Incidence and risk factor for short term postoperative cough after thyroidectomy

There has been a substantial increase in the proportion of thyroid cancer cases globally, on the one hand, because the prevalence has really increased, on the other hand, because of the prevalence of thyroid color Doppler ultrasound examination [1]. Usually the disease is asymptomatic, surgery is the first choice of treatment, and thyroidectomy is considered as an effective and safe option for most patients who have low chance of suffering permanent vocal distortion, swallowing difficulties, and hypocalcemia when performed by experienced surgeons [2‐4]; however, some postoperative symptoms such as nausea and vomiting, local sensory disturbance and transient throat pain are still inevitable after the surgery. In our cancer center, some patients who have done thyroidectomy may also suffer serious cough; postoperative cough can even induce postoperative hematoma [2]. Patients complain that this phenomenon is quite worrisome, and doctor-patient conflicts may even occur [5], especially in patients without a previous history of cough. This suggests the importance of preoperative communication for postoperative cough. Factors including smoking history, surgical extent, and operation time might be associated with postoperative cough. However, there are few studies available in terms of its prevalence and potential predictors.

Since its first introduction by Birring et al. [6], the Leicester Cough Questionnaire (LCQ) has been regarded as a reliable tool for evaluating the cough in adults by a number of researchers [7‐9] . Therefore, our goal was to prospectively analyze the postoperative cough in patients who undergoing thyroidectomy.

Ethics approval and consent to participate: Henan Cancer Hospital Research Ethics committee (approval number: HNZZ20170102) approved this study, written informed consent was obtained from all patients at initial treatment.

From January 2018 to December 2018, adult (≥18 years) patients undergoing primary thyroidectomy were prospectively tracked. The exclusion criteria were as follows: the patient had chronic cough associated with smoking or gastroesophageal reflux or with other causes; resection of the trachea or larynx was performed; the recurrent laryngeal nerve was invaded by the tumor or metastatic nodes resulting in recurrent laryngeal nerve paralysis; and there was pulmonary infection. The symptom of cough had to begin on the first day after the operation, and was defined and assessed by our research group based on previous studies [10]. Patients who had a history of smoking/drinking at the time of diagnosis or had quit smoking/drinking for less than 1 year were defined as smokers/drinkers [11]. The operation time was defined as “the time from the beginning of endotracheal intubation to the point of extubation” [12]. Data regarding age, sex, BMI, operation time, postoperative pathology, operation type, and drinking and smoking status were collected and analyzed. Based on the Chinese Nutrition Society, overweight refers to a BMI from 24 to 28, and obesity refers to a BMI above 28.

All patients received an open surgery under general anesthesia using both intravenous anesthesia and inhalation anesthesia, the frequently used narcotic drugs included cisatracurium, propoxate and fentanyl. After the operation, patients needed to stay for a short time in the recovery room and then they would be transferred to the ward. The extent of operation of the primary tumor consisted of two types: unilateral thyroid operation referred to surgery involving only one thyroid lobe, and bilateral thyroid operation referred to surgery involving both thyroid lobes. In our hospital, central neck dissection was routinely performed for thyroid papillary and medullary carcinoma. Lateral neck dissection was performed if there were positive nodes at level 3 or 4 according to frozen sections. All patients had atomised inhalation after surgical treatment.

All patients enrolled needed to complete the Mandarin Chinese Version of LCQ [13] preoperatively in the ward, and patients with postoperative cough were required to complete the LCQ at 2 weeks and 4 weeks postoperatively via the out-patient department, email, or WeChat. The LCQ was usually used for measuring chronic cough, but recent evidence showed there was also high validity and responsiveness in assessing acute cough or postoperative cough [7, 14, 15]. The LCQ is easy to complete taking less than 5 min by themselves. There are 19 items in total, each item represents an adverse event caused by cough. The responses were scored by a 7-point Likert scale. The 19 items were divided into three areas that considered the psychological effects (for instance the impact of cough on embarrassment/anxiety), physical effects (for instance the impact of cough on chest and stomach pain), and social effects (for instance the effect of cough on work/daily life and entertainment life). A total score and three domain scores were calculated, the score in each domain is between 1 and 7, and the total score is between 3 and 21; the higher the score, the better the health [16].

Recurrent laryngeal nerve injury and parathyroid injury are common complications after thyroidectomy and common causes of doctor-patient conflict [17, 18]. However, with the increasing demand for high-quality medical treatment, severe postoperative cough has become an aspect of concern in recent years. Our findings have shown that postoperative cough was relatively common after thyroid surgery, with an overall prevalence rate of 28.5%, and it was more common in patients with malignant disease than in patients with benign disease. In a study by Jung et al. [19], the author’s goal was to assess the effect of a humidifier with heated wire circuits on the prevalence and severity of cough after thyroid surgery, and they found that in patients undergoing active humidification of inspired gases, the prevalence of postoperative cough was significantly decreased compared to patients without a heated humidifier. In the current study, all patients had routine postoperative aerosol inhalation of budesonide aerosol and ambroxol, twice a day for 5 days. The overall prevalence of postoperative cough was in agreement with the findings of the above study.

The prevalence of postoperative cough in patients with malignant diseases was higher than that in patients with benign diseases. This finding was interesting. Most previous authors have analyzed cough in patients with a flexible reinforced laryngeal mask airway or a plain endotracheal tube [20‐22]; although they described that compared with conventional endotracheal intubation for the placement of flexible reinforced laryngeal mask airway during operation can reduce the incidence and severity of laryngo-pharyngeal symptoms, no authors have evaluated whether the type of pathology affects postoperative cough. The variation between the two groups might be explained by the following: surgical trauma, anesthesia, tracheal intubation, and recurrent laryngeal nerve dissection are all potential causes for cough following surgical procedures [23], and patients with malignant disease had longer operation times and more instances of routine central neck dissection in the current study.

Risk factors for postoperative cough following surgical procedures have been occasionally analyzed. Lin et al. [12] demonstrated that a long duration of anesthesia time, female sex, subcarinal node resection and lower paratracheal node resection were independent risk factors for postoperative cough in non-small cell lung cancer patients. The correlation between lung surgery and postoperative cough was reported by Chen et al. [24]. They found that the more aggressive patients were, the higher the probability of postoperative cough. This study was the first to analyze the predictors of postoperative cough after thyroidectomy. Similar to previous reports, we also noted that operation time was an independent predictor for postoperative cough. Moreover, smoking has been proven to be related to chronic cough by Colak et al. [25], and there was also a positive linkage with acute cough based on our outcome.

Another interesting finding was that postoperative cough was more common in patients with more than 3 positive central nodes. In our view, the most likely cause of this finding is associated with the branches of the recurrent laryngeal nerve. Small branches of the nerve, such as the tracheal branch, are often encountered during thyroidectomy, and they may inadvertently be resected during central neck lymph node dissection. However, the actual frequency of branch excision has not been clearly documented. More research is needed to clarify this issue. Another potential explanation is the tracheal thermal damage associated with the usage of high-frequency electric surgical knives and ultrasonic scalpels during operation.

It is important to evaluate the impact of postoperative cough on quality of life. The LCQ is a reliable method that has been used as an outcome measure in many clinical trials [15, 23, 26]. Lin et al. [12] described that in patients receiving video-assisted thoracoscopic surgery for lung cancer, the mean postoperative total score was 16.35, which was significantly lower than the mean follow-up score after 1 month, but the authors did not provide the data of preoperative levels. In our previous study, we found that thyroidectomy was significantly associated with a decreased LCQ score compared to baseline scores, but the study did not report when the LCQ scores returned to preoperative levels. In the current study, we were the first to note that for patients undergoing thyroidectomy regardless of the presence of benign or malignant disease, the mean postoperative LCQ score returns to baseline level in 4 weeks. The time interval found in this study was significantly shorter than in patients undergoing lung surgery, which can be attributed to the different types of operation.

Moreover, we found that the 2-week mean LCQ score was quite lower in patients with malignant disease than in those with benign disease. One possible reason for this difference is that patients with malignant disease were associated with a longer operation time and more surgical trauma, including routine dissection of the recurrent laryngeal nerve.

We must admit that the study can be limited. Firstly, cough assessment usually consists of objective and subjective measures. Although the LCQ is a reliable method for subjective assessment, more objective analyses are needed to clarify postoperative cough after thyroidectomy. Secondly, related intubation laryngitis, or laryngeal trauma can also cause coughing after surgery. However, for patients with cough, endoscopy is not used for routine laryngeal examination. Any misclassification would deviate our analysis. Last but not the least, the BMI and lifestyle characteristics were significantly different in this group to a western group, it remained unclear whether this finding could be confirmed in western studies.

In summary, compared to patients with benign disease, patients undergoing thyroid cancer surgery had a higher prevalence of postoperative cough and a lower 2-week postoperative LCQ score. However, the 4-week postoperative LCQ score returned to the preoperative level in patients undergoing thyroidectomy regardless of the presence of benign or malignant disease. The factors of smoking and operation time were the most important predictors for postoperative cough after thyroidectomy.