Early dysphagia following anterior cervical discectomy and fusion: a centre experience

Abstract

Objective

Anterior cervical discectomy and fusion (ACDF) is a commonly performed surgical procedure in patients with cervical spine radiculopathy and/or myelopathy. It’s considered safe, but one of its most common complications is postoperative dysphagia, which can negatively impact the patient’s quality of life, and increase health costs. In this study, we discuss our experience with post-ACDF dysphagia (PAD) in terms of incidence, risk factors and approach to management.

Results

In total, 196 eligible patients were included with a mean age of 50.38 years (SD = 11.18); 107 patients (54.6%) were males, and 89 (45.4%) were females. The incidence of PAD was 5.6% (11/196 patients). No significant association could be found between the development of PAD and the assessed patient-related factors (age, gender, body mass index, and comorbidities) nor surgical factors (number of operated levels, use of anterior plate, operative time).

Introduction

Anterior cervical discectomy and fusion (ACDF) is a commonly performed surgical procedure in patients with cervical spine radiculopathy and/or myelopathy. It’s considered safe, and serious complications, such as oesophageal or vascular injuries, are rare. However, post-ACDF dysphagia (PAD) is common complication, and can negatively impact the patient’s quality of life and increase health costs [1].

The topic of PAD has been assessed in a lot of studies in the past. However, dysphagia has been variably defined, or not defined at all. This resulted in an extremely wide range of outcomes with the reported incidence having been as low as 1%, and as high as and 79% [2, 3]. In this study, we present our experience with PAD in terms incidence, risk factors and approach to management.

Methods

Study design, eligibility and data collection

This is a cross-sectional study based on the electronic health records (EHR) of patients who underwent ACDF at Dr. Sulaiman Al-Habib Hospital (Qassim Hospital, Saudi Arabia) in the period between January of 2017 and December of 2023. All patients who underwent ACDF in the specified period for cervical radiculopathy and/or myelopathy secondary to disc disease and/or spondylosis were included. Patients were excluded if the indication of their surgery was traumatic or they suffered from dysphagia preoperatively.

Eligible patients were identified and several patient-related and perioperative variables were collected. Patients-related variables: age, gender, body mass index (BMI) and comorbidities (DM and hypothyroidism); perioperative variables: number of operated levels, anterior plate placement and operative time.

The surgical procedure, postoperative protocol and definition of dysphagia

All included patients undergone cervical discectomy and fusion through a right-sided anterior approach (Smith-Robinson approach). Following soft tissue dissection, and once the anterior surface of the cervical vertebrae is reached, the surgical field patency is maintained with Casper retractor system. Discectomy is performed under a surgical microscope, and PEEK cage with a biologic bone graft (i-factor) is utilized. For some patients, depending on the surgeon’s choice, an anterior plate is placed. Following final surgical site assessment and haemostasis, a surgical drain is placed and the wound is closed anatomically. Postoperatively, patients are allowed to start soft diet on the same evening and intravenous fluids are discontinued. The following morning, the diet is advanced as tolerating and medications are changed into oral pills. The surgical drain is removed 24–48 h post-operatively, and the patient is discharged on the third postoperative day.

In this study, we define dysphagia as any difficulty of swallowing that leads to one or more of the following: (1) inability to discontinue intravenous fluids by 12 h postoperatively, (2) inability to tolerate oral medications on the first day postoperatively, (3) extension of hospital stay, or (4) readmission following discharge from the hospital.

Data analysis

Data of the eligible patients was documented in a Microsoft Excel file where the necessary cleaning was done, then imported into SPSS software version 23 for windows. The date was summarized using frequencies/percentages for categorical variables, and means/SD for continuous variables. Chi-square and Fisher exact tests were used for significance testing of the assessed categorical variables; while independent sample-t test and Mann-Witney test were used for the continuous variables with the significance level set at 5%.

Results

The patients’ characteristics, perioperative factors and their analysis with the development of PAD are presented in Tables 1 and 2. In total, 196 eligible patients were included with a mean age of 50.38 years (SD = 11.18). 107 patients (54.59%) were males and 89 (45.41%) were females. The most common comorbidities were (DM; n = 52, 26.53%), and hypothyroidism (n = 11, 5.61%). Most patients underwent a single or two-level ACDF; 81 patients (41.33%) for each.

Table 1 Comparison of patient-related factors between patients with and without post-ACDF dysphagia

Table 2 Comparison of perioperative factors between patients with and without post-ACDF dysphagia

The incidence of PAD was 5.6% (11/196 patients). In all cases of PAD, the onset was in the first or second day postoperatively, except in one case, which was on the 5th day (following discharge), and secondary to an oesophageal injury.

No significant association could be found between the development of PAD and the assessed patient-related factors (age, gender, BMI, DM and hypothyroidism), or surgical factors (number of operated levels, use of anterior plate and operative time).

Discussion

ACDF is a commonly performed surgical procedure, and PAD, depending on how it’s defined, is the most common complication associated with it [1]. In fact, some degree of dysphagia is probably inevitable. Because of this, many studies have evaluated PAD.

PAD has been variably defined in different studies, and in many, not defined at all. This resulted in an extremely wide range of results with the reported incidence ranging between 1% and 79% [2]. When defined, the most common used definition is based on the Bazaz scale, which is the first tool devised to diagnose dysphagia following ACDF [4]. It diagnoses patients with mild, moderate or severe dysphagia based on the type of food and frequency of dysphagia. In this scale, mild dysphagia is present when patients experience it only to solids and only rarely. However, the significance of diagnosing this mild form of dysphagia is questionable. In addition, several drawbacks of this scale have been pointed out, such as being oversimplified, which may lead to a lack of discrepancy between patients; and not self-administered, which might introduce bias by the therapist administering it [5].

In our study, we defined PAD as a difficulty in swallowing that led to any of the following:

(1) inability to discontinue intravenous fluids by 12 h postoperatively, (2) inability to tolerate oral medications on the first day postoperatively, (3) extension of hospital stay or (4) readmission following discharge from the hospital. Although this definition obviously overlooks mild forms of dysphagia, it offers several advantages. First, it provides a figure that could be more relevant finically. Also, it minimizes information bias that could be introduced by our study design being based on EHR, and interpretation bias that could be present in studies depending on clinician-administered tools, such as the Bazaz scale [6, 7].

The risk factors of PAD have always been a matter of debate. Factors, such as gender, number of operated levels, BMI and others have been considered. Many studies have assessed them, but reported different and conflicting results [2, 3]. Our study has assessed several patient-related (Table 1) and perioperative (Table 2) factors, but failed to show any significant association with PAD. Nevertheless, we should point out again that we defined dysphagia differently, and assessed only early postoperative dysphagia.

Several studies have evaluated and proposed strategies to prevent and/or improve PAD. This includes preoperative steroid use and tracheal/oesophageal traction exercises [8, 9]; intraoperative retropharyngeal steroid use and decreased endotracheal cuff pressure during neck retraction [10, 11]; postoperative steroid use [12]; and perioperative laryngeal rehabilitation therapy [13]. However, despite how common ACDF is, there are no agreed upon standards on how to approach PAD.

At our institution, some patients are prescribed a short course of postoperative steroids either prophylactically or therapeutically, and the decision is based on the surgeon’s judgment and experience. When a patient does complain from dysphagia, in most cases, they are reassured and managed expectantly. Imaging studies, such as lateral neck radiograph and gastrografin swallow, are performed in one of two situations. First, if the patient develops additional symptoms (fever, surgical site swelling or discharge). In our experience, a patient developed PAD on the fifth day with associated fever, and imaging studies showed that he his symptoms were secondary to oesophageal injury [14]. The second situation is when the dysphagia is worse than expected in severity and/or duration; a subjective judgment that is also based on the surgeon’s experience, which reflects the lack of standardized guidelines on how to approach PAD.

Conclusions

PAD is a common and well-known phenomenon. The incidence of PAD that could result could prevent adequate oral intake, lead to extension of hospital stay or re-admission is 5.6%. The occurrence of PAD is largely unpredictable. Future research needs to focus on how to approach PAD. This includes how to improve PAD symptoms, when to be managed expectantly, and when/what further investigative studies need to be performed.

Limitations

This study has three limitations. First, its design being retrospective and dependent on the patients’ EHR. Second, it does not include milder forms of dysphagia. Finally, the sample size is small compared to other similar studies.