Age, gender, BMI, operative time, blood loss, procedure type, revision surgery and operative levels are not risk factors of postoperative dysphagia.

Background
Anterior cervical decompression and discectomy, plate and screw instrumentation, bone graft and fusion, (referred to as ACDF, Anterior cervical discectomy and fusion) was the most common procedure for cervical spondylosis. This proved to be effective and safe in clinical practice. On the other hand, patients undergoing ACDF surgery presented with postoperative dysphagia, which was one of the most frequent complications according to the literature. The occurrence of postoperative dysphagia in ACDF group varied differently from 4%-71% 1-25 , but there were rare cases of severe dysphagia. Meanwhile, a portion of patients undergoing posterior cervical spinous process-splitting laminoplasty (PC procedure) also develop dysphagia after surgery, which has rarely been mentioned in previous research.
Domestic studies concerning the mechanisms of postoperative dysphagia development in short-and long-term outcomes after cervical spine surgery were relatively few. Many potential risk factors have been reported, including age 1,2 , gender 3,4 , intraoperative esophagus traction 5,6 , long segmental surgery 7,8 , postoperative persistent pain 9,10 , prolonged procedure time 11 , high surgical position 12 , revision surgery 13 , esophageal intubation 14 , and massive blood loss 15 . In addition, other risk factors were associated with dysphagia development such as thickness of internal plate 16 , intraoperative injury of esophagus 17 , postoperative cervical hematoma formation 18 , the use of BMP-2 19 , postoperative soft swelling 20 , scar formation and tissue adhesion 21 , internal compression of the esophagus 22 and recurrent laryngeal nerve injury [23][24][25] .
Meanwhile, to access whether other confounding factors affected the occurrence of dysphagia, we collected clinical and operative characteristics. Logistic regression was used in analyzing these clinical data, including gender, age, BMI, operation time, blood loss, usage time of neck collar, type of surgical procedure (anterior/posterior), revision surgery, number of surgical segments, the superior segment of surgical cite, whether the third cervical vertebrae was included. The exclusion criteria were listed as following: i) Patients could not make routine follow up; ii) Preoperative and Postoperative JOA scores, general information data, cervical X-rays were missing; iii)

Patient Selection
Patients presented with dysphagia before surgery, including 3 of 198 patients in the ACDF group and 4 of 152 patients in the CDR group.
Surgical Protocols AC procedure.

Page 4/14
The AC procedure was performed via the standard Smith-Robinson approach to the cervical spine under general anesthesia and endotracheal intubation. A right-sided approach was typically performed if the surgical elds were above the C6 level; under C6 level we would like to choose the left-sided approach to decrease the potential injury to the right recurrent laryngeal nerve. C-arm uoroscopy was used to mark an appropriate skin incision. Blunt dissection was used to expose the anterior aspect of the cervical spine. Deep retraction of the soft tissue was achieved with smooth-edged self-retaining retractors. Direct decompression is performed by removing the offending discs and osteophytes on the preoperative radiographs. Bone fusion was obtained using either autologous iliac bone graft or coral bone allograft.
Anterior cervical plating with locking screws was used in all cases involving instrumentation ( Fig. 2.a). All patients were asked to use a rm cervical collar for at least 2 weeks postoperatively.
PC procedure.
The patients of PC group were treated with spinous process-splitting laminoplasty, as rst described by Kurokawa in 1982 and improved by us during recent years. Like AC procedure, patients were under general anaesthesia and endotracheal intubation. A midline linear incision was performed, and soft tissues were divided to expose the posterior elements. The dorsal part of each spinous process was removed. Lateral gutters for the hinge were created using a diamond burr. The lamina was then cut at the midline by using a threadwire-saw instead of diamond burr, which markedly shortened the mean time of surgery and lessened the mean blood loss. Coral bone grafts were shaped to t the spaces and secured with wire. As a general rule, C4-C6 laminoplasty was performed to provide extensive decompression while conserving the C3 and C7 spinous process ( Fig. 2.b), which was different from Kurokawa's method. As an exception, C4-C7 laminoplasty was performed for C6/7 disc herniation or stenosis. At least Two weeks of neck immobilization with a collar after surgery were recommended to these patients.

Data Collection
The presence and severity of dysphagia were recorded at early postoperative, 1, 3, 6 and 12 months after the procedure, by face-to-face questioning or telephone interview. The dysphagia grading system de ned by Bazaz et al 1 was used (Table 1). Statistical Analysis: Comparison of the overall incidence of dysphagia between the AC and PC group early preoperatively and at each time period after surgery (1, 3, 6, and 12 months) was performed according to the results of telephone interview. A Logistic regression analysis was also used to investigate the association of dysphasia with several documented risk factors, with dysphasia as the dependent variable. The independent variables included age, gender, BMI, blood loss, length of surgery, Usage time of neck collar after operation, primary or revision, AC or PC procedure (procedure type), and operative levels. All reported p values are two-tailed, and the level of statistical signi cance was p < 0.05. All statistical analysis was performed using SPSS, version 19.0(SPSS Inc., Chicago, IL, USA).

Results:
For AC group, the prevalence and severity of dysphagia at early postoperative was 12.8% ( Fig. 1)   Logistic regression model showed that usage time of neck collar after operation may lead to postoperative dysphagia. Age, gender, BMI, procedure type (AC versus PC), number of operated levels, operative time, revision surgery, and blood loss were not associated with an increased likelihood of dysphagia after AC and PC cervical procedure (Table 3).

Discussion
The occurrence of postoperative dysphagia after cervical spinal surgery in several studies ranged from from 4%-71% 1-25 . In recent years, similar studies have been reported in China. These differences in the occurrence of dysphagia may be related to inclusion criteria, the design scheme and statistical method. In this study, the occurrence of dysphagia was 12.8% of patients in ACDF group and 9.4% in PC group. Dysphagia most commonly begins in the immediate postoperative period, but may also be delayed, beginning more than 4 weeks after surgery in some patients. Most cases are mild and resolve within 1 week. In a small subset of patients, these symptoms are permanent. Despite this, dysphagia has no signi cant impact on patient satisfaction with surgery.
A total of 354 patients were followed up by researcher in this study, including 172 patients of AC group and 182 patients of PC group. The results showed that dysphagia was 12.8% and 9.4% after AC and PC procedure in the early postoperative period. Symptoms then decreased to 9.3% and 6.0% at 1 month, 7.0% and 4.4% at 3 month, 4.6% and 2.2% at 6 months, 3.5% and 1.6% at 12 months separately in the AC and PC group. No signi cant difference was found between AC and PC group in the incidence and severity of dysphagia early postoperatively or at any follow-up time after surgery. Since the incidence of postoperative dysphagia was nearly the same between these two groups, we hold the belief that the procedure type was not the key factor in the mechanism of developing dysphagia. Most of the improvement occurred in the rst 3 months postoperatively. Most of the dysphagia was graded as ''mild'' and ''moderate''. Only 1.1% of patients at 1 year after cervical surgery had dysphagia that was graded as severe (Table 2).
Swallowing requires precise coordination of neuromuscular events in different organs, including oral, pharynx, laryngeal and esophagus 10 . The process of swallowing is divided into four parts: the rst two named as oral pre-phase and oral phase, are voluntary movements involuntary dominated by human awareness; the last two named as throat phase and esophageal phase are involuntary movements. In cervical spinal surgery, Total or partial neuromuscular dysfunction during any stage of these parts may lead to the symptom of dysphagia after operation.
The To assess whether reported risk factors affected the occurrence of dysphagia, we collected clinical and operative characteristics. Logistic regression was performed in analyzing these clinical data, including gender, age, BMI, operation time, blood loss, type of surgical procedure (anterior/posterior), revision surgery, number of surgical segments, the superior segment of surgical cite, usage time of neck collar, whether the third cervical vertebrae was included. No direct association was observed between these data and postoperative dysphagia. But we nd that long term use of neck collar may lead to postoperative dyaphagia (Table 3). That may partly because neck xation for so long time is not conductive to eating and drinking. So we suggest that neck collar should be worn no longer than four weeks after cervical spine surgery.
The assessments of dysphagia in this retrospective study were based on the subjective experience by patients. Most patients presented with dysphagia after operation lack anatomical examinations such as barium meal test, and laryngoscopy 18 . In the future, prospective studies with larger samples are required to achieve more reliable conclusions. They can further provide guidance for surgical approach and lower the rate of postoperative complications.

Conclusions:
Dysphagia is a known complication after cervical spine surgery. The overall incidence of dysphagia was 12.8% after AC and 9.4% after PC procedures in our study, and the incidence and severity both decreased over time. Long term use of neck collar after surgery is related of postoperative dysphagia.  Comparison of the overall incidence of dysphagia between the AC and PC group early preoperatively and at each time period after surgery (early postoperative, 1 month, 3 months, 6 months, and 12 months).

Figure 2
X-ray lateral cervical photograph of patients following ACDF/posterior cervical spinous process-splitting laminoplasty. F 2-a. photograph of patient following ACDF F 2-b. photograph of patients following cervical spinous process-splitting laminoplasty.