Abstract
Objective
To compare the complication rates of bowel perforation during colonos-copy performed with or without anesthesia.
Methods
We retrospectively analyzed 9501 case records of colonoscopy performed at the Chi Mei Medical Center between 2000 and 2004, and compared the rates of bowel perforation between patients receiving anesthesia during the procedure versus those that did not receive anesthesia. Poisson distribution was used for statistical analysis.
Results
Only one case of perforation was found in the non-anesthetized group (n = 2460) compared with two cases in the anesthetized group (n = 7041). No statis-tical difference was found (p = 0.6173; risk ratio = 0.6988; 95% confidence interval by Byar’s method = 0.063−7.705).
Conclusion
During colonoscopy, anesthesia per se does not affect the rate of bowel perforation.
Keywords
anesthesia; colonoscopy; intestinal perforation;
1. Introduction
Colorectal cancer is the fourth most common malig-nancy in England (data from national statistics: http://www.statistics.gov.uk/cci/nugget.asp?id = 915) and is the second leading cause of cancer death in the United States. In Taiwan, colorectal cancer is the second leading cause of cancer-related death, with a crude incidence of 32.9 cases/100,000 and a mortality rate of 15.5 cases/100,000 (data from Bureau of Health Promotion, Department of Health, Executive Yuan, Taiwan). Routine colorec-tal screening can help reduce the mortality rate.1,2 Colonoscopy is the most common diagnostic tool for disclosing colon and rectal diseases and is a routine procedure for patients with abdominal symptoms suspicious of large bowel diseases.3 Unfortunately, colonoscopy is uncomfortable for most patients and sharp pain is the most frequent complaint during the procedure. The advantages and disadvantages of using analgesics and sedatives during colonoscopy are widely debated. The major complications of colonoscopy are bowel perforation and post-polypectomy bleeding.1,2 The perforation incidence is relatively low: 0.03−0.65% for diagnostic colo-noscopy and 0.073−2.14% for therapeutic colonos-copy.3,4 Since 1974, colonoscopy under anesthesia has been considered a risk factor for perforation. The incidence of large bowel perforation during colonoscopic procedures under anesthesia was re-ported by Livstone et al,5 Rogers et al,6 Koyama7 and Smith.8 However, these investigators showed neither objective data nor conclusive findings. Kjaergard et al9 suggested that anesthesia performed by ex-perienced staff is harmless during colonoscopy, and that general anesthesia does not increase the risk of perforation of the large bowel. However, their re-port was based on a small number of cases and no comparison was made between perforation rates in procedures done with or without anesthesia. Three other studies (Dillon et al,10 Le Guern et al,11 and Heiman et al12) were also based on small samples.
Therefore, we performed this study to deter-mine whether the perforation rate during colonos-copy is affected by anesthesia in a large population of patients.
2. Methods
2.1. Patients
A total of 9501 cases of colonoscopy performed at the Health Check Service, Chi Mei Medical Center, between January 2000 and December 2004 were retrospectively analyzed. The cases were first sorted into those that received anesthesia group (n = 7041) or did not receive anesthesia (n = 2460). The use of anesthesia was at the patient’s voluntary discre-tion. The principles of the Declaration of Helsinki were strictly followed without any disclosure of personal information or suffering during our study.
2.2. Monitored anesthesia care and procedure of colonoscopy
Patients who requested anesthesia received a com-plete preanesthetic examination and assessment by an anesthesiologist before colonoscopy. All anesthetic procedures were carried out by board-certified anesthesiologists. Patients received standard an-esthesia care and monitoring throughout the pro-cedure and in the postanesthesia care unit. Blood pressure, heart rate, electrocardiogram and SpO2 were monitored conventionally (Datex Ohmeda S5 system; GE Healthcare, Milwaukee, WI, USA). Oxygen support was facilitated by a nasal prong. In case of low oxygenation (SpO2 < 90%), assisted ventilation with an ambu bag (FiO2 60%) was applied until SpO2 reached 97% before allowing spontaneous breath-ing. When bradycardia occurred as a consequence of colonoscopic manipulation, atropine (0.02 mg/kg) was given. Intravenous fluid supplement with ephedrine (5−10 mg intravenous bolus) was given in case of hypotension. Resuscitation drugs and equipment were available for unexpected cardio-vascular events. Patients in the anesthetized group received an initial dose of intravenous midazolam (1−2 mg) or propofol (1.0 mg/kg), followed by pro-pofol infusion (25−100 μg/kg/min). Alfentanil (0.5 mg) was also given in case of restlessness. The anes-thetic depth was aimed to keep the patient at Ramsay sedation scale 5. After colonoscopy, patients recovered in the postanesthesia care unit. Patients were followed-up by telephone to assess possible discomfort and complaints 24 hours after discharge. Follow-up calls were made daily for the following 3 days. All colonoscopic procedures were performed by senior gastroenterologists with a minimum of 5 years of clinical experience.
2.3. Case screening
Screening for perforations was done by reviewing the International Classification of Diseases 9th revi-sion (ICD-9) codes. ICD-9 code 569.83 represents perforation of the intestine, accidental puncture or laceration during a procedure; code 998.2 repre-sents accidental cut, puncture or perforation; and code E870.4 represents hemorrhage during medi-cal care by endoscopic examinations. iPrice software (Informa Research Services Inc., Calabasas, CA, USA) was used to design the payment system and these payment codes were used to determine which cases were done under anesthesia. In addition, the med-ical history of the screened cases was reviewed to exclude episodes unrelated to colonoscopy.
2.4. Statistical analysis
Poisson’s distribution was used to determine the relationship between pairs of variables. Values of p < 0.05 were considered statistically significant. All analyses were performed using SPSS version 11.0 (SPSS Inc., Chicago, IL, USA).
3. Results
A total of 5549 men and 3952 women undergoing colonoscopy were reviewed. The mean age was 51.17 ± 12.03 years (range, 22−74 years). No differ-ences in demographic characteristics were found between the anesthetized and non-anesthetized atients (Table 1). The overall perforation rate was 0.03% (n = 3) in 9501 procedures, with one case in the non-anesthetized group (n = 1/2460), and two in the anesthetized group (n = 2/7041). No significant difference in perforation rate was found between the two groups (p = 0.6173; risk ratio = 0.6988; 95% confidence interval by Byar’s method = 0.063−7.705).
All three perforated cases received diagnostic ex-amination alone, which was performed by different gastroenterologists. The case in the non-anesthetized group was a 60-year-old man with a history of hy-pertension. The perforation was found 1 day after the procedure with the chief complaint of abdomi-nal pain. A computerized tomographic scan of the abdomen confirmed a perforation at the sigmoid colon. Laparotomy was done and he recovered smoothly and was discharged 7 days later. The first case in the anesthetized group was a 59-year-old woman with a history of abdominal total hysterec-tomy. This patient suffered from bowel bleeding and perforation at the rectosigmoid colon after en-doscopic polypectomy with biopsy forceps. Emer-gency exploratory laparotomy was performed and she was discharged 9 days later without adverse events. The second case in the anesthetized group was a 59-year-old woman without remarkable med-ical history. This patient complained of severe ab-dominal pain 2 hours after the procedure without operative measures. Abdominal ultrasound revealed massive bleeding and a suspected perforation. Emer-gency exploratory laparotomy was carried out and the bowel perforation was found at rectosigmoid junction. She was discharged 8 days later without any sequelae.
4. Discussion
Our results showed no significant effect of anesthe-sia on the incidence of bowel perforation based on a large sample of colonoscopy cases. Our perfora-tion rate (0.03%) was comparable to previous re-ports of diagnostic colonoscopy (0.03−0.65%).3
Indications for anesthesia in a patient undergo-ing colonoscopy include: (1) unresponsiveness to colonoscopic injury; (2) cardiopulmonary adverse effects;13 and (3) apnea with upper airway obstruc-tion during the procedure. Risk for adverse outcomes is low under carefully titrated anesthesia. Anesthesia could attenuate the responses to pain during colo-noscopy, but it does not necessarily increase the perforation rate, as revealed by our results, which are contradictory to the prevailing belief of many gastroenterologists.
As previously reported (Eckardt et al,13 Takahashi et al,14 Ristikankare et al,15 and Thiis-Evensen et al16), patients can withstand colonoscopy procedures with-out anesthesia. Nevertheless, many patients prefer anesthesia to reduce pain and anxiety during the procedure.17,18 Pain induced by the colonoscope is multifactorial and it is not easy for gastroenterolo-gists to manage the pain without analgesics or an-esthesia. The study by Takahashi et al14 showed that people with lower body mass index and of younger ages were more prone to report pain during colo-noscopy. In addition, inappropriate insertion time, poor preparation status, previous hysterectomy and the use of antispasmodic agents are other factors that may cause pain during the procedure.19 One perforation case in our study might have been re-lated to a previous hysterectomy and caused tech-nical difficulties during the colonoscopy. Although some people can cope with moderate to severe pain due to colonoscopy, anesthesia provides an anxio-lytic effect, amnesia,20,21 and reduced cardiovascu-lar stress, all of which are beneficial to patients from a therapeutic point of view.22
The drugs used in our study were propofol and midazolam, with or without alfentanil. One study by Lee et al23 showed that using in vitro preparations, propofol inhibits spontaneous and Ach-induced con-traction of human colonic smooth muscles. Opioid receptors are also known to modulate colonic func-tion. For example, delta 2 opioid receptor agonists are potent inhibitors in human colonic circular muscle24 and kappa opioid receptor agonists might cause dose-dependent attenuation of the presser and visceromotor responses to colorectal disten-sion.25 Benzodiazepines have hypnotic, sedative, anx-iolytic, amnesic, anticonvulsant and centrally elicited muscle relaxant properties. All of these drugs can reduce colon spasms during colonoscopy.
Although colonoscopy is a relatively safe proce-dure, bleeding and perforation are the two main complications. Bleeding can be treated conserva-tively but perforation usually requires surgical inter-vention. Three possible mechanisms are responsible for colonoscopy-related bowel perforation: (1) direct mechanical trauma by the colonoscope or biopsy forceps; (2) barotraumas from overzealous air insuf-flation; and (3) invasive therapeutic procedures.26 The skill level of the gastroenterologist also affects the completion rate of colonoscopy.27 Perforation of the bowel could be due to mechanical trauma by the shaft of the colonoscope or by excessive pneu-matic stretching during advancement and rotation of the scope.28,29 Perforation during therapeutic colonoscopy usually occurs in heat biopsy or polypec-tomy, during which thermal injury at the operative site is also likely to occur. This may cause a smaller injury with less contamination in comparison with mechanical tears during diagnostic colonoscopy.30
Symptoms and signs range from absence of, or localized, peritoneal signs to persistent abdominal pain associated with hypotension and decreasing hemoglobin level. Treatment includes conservative therapy and operative management. Conservative therapy may include bowel rest and intravenous an-tibiotics. Although this issue remains controversial among surgeons, one study by Ker et al31 concluded that the method could be safe and cost-effective, with low mortality and morbidity rates. Operative management involves prompt surgical intervention (e.g., abdominal operation) and laparoscopic repair of the colonic perforation. Laparoscopic repair of colonic perforation is widely accepted and benefi-cial to the patient because the procedure is rela-tively safe, and can be used for large perforations, in addition to reducing postoperative colonic ste-nosis.32 Less invasive conservative management has re cently become indicated when the patient has good general health status in the absence of peritonitis.
This study is a retrospective analysis of case re-cords. Although the search for perforation cases was done as thoroughly as possible, some cases may have been missed. For example, some patients might re-ceive emergency exploratory surgery at another hos-pital after the end of our follow-up period (usually 3 days). Furthermore, minor complications were not considered in this study, meaning we may have overlooked some anesthesia-related complications. Nevertheless, perforations during colonoscopy are multifactorial. Polypectomy, biopsy, anesthesia, pre-vious surgical history and skill level can affect the perforation rate. Further improvements of tech-niques require prospective studies using analysis of covariance with close follow-up of patients after colonoscopy.
In conclusion, anesthesia does not increase bowel perforation rate in colonoscopy. Anesthesia should be considered as a safe and well accepted proce-dure together with colonoscopy.