|Home | About | Journals | Submit | Contact Us | Français|
Correspondence to: Dr. Daw Shyong Perng, Department of Gastroenterology, E-DA Hospital, I-Shou University, No. 1, Yida Road, Jiaosu Village, Yanchao District, Kaohsiung 82445, Taiwan, China. machinozomu/at/mail2000.com.tw
Telephone: +886-7-6150011 Fax: +886-7-6150011
AIM: To investigate whether the incidence of hiccups in patients undergoing esophagogastroduodenoscopy (EGD) or same-day bidirectional endoscopy (EGD and colonoscopy; BDE) with sedation is different from those without sedation in terms of quantity, duration and typical onset time.
METHODS: Consecutive patients scheduled for elective EGD or same-day BDE at the gastrointestinal endoscopy unit or the health examination center were allocated to two groups: EGD without sedation (Group A) and BDE with sedation (Group B). The use of sedation was based on the patients’ request. Anesthesiologists participated in this study by administrating sedative drugs as usual. A single experienced gastroenterologist performed both the EGD and the colonoscopic examinations for all the patients. The incidence, duration and onset time of hiccups were measured in both groups. In addition, the association between clinical variables and hiccups were analyzed.
RESULTS: A total of 435 patients were enrolled in the study. The incidences of hiccups in the patients with and without sedation were significantly different (20.5% and 5.1%, respectively). The use of sedation for patients undergoing endoscopy was still significantly associated with an increased risk of hiccups (adjusted odds ratio: 8.79, P < 0.001) after adjustment. The incidence of hiccups in males under sedation was high (67.4%). The sedated patients who received 2 mg midazolam developed hiccups more frequently compared to those receiving 1 mg midazolam (P = 0.0028). The patients with the diagnosis of gastroesophageal reflux disease (GERD) were prone to develop hiccups (P = 0.018).
CONCLUSION: Male patients undergoing EGD or BDE with sedation are significantly more likely to suffer from hiccups compared to those without sedation. Midazolam was significantly associated with an increased risk of hiccups. Furthermore, patients with GERD are prone to develop hiccups.
Esophagogastroduodenoscopy (EGD) and colonoscopy are important diagnostic tools for evaluating gastrointestinal diseases or for the early detection of malignant lesions. Although same-day bidirectional endoscopy (EGD and colonoscopy; BDE) has many advantages, including shorter hospital stays and expedited decision making for patient care, BDE remains a difficult and longer endoscopic procedure. Sedation is usually used to alleviate patient discomfort. Several regimens are available for inducting moderate sedation, but the combination of propofol with an opioid[2-4] and/or a benzodiazepine[5,6] is most commonly used. The combined regimens could improve the safety profile of propofol by reducing its dose-dependent potential to induce deep sedation, in which patients may develop inadequate spontaneous ventilation and may require assistance to maintain a patent airway. In addition to hemodynamic or respiratory depression, the other frequently encountered clinical situation during endoscopic procedures with sedation is hiccups.
Hiccups are an involuntary spasmodic contraction of the diaphragm and accessory muscles followed by the sudden closure of the glottis. Hiccups during an endoscopic procedure are commonly acute and temporary, but the intermittent involuntary contraction of the diaphragm may interfere with spontaneous breathing and hamper endoscopic investigations and interventions. In addition, this powerful reflex can result in markedly negatively intrathoracic pressures and increased abdominal pressures. Accompanied by gastric distension due to the air inflation that is required during EGD, hiccups can induce reflux of gastric contents into the esophagus and can cause regurgitation and aspiration, particularly in sedated patients[10,11].
Although their exact mechanism remains unknown, hiccups may be induced by stimulating the postulated hiccup reflex arc, which is a complex interaction mainly composed of the phrenic and vagus nerves, the sympathetic chain and accessory nerves connecting to the glottis and the inspiratory intercostal muscles[12-14]. Vagus and phrenic nerve irritation, central nervous system disorders, toxic-metabolic disorders, and psychogenic factors are 4 possible mechanisms leading to hiccups[12,14,15]. As described above, gastric distention during endoscopy stimulates the phrenic nerve and causes hiccups. Drug-induced hiccups are uncommon, but anesthetics and sedative agents, such as benzodiazepines, opioids, and intravenous general anesthesia drugs, are suspected agents in cases of drug-induced hiccups. On the other hand, the sedation of patients to relieve anxiety and discomfort appears to decrease the incidence of hiccups during endoscopic procedures. Therefore, whether sedation can protect the patients undergoing endoscopic procedures against the onset of hiccups is unknown and requires further investigation.
Gastroesophageal reflux disease (GERD) is a common outpatient gastroenterology problem[16,17]. Patients with GERD may present a variety of symptoms, ranging from typical symptoms, such as heartburn or regurgitation, to atypical symptoms, such as chest pain, asthma, laryngitis, or chronic cough. Hiccups represent an atypical symptom of GERD, with a prevalence of 4.5% to 9.5% in GERD patients[18,19]. The major pathophysiology of GERD includes increased episodes of transient lower esophageal sphincter (LES) relaxations, ineffective esophageal motility, and reduced LES tone[20-22]. Interestingly, hiccups could also induce the transient decrease of LES tone[8,20] and even the loss of LES tone. Therefore, the increased probability of hiccups in GERD patients during sedation requires further investigation.
The goals of this study were to determine the incidence of hiccups in patients undergoing EGD or BDE with or without sedation, the possible causes of hiccups in sedated patients and the association of hiccups with GERD in sedated patients.
This prospective study was approved by the Institutional Review Board of the E-DA hospital (EMRP-098-125). Consecutive adult (18 years or older) patients scheduled for outpatient EGD or same-day BDE at the gastrointestinal endoscopy unit or the health examination center were included. The exclusion criteria included history of esophageal, gastric or colorectal cancer; history of esophageal or gastric surgery; esophageal or gastric varices; allergy to propofol or eggs; and American Society of Anesthesiology risk Class 3 or higher.
The participating patients were allocated into 2 groups: EGD without sedation (Group A) and BDE with sedation (Group B). Sedation used for endoscopic examination was based on the patient’s explicit request. The sample size was calculated from the results of our pilot study of patients who underwent EGD without sedation. We assumed that the patients with sedation would show a different (approximately 15%) prevalence of hiccups than the nonsedated patients. Therefore, 195 patients per group were required [where α = 0.05 (two-tailed), β = 0.2]. At least 400 patients were studied to allow for any loss of follow-up.
After consent was obtained, demographic, medical and drug history data were recorded for each patient. On arrival in the endoscopy room, each patient was routinely given supplemental oxygen (4 L/min) thorough a nasal cannula. In addition, an intravenous route was set up in patients of Group B, and no intravenous fluids were given before the initiation of sedation. Routine patient monitoring included electrocardiography, noninvasive arterial blood pressure measurement, and pulse oximetry.
In all the participating patients (Group A and Group B), 20 mg hyoscine N-butyl bromide (Buscopan) was administrated intravenously as premedication except in those with glaucoma, obstructive uropathy, coronary heart diseases, or a history of allergies to anticholinergic drugs. Anesthesiologists participated in this study by administrating sedative drugs. Fifty microgram of fentanyl plus midazolam were given, followed by the administration of propofol approximately 1 to 2 min later. However, the choice of whether to use midazolam as an adjuvant and the administration dosage were determined by the anesthesiologists without further input from the researchers. To avoid any pain from injecting propofol, lidocaine was mixed with propofol before administration (1 mL of 2% lidocaine in every 20 mL of 1% propofol). An initial bolus of 1.5 mg/kg of propofol was given and then titrated in 10- to 20-mg increments to achieve an adequate level of sedation.
A single experienced gastroenterologist performed both the EGD and the colonoscopic examinations for all the patients. The EGD started when an adequate level of sedation (in which the patient was asleep, not responding verbally but ventilating spontaneously) was achieved. Once the EGD was finished, the subsequently colonoscopy was preformed immediately. During the endoscopic procedure, the need to add more propofol was estimated by the patient’s pain response (i.e., moans, grimaces, gag reflex, and movements).
Predefined complications (hypotension, bradycardia, airway obstruction, hypoventilation, and hypoxia) were managed according to the hospital’s routine protocol. At the end of the procedure, the patients were transferred to the recovery room. The time when the patient was ready for discharge according to hospital criteria (oxygen saturation > 95% on air, heart rate > 60 beats/min, systolic blood pressure ± 30 mmHg from preoperative values, orientated, pain score < 4/10, and no nausea or vomiting) was recorded.
Demographic data were recorded, including each patient’s age, gender and body weight. Medical data were collected, including diabetes, hypertension, cigarette smoking/alcohol consumption, cerebrovascular diseases, gastrointestinal disorders, GERD-like syndromes, liver diseases, psychological disorders, and history of intractable hiccups. The patients’ drug histories regarding agents suspected of causing hiccups [i.e., non-steroidal anti-inflammatory drugs (NSAIDs), benzodiazepines and aminophylline] was also recorded. Sedative regimens, midazolam and propofol dose titrated to an adequate level of sedation for EGD insertion and total propofol dose were recorded. Oxygen saturation, heart rate and arterial blood pressure were recorded every 5 min during sedation. All the times were recorded from a continuously running stopwatch. Endoscopy time was defined as the time from the insertion and until withdrawal of the colonoscope from the anus. If hiccups occurred during the procedure, the time hiccups occurred, the time hiccups subsided and the administrated doses of propofol were recorded. After the endoscopic examinations, the patients with a diagnosis of GERD and other gastrointestinal disease revealed by the EGD were collected.
All the continuous data were tested for normality. Normally distributed continuous data were expressed as the mean ± SD, and t tests were used to compare the means of continuous data. Skewed data were expressed with medians and ranges and were compared using Wilcoxon’s ranked sum test. Categorical data were analyzed using the Pearson χ2 test with Yates’ correction or Fisher’s exact probability test. When there were more than two groups, normally distributed data were compared using analysis of variance, and skewed data were compared using the Kruskal-Wallis test.
In all tests, a 2-tailed P < 0.05 was considered statistically significant. The relative magnitudes of the associations between individual (categorical and continuous) variables and the likelihood of the occurrence of hiccups was compared using crude odds ratios (ORs). The precision of the estimated ORs was assessed by examining the 95% confidence intervals (CIs). A multivariate logistic regression model containing all the candidate variables was used to examine the independent contribution made by each variable, while controlling for all the variables. This resulted in an adjusted OR and a calculated 95% CI. The statistical analyses were performed with SPSS for Mac version 19.0 (SPSS Inc., Chicago, IL).
A total of 435 patients were invited to participate in our study. Ten patients were excluded from the analysis because of incomplete data collection. Therefore, a total of 425 data sets were eligible for analysis: 215 data sets were from the non-sedation group, and 210 were from the sedation group. Table Table11 summarizes the baseline characteristics of the patients in this study. There was no significant difference between the 2 groups regarding gender, body weight, hypertension, psychological disorders, or aminophylline intake history. However, there were more patients in Group A who took NSAIDs and benzodiazepines and had histories of diabetes, cigarette smoking/alcohol consumption, cerebrovascular diseases, gastrointestinal disorders, GERD-like syndromes and liver diseases than in Group B.
Though the indications for endoscopic examination in our patients were including epigastralgia, abdominal fullness, dysphagia, change in bowel habit, positive fecal occult blood test, and bloody stool etc., the indication in most patients was for heath check-up. Hence, these significant differences in baseline characteristics were because most Group B participants undergoing endoscopic examination were for health check-up.
The incidences of hiccups in Group B (under sedation) and Group A (without sedation) were 20.5% and 5.1%, respectively. The ORs and 95% CI for the occurrence of hiccups induced by sedation were calculated. The use of sedation for patients undergoing endoscopy was significantly associated with an increased risk of hiccups (OR, 4.78; 95% CI: 2.39-9.56). This association remained significant and independent after adjustment for age, diabetes, cigarette smoking/alcohol consumption, cerebrovascular diseases, gastrointestinal disorders, GERD, liver diseases, and the use of NSAIDs or benzodiazepines (adjusted OR, 8.79; 95% CI: 3.27-23.60).
Table Table22 summarizes the characteristics of the patients who developed hiccups and those who did not during sedation. Gender was the only significant difference between these groups (P < 0.05). The majority of the patients who developed hiccups during sedation were male (67.4%).
The onset time (counted from the insertion of EGD) of hiccups was 6 min (1-21 min) [median (range)], and their duration was 1 min (1-19 min) [median (range)]. All the hiccups occurred before the end of the EGD in sedated patients, except in 3 patients who suffered from hiccups after the insertion of the colonoscope (onset time was 21, 17 and 16 min). Two patients suffered from hiccups that lasted for 19 and 15 min.
In the pharmaceutical characteristics of the development of hiccups, the effect of the midazolam dose was observed (Table (Table3).3). The patients who received 2 mg midazolam (1-2 mg) (range) developed hiccups more frequently compared with patients who received 1 mg midazolam (range: 0-2.5 mg, P = 0.028).
Diagnostic findings of endoscopy were listed in Table Table3.3. The patients with diagnosed GERD were more prone to develop hiccups (P = 0.018).
To the best of our knowledge, this is the first study to examine the prevalence of hiccups in patients undergoing EGD or BDE. In this prospective study, sedation was associated with a significantly increased occurrence of hiccups. Midazolam, which was used in our combined sedation regimens, was associated with the onset of hiccups. In addition, the patients diagnosed with GERD revealed by EGD were more prone to develop hiccups while undergoing endoscopic procedures with sedation.
Clinically, most hiccup episodes begin with an acute onset, are benign, and are self-limited, typically ceasing within minutes. However, the sudden onset of hiccups may become a safety hazard while patients are sedated. Hiccup-associated acute negative intrathoracic pressure may occur, resulting in hypotension and bradycardia. This effect is attributed to decreased vascular resistance resulting from increased dilation and volume of the thoracic aorta. In our study, no statistically significant changes occurred in any hemodynamic measure following the onset of hiccups. Whether hiccup-associated systolic hypotension is deleterious to cardiovascular function in adults is unclear, but it remains a plausible etiology of pathological hemodynamic changes in those patients with underlying heart disease.
Hiccups could influence the respiration of the patient during anesthesia. Unexpected pulmonary aspiration has been diagnosed in positron emission tomography (PET) screening following panedoscopy under conscious sedation. In our patients, we observed no respiratory distress associated with cyanosis or desaturation on pulse oximetry in patients with hiccups during the endoscopic examinations or in the recovery room. Although aspiration during sedation may be silent and uneventful, hiccups in sedated patients do carry clinical risks[10,28].
Acute gastric distention due to air inflation during endoscopy might result in hiccups by stimulating gastric vagal afferent activity. To avoid any influence of the gas inflation that is required during a colonoscopy, we conducted EGD before colonoscopy in the patients of Group B. Furthermore, to avoid inter-observer variation, a single physician performed all the EGD and colonoscopy procedures. In the sedation group, the hiccups all occurred during the EGD, and almost all of the hiccups ceased before the initiation of colonoscopy. Further, the incidence was significantly greater than in the nonsedated group (Group A). Therefore, causes other than this mechanical factor could have induced hiccups. Psychogenic factors (e.g., anxiety, stress, and excitement) are other possible causes of hiccups. Therefore, sedation for EGD or BDE relieves patient anxiety and seems to decrease the incidence of hiccups. However, Group B did not show a lesser incidence of hiccups than Group A. In fact, the use of sedative drugs during endoscopic procedures favored the onset of hiccups.
Although there was insufficient evidence in Thompson’s review to conclude that any specific drug induces drug-related hiccups, corticosteroids and benzodiazepines are the most frequently suspected agents in cases of drug-induced hiccups[15,29]. Midazolam is a benzodiazepine with rapid onset, brief duration of action and an amnestic effect, and it is commonly administered in combination with an opiate and/or propofol to achieve adequate sedation for an EGD or colonoscopy. In addition, midazolam has gained popularity for use in procedural sedation and anxiolysis in pediatric patients. Midazolam is water-soluble and can be safely administered in a variety of ways. Marhofer et al had investigated that the incidence of hiccups in pediatric patients who were pre-medicated with two different doses of rectally administrated midazolam for minor surgery. Their patients in group A received 0.5 mg/kg of midazolam, while their patients in group B were treated with 1 mg/kg. Twenty-four percent of the children developed hiccups, but no statistically significant difference was noted between the two doses (22% in group A vs 26% in group B). Hiccups were more common among younger children (i.e., those aged 5-6 mo vs 20 mo). In our sedated patients, the total incidence of hiccups was similar, but age was not a statistically significant variable. In addition, the development of hiccups was associated with higher doses of midazolam.
Interestingly, drug-induced hiccups are reported more commonly in men than women. In our study, the majority of patients who developed hiccups during sedation were male. This finding is consistent with recent studies revealing a significant male predominance of hiccups in patients receiving cytotoxic chemotherapy[32,33]. However, the mechanisms for the male predominance of hiccups and the midazolam-induced hiccups are still unknown. Hiccups might be induced by the influence of midazolam on a supraspinal hiccup center localized in the brain stem. Another factor causing midazolam-induced hiccups could be a direct stimulation of the inspiratory muscles, specifically on diaphragm contractility. By contrast, Fujii et al indicated that midazolam decreases the contractility of the diaphragm in a dose-dependent manner in an animal study. Furthermore, intravenously administered midazolam has been successfully used in patients with terminal hiccups. The discrepancies among these results perhaps reflect the complicated etiology of hiccups; further studies are required to improve our understanding of their etiology. Nevertheless, our study provides further evidence for the association of midazolam administration with the onset of hiccups.
Graham showed that there is no detectable LES tone during a hiccup attack. Hiccups add tension to the phrenoesophageal ligament and thus have the same sphincter-dilating effect. Therefore, hiccups could induce episodes of transient decrease of LES tone[8,20] that result in reflux both in normal subjects and in esophagitis patients. Furthermore, LES function can be overcome by the transient peritoneal-pleural gradient created during hiccups. In the study of Werlin et al, 27% transient increases in intra-abdominal pressure, such as would be caused by a hiccup, were associated with reflux. According to Vanner, approximately 40% of patients who hiccup after induction of anesthesia develop detectable gastroesophageal reflux.
The most important aspect of the pathophysiology of GERD is the competency of the anti-reflux barriers. Factors contributing to their integrity include the LES pressure, the presence or absence of a hiatal hernia, and the occurrence of transient LES relaxation[22,41]. Transient LES relaxation is considered the most common pathophysiologic event at the time of a reflux episode[22,41]. Hence, the fact that patients with GERD are prone to develop hiccups during sedation might be due to previous abnormalities of LES.
This study had several weaknesses. Because of the use of sedation is a self-pay service and based on the patients’ request, it is difficult to perform a randomized controlled trial to detect the incidence of hiccups in patients undergoing EGD or BDE with or without sedation. Though there is potential for bias in our prospective and cross sectional study, the study design could examine the association among the variables (e.g., sedation vs hiccups in our study). In addition, the fact that the choice of midazolam as an adjuvant was not random limits the ability to draw conclusions about its hiccup-inducing effect, but it raises interesting hypotheses and provides pilot data for testing in the future.
In conclusion, sedation was associated with the occurrence of hiccups in patients undergoing gastrointestinal endoscopic procedures, and hiccups occurred primarily in males. Midazolam, a sedative drug, was significantly associated with an increased risk of hiccups. Furthermore, patients with GERD were prone to develop hiccups.
The authors thank Hon-Yi Shi, PhD for revising the manuscript.
During esophagogastroduodenoscopy (EGD) and colonoscopy procedure, sedation is usually used to alleviate patient discomfort. In addition to hemodynamic and respiratory depression, hiccups are the other frequently encountered clinical situation during endoscopic procedures with sedation. Hiccups, though commonly acute and temporary, may interfere with spontaneous breathing and hamper endoscopic investigations and interventions. However, the exact mechanism of hiccups remains unknown.
There are 4 possible mechanisms leading to hiccups, including vagus and phrenic nerve irritation, central nervous system disorders, toxic-metabolic disorders, and psychogenic factors. Anesthetics and sedative agents are suspected agents in cases of drug-induced hiccups. On the other hand, the sedation of patients to relieve anxiety and discomfort appears to decrease the incidence of hiccups during endoscopic procedures. Therefore, whether sedation can protect the patients undergoing endoscopic procedures against the onset of hiccups is unknown and requires further investigation. In addition, hiccups represent an atypical symptom of gastroesophageal reflux disease (GERD), the increased probability of hiccups in GERD patients during sedation also requires further investigation.
To the best of our knowledge, this is the first study to examine the prevalence of hiccups in patients undergoing EGD or colonoscopy. Sedation for endoscopic procedures relieves patient anxiety and seems to decrease the incidence of hiccups. However, in this prospective study, the use of sedative drugs during endoscopic procedures favored the onset of hiccups. Although there was insufficient evidence to conclude that any specific drug induces drug-related hiccups, benzodiazepines are the most frequently suspected agents in cases of drug-induced hiccups. Midazolam, a benzodiazepine with rapid onset and brief duration of action, was associated with the development of hiccups, especially in higher doses. In addition, the study also shows that patients with GERD are prone to develop hiccups during sedation.
The study results suggest that the development of hiccups was associated with higher doses of midazolam. Therefore, avoid using midazolam or decreasing the midazolam dosage in combined sedation regimen for endoscopic procedure can decrease the incidence of onset of hiccups.
Hiccups: Hiccups are an involuntary spasmodic contraction of the diaphragm and accessory muscles followed by the sudden closure of the glottis; GERD: Patients with GERD may present a variety of symptoms, ranging from typical symptoms, such as heartburn or regurgitation, to atypical symptoms, such as chest pain, asthma, laryngitis, chronic cough, or hiccups. The major pathophysiology of GERD includes increased episodes of transient lower esophageal sphincter (LES) relaxations, ineffective esophageal motility, and reduced LES tone.
The results of the study are interesting and it was well written. Nevertheless, there are some major points of concern.
Peer reviewers: Cesare Tosetti, MD, Department of Primary Care, Health Care Agency of Bologna Via Rosselli 21, 40046 Porretta Terme (BO), Italy; Diego Garcia-Compean, MD, Professor, Department of Gastroenterology, Faculty of Medicine, University Hospital, Autonomous University of Nuevo Leon, Ave Madero y Gonzalitos, Monterrey, NL 64700, Mexico
S- Editor Cheng JX L- Editor A E- Editor Zheng XM